The assembly layer is going away.

The assembly layer has been around since the beginning of digital product development. For engineers, it includes things like writing a function that calls an API and stores the results in a database. For PMs, it includes culling through experimentation data and collecting significant results into a slide deck. For designers, it includes specifying a button hover state for the thousandth time. For researchers, it involves going through open-ended survey results line by line and tagging each result.

“White-collar assembly work” is the digital equivalent of all of the various trades that go into building a house. It is not the conception and planning of the house, but rather the drywall, flooring, electrical, and dozens more essential, skilled functions. This work is noble and important, but in the digital world, a lot of it is rapidly being eaten by AI.

The next question to ask yourself then is am I an assembler?

I would argue that at least in design, most of us are some percentage assembler. It could be 5%, it could be 100%, but it’s rarely zero. I run the largest design team at Microsoft AI, and even I spend some of my time nudging things around in Figma, trying to make decks look sharp, and performing some truly lizard-brain level data entry.

Orchestral Manoeuvres in the Dark

In order to prepare for a world where the nature of the work is changing so dramatically, Microsoft and many other companies are moving towards more of an Individual Contributor empowered environment. Some people assume this means all managers are going away. This is not the case. What it actually means is:

1. A true dual-track career path. Tech companies have been saying managers and ICs are rewarded evenly for 15 years now, but it’s often not the case in practice.
2. A requirement that every hire, whether IC or manager, is deeply involved in the product-making details.
3. A shifting of decision-making to the people closest to the work, where context is richest.

When people think of an orchestra of Individual Contributors, they picture everyone as a violinist. Perhaps you, as a lifelong IC, consider yourself a violinist.

What it really means though, is that everyone is a conductor now. It is your job to wave your arms expertly, and the machines will play the violins. As many violins as you want. And trumpets. And pianos. And sousaphones. Even in the dark, while you sleep, if you so desire.

What some companies are getting wrong right now is assuming those conductor jobs are mainly for PMs. Perhaps that is what’s behind this odd looking State of the 2026 Job Market survey. The data comes from a company called TrueUp, which I hadn’t heard of until today, but that’s probably just a me thing. Perhaps the data reflects reality, or perhaps design jobs aren’t accurately tracked by this company, but either way, this is not what I or a lot of my colleagues at other companies are seeing. If anything, most cross-functional teams are more underwater on design than on other functions.

Don’t You Forget About Me

Regardless, one thing that has absolutely changed is what we look for in the designers we interview. If you are in the job market now or will be at some point in the future, this post is designed to help you shape yourself into a better candidate for the design jobs of today and tomorrow.

As I first mentioned a year ago in The Future Favors the Curious, we are at a moment in time when orientation towards the future may supersede hard-earned experience with the past. Decades of wisdom in design can be a wonderful foundation on which to build even more expertise, but it can also mean unhelpful dogma about what the product-making process must look like. What’s helpful? A chicken-sexing like feel for identifying experiences that will and won’t resonate. What’s unhelpful? An insistence that static redlines be scrutinized by multiple levels of leadership and then handed off to engineering.

Some parts of AI-native product-making require rewiring your brain to know what is helpful and what is wasteful. How often have you seen a flawed project that an entire cross-functional team has worked on for weeks or months only for you to say “ugh, we could have saved thousands of person-hours if you did X, Y, and Z earlier”? How would your reaction change if you saw the same work but an enterprising designer got it working with live code in a single afternoon? “What a giant waste” turns into “That’s a helpful starting point!”

I should mention that the goal here is not to build a company full of product-makers who collaborate entirely with their computers and not their teammates. Ex-colleague Jess Rosenberg wrote a great piece that you should read in full. My favorite passage:

“I do think there’s something worth mourning in the transition from ‘we’ to ‘I.’ The solo builder, empowered by AI, is a kind of miracle and also a kind of loss. The miracle is obvious: democratized creation, reduced gatekeeping, the long-tail flourishing of ideas that never would have found support in the old institutional structures. The loss is subtler: the slow atrophying of the collaborative muscle, the gradual forgetting of how to think with others, the replacement of human to human dialogue & conversation with prompting.”

Let us not sit idly and create environments where people talk passionately to their computers and only reluctantly to their teammates.

Let us also not turn ourselves into output-obsessed monsters, spending 16 hours a day spinning up as many agents as we can, unable to sleep without wondering whether we could burn even more tokens on some other phantom obligations. AI Pyschosis appears to be real, and it’s important we modulate ourselves away from it, towards balanced lives in the real world. The goal isn’t to work 9am to 9pm, 6 days a week forever. It’s to work a healthy workweek — whatever that means to you — and have microchips carry out your well-conceived plans asynchronously. Don’t forget that the machines are supposed to work for us, and not the other way around.

Everybody Wants to Rule the World

So what has changed in the candidates we look for today? You probably guessed it, but it comes down to more conducting and less assembly. Essentially, can you guide a small group of humans and a large group of GPUs to produce the very best experiences in the world in timeframes unheard of until 2026? A year ago, it was enough to simply be experimenting and comfortable with LLMs. Now, it’s hard for me to imagine hiring anyone who is still just dipping a toe in. The uplift in superpowers is just too great to keep doing things the classic way. If you’re a designer and you bristle at this, think of how our friends in engineering are feeling. We may not even be looking at code at all in two years, but at least we’ll always be looking at, critiquing, and playing with design.

Aside from conductor skills, we’re also focusing on people who:

1. Take token design just as seriously as pixel design.
2. Are comfortable in engineering-dominated environments like VSCode and Terminal.
3. Don’t wait for others to tell them what to do.
4. Push the frontier forward, rather than just ape what very-online trendsetters are doing.
5. Make working software, not just the blueprints for it.

On that last point, Joel Lewenstein, my counterpart at Anthropic, had a really great answer to a question on the By Design Podcast about design being “downstream of engineering” at most companies. Joel said:

“Actual usable software is the lingua franca of Anthropic, and whoever can make that drives decision making, ideation, and roadmaps.”

Traditionally, the people who have ruled the tech world have been mostly engineers and very technical PMs, but leading roles are now available to literally everyone. Designers are in one of the best positions to raise their influence under these new conditions.

Let’s Dance

So how do you, as a design candidate, stand out and prove you have these modern skills that companies are looking for?

1. Do a self-assessment to see where you might be behind. Don’t rely on hiring managers to see your hidden potential. Other candidates for the roles you’re applying to are likely including entire working applications they built, not just static mockups of projects from five years ago. Don’t fret about who might be more advanced than you. There will always be someone. Just make sure you are comfortable with the new instrument panel and have some tangible artifacts to prove it.
2. Rebuild your portfolio site if necessary so it contains a very small number (3-5) of things you’re proud of. Using the LLM of your choice, you can rebuild your portfolio in a single night. You should probably spend more time on it than that, but the days of procrastinating weeks of handcrafted new HTML/CSS/JS are over. Coaxing an LLM to present your portfolio in a compelling way is itself proof of fluency.
3. Follow the tried-and-true format of presenting design work. What problem were you trying to solve? How did you know it was a problem? What variety of solutions did you try? What solution did you land on? Was it successful and if so, how did you know? This format is straightforward and hiring managers are used to evaluating it. It will ensure the depth of your process is communicated.
4. Many designers, whether right out of school, or 30 years into their careers, haven’t done “fully AI-native work” at their companies or on real projects yet. Don’t let that stop you. Build a casual game. Create a single-purpose app for a problem you’ve always wanted to solve. Hiring managers would rather see thoughtful, modern work on small or even imaginary projects than unimpressive work for billion dollar companies. Show what you can do, not just what the people who currently pay you ask you to do.
5. Don’t be shy about applying for positions you “are not qualified for”. Many companies (including us) post more positions at the Senior and Principal level because our systems follow an inflexible “one role, one level” policy, and we’d rather see seasoned candidates if we are forced to choose. That said, I’d hire a fresh-out-of-school designer with a great start and a demonstrably high ceiling over someone with 30 years of experience building unimpressive things. If your stuff is good, trust me that no one is going to laugh at your application. If anyone, you’ll probably stand out even more.
6. As always, applying cold is worth a shot but rarely the best way in the door. If you see a position or a company you’re interested in, find out exactly who the hiring managers are, and figure out a way to be helpful to them. Maybe you have an idea for a new feature or a way to fix a broken part of one of their products. Produce a working prototype of your solution. Maybe you are organizing an event they might feel honored speak at. Introduce yourself and ask them. They may be too busy to say yes, but the point is to break through the pack. Your ability to stand out with a hiring manager is really the first test of your creativity, and you’d be surprised at how many of your co-applicants (99%?) barely even try.

With those tips in mind, here are some of the many roles we have open at Microsoft AI right now:

• Principal Product Designer – News & Content Team (Puget Sound Preferred)
• Sr. Product Designer – News & Content Team (Puget Sound Preferred)
• Principal UX Research Manager (Puget Sound or Mountain View)
• Principal UX Researcher (Puget Sound or Mountain View)
• Principal Product Designer – Copilot Team (Mountain View)
• Senior Product Designer – Copilot Team (Mountain View)

Should I Stay or Should I Go

In trying to decide what legendary ’80s song to end this little post with, I think The Clash sums up the psyche of the modern designer best:

“If I go, there will be trouble. And if I stay, it will be double.”

A lot of us look around at all of the potential damage AI can cause and our first reaction is to turn away from it. Every time another slop cannon publishes a piece about how much more prolific they’ve gotten since getting the agents to fight each other, The Way of the Luddites seems more attractive than ever.

Then you think back to all of the previous inventions that threatened designers — lithography, the Linotype machine, Photoshop, the internet itself — and you remember this train ride has always been bumpy. Diving into the next part of the journey with both feet isn’t going to speed up or slow down the train, but it might end up taking you to places you never imagined… and that, for now, is worth the ride.

In 1995, halfway through college, I determined I was going to be a designer. At that time, 99.99% of “professional designers” had no experience designing anything for the internet. You could have tasked some of the most accomplished designers at the time — Jony Ive, Paula Scher, David Carson, Paul Rand, Massimo Vignelli — with creating a simple 468×60 banner ad and none of them would even know what you were talking about.

Over the next five years, the world would discover what banner ads and the internet were, but most experienced designers stayed put in print, television, or whatever field they were already comfortable with. The designers who would go on to help shape the internet were from two groups:

• People brand new to the industry
• People who looked forward to starting over and getting good at something new

I was somewhere between the two groups, having spent a few years in print but leaning more into a childhood love for new technology than anything else.

Within a few years, the demand for internet-native designers exploded and created the thriving job market we have enjoyed almost uninterrupted until now. Analog design still produces some of the most amazing creative work on the planet, but the growth of that part of the profession has not been the same.

Running design at Microsoft AI and having done a decent amount of hiring lately, I can tell you that the patterns emerging now are exactly as they were in 1995. There is a giant population of designers who have a bunch of really great skills. Some of those designers will decide they are content doing the same sort of design they have done for their whole careers. Others will decide to learn as much as they can about AI and prepare for an industry that will look very different in 5 to 10 years. Finally, there is another group of people who have no design experience whatsoever but are so enamored with this new technology that they will teach themselves very useful skills in a short amount of time. Do not underestimate this third group as it’s easier than ever to fake it ’til you make it right now.

If you are content in your comfort zone, that’s perfectly fine. If, however, you are looking to lean into what’s next, here are some suggestions.

You are not (far) behind

One thing I tell people who are trying to get into running is “you are always only a month away from being in decent running shape”. What I mean by that is, your first run is going to be 5 minutes long before you tap out, but if you do it every day for a month, you’ll be able to run at least a mile or two.

I see so many people trying to fake their AI experience by adding those two letters to every job they’ve had for the last 10 years. You are not fooling anyone. 99% of designers’ “AI experience” has come in the last two years. You will very occasionally find someone who was super early, but even that doesn’t mean much, as they might not be very good. It’s like saying in 1995 that you had a Commodore 64 growing up. Cool, but not exactly an endorsement of your actual design skills.

Someone may have told you to pepper your résumé with references to AI so it gets past certain filters, but we run a big operation here at Microsoft with thousands of designers and we don’t filter like this. Furthermore, when I see someone doing this, I actually take it as a sign that I should be suspicious of what else is on their résumé. When I was in college, I was a rampant gilder of my résumé, so I understand the temptation. “Salad Bar Director” did not actually get me any jobs though. Don’t do it. Just be honest. In terms of calendar years, your competition doesn’t have much more AI experience than you do. The clock is ticking though, and designers are already learning to do extraordinary things each time a new model or technique emerges. Pay attention. Learn by doing.

Know how our field is changing

For as long as design has been a profession, there has been a debate about the relative value of generalists vs. specialists. My view has been that in digital product design, you usually want generalists, but every so often, there is no substitute for a specialist. Some people are looking at AI as a specialty right now, and that may be true in some parts of engineering like model training, but for design, it’s a tool that will make generalism more common and accessible. Even pre-AI, I have always preferred hiring designers who I can give an entire project to. I want people who can frame up a user problem, explore the solution space with functional prototypes, produce high fidelity production designs/code, and also be involved in the testing and refinement of said solutions. This is a LOT to ask of most designers, but the best generalists can usually play all four quarters of the game.

With the help of AI, there are fewer barriers to designers participating in all parts of the product making process. Want to quickly research a problem you suspect exists in the world? There are now tools to help you write and analyze that survey. Want to quickly mock up a dozen high fidelity interfaces just to get a discussion going? Now you don’t have to spend days on that. Need hundreds of illustrations in a consistent style? Plenty of tools for that. And perhaps the biggest barrier for some designers over the last couple of decades: turning your ideas into functional code. That future is now here.

For all the excitement about these advancements, they also raise uncomfortable questions. The most obvious is: do we need fewer designers now? To that, I would say yes, we need fewer designers (and engineers, and everyone else in product development) for the same amount of output. Do we need fewer designers overall though? That remains to be seen. The growth of the economy since the beginning of time has been based on productivity increases leading to greater total output of our population. So if you think we have already maxxed out our total output of products and services in the world, I would expect a nosedive in the amount of designers and engineers needed. But if you think we are only scratching the surface of product development, you should expect a future where millions of designers and engineers continue to do great things but much more prolifically. Think of the farmer 1000 years ago who could feed a few families vs. the farmer of today who can feed an entire town.

In order to prepare for this, you should anticipate companies initially employing fewer designers (and engineers, and PMs) to do the same amount of work they were doing yesterday. Managers aren’t going completely away, but organizations will flatten somewhat as companies concentrate their budgets on people close to the metal. As my colleague Jon Friedman said the other day, “there will be people directing AI, and people directing people and AI”.

Companies will increasingly ask the question they should always be asking regardless of technology trends: what is the smallest team of people I need to create the best version of this product? The difference this time is that we now have a transformative technology that can turn an inefficient team of 3 designers, 2 researchers, 2 PMs, and 50 engineers into a one-pizza squad of 1 designer, 1 researcher, 1 PM, and 7 engineers. Furthermore, the lines between job functions are blurred such that each member of that team can do some of what another member can. Engineers can help with design, design can help with research, research can help with PM, etc. There will even be one-person startups where the same person does everything and eats the whole pizza!

Another way our field is changing is that the term “product designer” will encompass more of the humanities than it did before. Our currency used to be mostly pixels, but now it is also words, personalities, and other emerging areas of focus. Content Designers and UX Writers may argue this has been the case all along, but until now, the market for pixel-focused designers has been much stronger. The best thing you can do to prepare for this expansion of design’s role is to poke around and get comfortable with all of the new powers you have at your fingertips. Go “program” the personality of a GPT for an hour. The process requires nothing more than plain English.

Learning is free

When I got into this industry, you needed to buy books and take expensive classes in order to learn how to use tools like Adobe Illustrator. There is nothing about designing for AI that requires you to spend significant money to learn. It’s all about the time you are willing to invest.

Furthermore, learning about the general field of design requires years of getting familiar with hundreds of concepts, but AI is such a nascent technology that you can count the new essential design skills on two hands right now. Image generation, prompting, evals, generative UI… there aren’t thousands of AI-specific concepts yet, so get started now.

From an employer perspective, I also don’t want to hear that you want a job here so you can start getting into AI. Just start. The most impressive applicants I’ve seen are people who have, on their own, created personal experiments with the sole purpose of learning the technology. I don’t need to see that you have used AI to help a multi-billion dollar company increase its profits by 200%. I’m also impressed seeing a series of album covers you made for your garage band using Midjourney, or how you have used a combination of your design skills, Recraft, and Lovable Dev to build a site for your non-profit in a weekend. There is no better quality in a teammate than initiative, and by showing you’ve cleared your own trail into the world of AI, you’ve shown me you might be curious enough to be an industry great one day.

Aside from understandable demands on your free time, there is nothing structural holding you back from knowing just as much about designing with AI as your competitors in the job market. If you’re looking for a couple of good newsletters to get started with, try Heather Cooper’s Visually AI or Xinran Ma’s Design with AI.

Don’t get discouraged by the hunt

I’ve worked at some of the smallest and biggest companies in the world and let me just tell you that we all have significant problems with our interview processes. If you get rejected by a company, it might be you, but there is just as much of a chance that it’s them. We’ve said no to hundreds of incredibly qualified candidates for a number of reasons, and only a fraction of the time is it because someone screwed up in their interviews. More often it’s because we are choosing one person between many qualified candidates. Or because we had to pause our process for any number of reasons completely unrelated to the candidates. Or because we screwed up in our evaluation of someone’s strengths and weaknesses.

If you’re not getting calls back from your applications, it could be because your portfolio needs work, or it could be because your approach needs work. More on each of those below. If you aren’t getting selected after interviewing, it could be because you truly didn’t interview well, but more likely it is because one person interviewed better than you did. That’s all it takes. Even at a giant company like Microsoft, it’s usually one person chosen per position. In other words, we are rarely able to hire every finalist who was qualified.

Get your book in order

These days, hiring managers will often pass judgement on your portfolio in less than 30 seconds. When you have 50 portfolios to go through in an afternoon, there are all sorts of things that will put candidates in the “No” pile quickly. If I can’t figure out how to navigate your site? No. If you have basic errors which show a lack of attention to detail? No. If I need to type in a password that you haven’t provided? No. The best portfolios show off a small handful of projects (3-5) straightforwardly, but deeply. Tell me:

1. What problem were you trying to solve?
2. Why did you think it was a problem? Research? Gut?
3. How did you explore the problem and solution space?
4. What did you end up with as a solution?
5. How did you measure if you were correct?

Importantly, remember to show a taste of #4 first for those of us with very short attention spans or a ton of stuff to get through. Wow me with the end result first and then show me how you got there.

Getting in the door is the same as it ever was

When I was in school, I remember being determined not to get jobs because of who I knew. I only wanted to get hired based on my portfolio. Only a couple of years into my career, however, I realized that that simply isn’t how the world works, and there is very little you can do to change that. There is no substitute for a trusted reference vouching for how valuable of a teammate you are. Do the social archeology to figure out the shortest line between you and someone close to the hiring manager. The reason this is so important is that what’s in your portfolio is only half your story. What you are like to work with is the other half, and to determine that, there is no better source than others who have worked with you.

To show you how important networking is: I’ve been lucky enough in my career to never be unintentionally unemployed. Every time I’ve left a job, it’s been on my own volition, and every time I’ve re-entered the job market, I’ve lucked out with multiple options. After I left my last job in 2022, I took a few months off before peeking my head up again. Within a few weeks, I ended up getting to offer stage with 5 companies, including Microsoft. I happened to be on LinkedIn one night and randomly saw a Head of Design post for Major League Baseball’s digital division in my feed. There was a one-click “Apply” button, so I hit it, just for kicks. 24 hours later I got a canned rejection notice. This whole paragraph contains enough cringey braggadocio already, but having specifically designed for ESPN, MLB, NBA, NFL, NHL, the Mariners, Twitter, NBC, and even having a current offer in hand from Disney/ESPN, you’d think I would have at least gotten a call back! Nope.

The point is: no matter how perfectly matched your qualifications might be, applying to a position cold is like buying a lottery ticket. You might get lucky, but the chances are incredibly small.

Creating opportunities for your future self starts many years before you need your next job. Build relationships inside the industry, especially with recruiters. I always shake my head when I see people in tech speaking ill of recruiters. Sure there are some bad ones, but in design, the best recruiters are often in charge of filling the best positions. Treating them badly is not only unkind, but it’s also specifically bad for your future self. Whether you are an early career individual contributor or a longtime design executive, make an effort to know the very best design recruiters in the world.

If you think that you, a member of the creative class , are somehow above the world’s best talent matchmakers , think again. They have furthered the careers of thousands of people who are better than both you and me… and the best ones are some of the nicest, most interesting people on the planet.

These are the fun times

When it comes down to it, your future in design is the sum of all of your actions that got you here in the first place. The skills you’ve built, the artifacts demonstrated in your portfolio, your helpfulness as a teammate, your reputation as a person, and now more than ever, your curiosity to shed your skin and jump into an undiscovered ocean teeming with new life, hazards, and opportunity. Someone will invent the next CSS, the next Responsive Design, the next sIFR, the next TypeKit, the next IE6 clearfix, and the next Masonry for the AI era. That someone might as well be you.

These periods of technological turnover are the most exciting times to be a designer. They are when we get to flap around chaotically and create the interaction patterns for the next couple of decades. Dive in while you can, because in a few years, we’ll all be back to making the same things again.

For now, the future favors the curious.

I usually write a lot about the companies I work at but have held off until now because we haven’t been hiring. Well now we are! We’re specifically looking for Designers and UX Engineers to work on our design system for Copilot. Some of these positions are on my team and based where we have offices (Puget Sound, the Bay Area, Atlanta, New York, Vancouver, Barcelona, Hyderabad, Beijing, and Suzhou) and some are on adjacent teams and can accommodate fully remote work. If you are a Designer or UX Engineer with a passion for design systems and AI, we’d love to chat.

So what has year one been like? The good and the “needs improvement”.

One of the reasons I decided to join Microsoft was I missed the joy of in-person product-making. I know not everyone feels the same way so I’m not trying to make any broad statements about local vs. remote work, but for me, it has been even more refreshing than I expected. I usually come in 3-4 days a week, while others on the team are anywhere from 0 to 5.

It’s funny, sometimes I will wake up on a Monday and think to myself “ahhhh, this is going to be a chill work-from-home day” and by the end of the day, I realize I’ve been staring into a screen on video calls for almost the entire day and how much that slowly saps my energy. Meanwhile, in-person days are filled with walks, whiteboarding, and energizing sessions with some of my favorite teammates I’ve ever had. I realize not everyone feels this way about being in the office from time-to-time, but I do. Even our fierce, interdepartmental karaoke battle helped bring a bunch of teams together who had never met before.

The other great thing about lucking out and joining when I did is that we are embarking on one of the rare paradigm shifts that occurs in technology maybe once a decade. The 1980s were about personal computers. The 1990s were about the internet. The 2000s were about smartphones. The 2010s were about the cloud. And the 2020s will be about AI. The really powerful thing about all of these developments is that they don’t replace each other, but rather they build on each other. AI is the result of everything that came before it. If you got into design to help shape the culture of the world around you, these are the moments you treasure.

These turning points are also wonderful because they give you a chance to reawaken to your Beginner’s Mind. I have learned more in my first year here than at almost any other time in my career. Not only does the pace of technological change in AI force you to build skills as you go, but there are so many amazing engineers, designers, researchers, writers, marketers and other creative people to learn from, that it happens almost automatically.

When I was interviewing here, not even my prospective new boss told me about any of what was behind the curtain. Only during my first week did I find out all we are working on to empower every person and every organization on the planet to achieve more. That is Microsoft’s mission statement, if you hadn’t heard it before. It’s an uncommonly good filter with which we can all ask ourselves every day “does this project actually do that?” It’s quite freeing as it gives you license to question projects at every stage of development.

Another thing I’ve loved about my first year here is that I joined a group that has figured out how to ship very quickly. That also has its downsides, as we have plenty of craft problems to solve, but it’s great working for one of the largest and most established companies in tech and being able to ship within weeks of designing something.

One more thing that’s blown me away is the Inclusive Tech Lab, where we work on new technologies to make our products inclusive to people all of abilities and walks of life. No one experiences technology the same way, and teammates like Dave Dame and Bryce Johnson do a ton of great work to make sure that’s top of mind for everyone.

Finally, one of the unsung benefits of working for a native Seattle company again is that Seahawks stuff is all over the place. Presentations, charity auctions, everyday office attire… you name it. It’s nice to not be the only one with good taste in football teams.

I am no corporate shill, however, so I must also be honest about some of the things that need improvement over here.

At the top of my list is that Microsoft has not yet fully embraced the role design plays at most other tech companies. We were engineering-driven in 1975 and we are squarely engineering-driven in 2023. The world, meanwhile, has changed in that time. It is no longer sufficient for complex things to work. They must also shed their complexity. People expect the products and services they spend their time and money on to delight. To overdeliver. To give them superpowers. Those sorts of qualities only materialize when you have supergroups building products.

In music, a supergroup is when a singer, a guitarist, a bassist, a drummer, a keyboardist (and so on) who are all at the top of their game come together to create an album. In tech, a supergroup is a researcher, a couple of designers, a product manager, and a Volkswagen Bus or two full of engineers. Up until a decade or so ago, a lot of tech companies followed the model of packing projects with as many smart engineers as they could find and only sprinkled in things like design and research as necessary. I still see some of this thinking in pockets over here. I’m trying to influence things, but it’s a delicate dance, especially when the company has had such enormous success by doing so many other things very well.

I think there are plenty of people here who still feel like being engineering-led is unequivocally good, but to those people I would say that in a modern tech company, design is engineering. It’s no better or worse, but it does have very different leverage in the building of a product.

On the plus side, Microsoft has never had as much design and research talent as it has right now, and we are increasingly looked to by the executive leadership team as lighters of the path. When we go into high-stakes meetings, we always go in with pixels and prototypes, which are uniquely good at cutting through bullshit and ambiguity. As a wise person once said, a prototype is worth a thousand meetings.

There are a ton of amazing designers and researchers who have been here for 10 and even 20+ years whose hard work has led to this moment of evolution for the company, and every day I am in awe of their perseverance.

Next on my list is how often we get in our own way with “procedural goo”. I’ve worked in plenty of large companies including Twitter, Disney, and MSNBC, and have never seen the level of approvals, paperwork, and rules that get in the way of speed and autonomy here. Just transferring someone from within my own org to a slightly different role within my own org took a dizzying amount of effort. At most companies, this would have been about 10 minutes of work: one minute from me and nine from someone in HR trying to navigate to the right screen in Workday.

Then we have acronyms. My GOD do we have acronyms. I actually liked acronyms before I got here! I usually think they are cute. After seeing a new one almost every single day since getting here, I have resolved to never use them either inside or outside of work. I even say “Cyan Magenta Yellow Black” out loud if I have to!

Finally, the last thing on my list — and this is where you come in — is dedication to craft. It is so tempting to try and “science” your way into viable products these days. Build the beginnings of a customer base through rudimentary product-market fit, and then fastidiously optimize your funnel, your game mechanics, your viral loops, your push notifications, and so on and so forth. These gains are not always easy to come by, but they are rooted in ruthless experimentation and allegiance to short-term data. Our north star is at least pretty pure — Daily Active Users — and that metric is usually a good indicator that you’ve made something people like, but doctrinaire allegiance to almost any singular metric can quickly make people forget why we are in this profession to begin with: to improve lives. Or to put it squarely in Microsoft parlance again: to help every person and organization on the planet achieve more.

If you ever find yourself asking the question “how can we increase Daily Active Users?” instead of “how can we make our product better for people?”, you’ve already lost. Metrics are trailing indicators of qualitative improvements or degradations you’ve made for your customers… they are not the point of the work.

Recently, we’ve made some excellent strides in prioritizing qualitative product improvements even when they fly in the face of metrics we care about, and it’s really gratifying to see. It reminds everyone that a product is the collision of thousands of details, and the crafting of these details requires taste.

Designers are often looked to as “owners of craft and taste”, but craft is very much a team sport. It’s not just how things look and feel but also how they work. I very much like how Nick Jones (channeling Patrick Collison) at Stripe put it in this video:

“What we put out there should quite plausibly be the best version of that thing on the internet“.

To do this takes an unbroken chain of excellence:

1. An idea that can improve lives if executed well.
2. Foundational research to light the path before design and coding begin.
3. Rich design explorations and prototyping to make the experience palpable.
4. Buy-in to build it at a level of quality that makes the team proud.
5. Impeccable UX engineering and UX writing to make sure every detail is dialed.
6. Well-conceived server-side engineering to make it scalable and maintainable.
7. Creative marketing to prime people for the experience.
8. … and finally, maybe more important that anything else on this list, the will to keep refining relentlessly after the experience is launched. This part is so often neglected as companies rush to build more things.

Some people would look at this list and think “yep, makes sense”. Others would look at it and think “sounds slow and not very agile”. The trick to balancing this level of quality with speed of development is realizing that it’s often more efficient to experiment in step 3 than it is in step 6. This is why so many modern tech companies realize that hiring more designers and researchers doesn’t waste time and money… it saves time and money. With less than a week of one designer’s time, we can produce a wide variety of prototypes to test with real people. Just recently, we created an entire GPT-powered research application without even bothering a single engineer.

Design is engineering.

Finally, a brief note about prototyping. I would argue that the most impactful innovation in the craft of product development over the last 20 years has been the rise of rapid design prototyping. Prototypes that demonstrate an experience are useful not just in usability testing, but also in selling ideas up and across the organization. Engineers hate working on things that haven’t been thought through or “appropriately politicked” yet, and if you can bring them a working prototype that has already been vetted with users and various stakeholders across the company, they will love you for it and work hand-in-hand with you to get every detail right.

Design prototypes are the currency of a high-craft, high-speed product development organization, and they are increasingly the currency of our team.

Alright, back to the hiring. I plan to hire against the entire growing list of products our team is responsible for: Copilot, Windows, Edge, Bing, Start, Skype, SwiftKey, and so much more… but for now, this is a concerted hiring effort centered around Designers and UX Engineers to help build out the emerging design system for Copilot and our suite of AI-powered products.

If this is you, please have a look at the following roles we’ve just posted:

• Designer II
• Senior Designer
• Principal Designer
• UX Engineer II
• Senior UX Engineer
• Principal UX Engineer

We’d love to work with you on the future of design systems at Microsoft!

In confident pursuit of that feeling, I’m thrilled to be joining Microsoft to run Design & Research for their Web Experiences organization.

I was Microsoft-adjacent 10 years ago at MSNBC.com in Building 25, but this will be my first time as a blue badge, so to speak. I’m also thrilled to be joining Liz Danzico and John Maeda, who have also started at Microsoft in the last several months. I’ve known them both for a long time and have wanted to work with them forever.

There are several things which drew me to this opportunity, but at the top of the list is the people. Not just Liz and John, but the thousands of teammates in Seattle, Vancouver, Hyderabad, Barcelona, Beijing, and many other cities. There are certainly some great solo efforts in tech, but almost all of the best work I’ve been around has been the result of getting the right people jammin’ with each other. In my first several days here, I’ve already met so many of those people, and I can’t wait to continue the momentum Albert Shum created by carrying the torch for one of the best Design & Research teams in the Pacific Northwest.

The second thing I’m super excited about is the scale of the work. I’ve worked on the largest sports site in the world and one of the largest social networks in the world, but the properties in this group reach well over a billion people. Between the (refreshingly fast!) Edge browser, MSN, Bing, and several other products, Microsoft has quietly built up one of the top five properties in the world in terms of traffic and reach. They’ve also done it with humility, knowing how far they are from being perfect. I also love that so many of these products can and will be so much more as we begin to use some of the technology that’s emerging within Microsoft. In my first two weeks, I’m already overflowing with ideas.

Finally, the other thing I’m most excited about is getting back into the office in a flexible hybrid environment. I’ve worked in-person for most of my career and remotely for the last four years, and where I’ve landed is that everyone’s preferences are different, and it’s just tradeoffs all the way down. Where you land depends on a mix of your personality, your life outside of work, and what type of job you have. For me personally, I very much like being around people and feel like I do better work when I am… but I also like getting back my commute time a couple of days a week and making daily jogging more convenient. Also, Henry (pictured above) enjoys the extra lap time. Microsoft’s hybrid policy is a nice balance, and it sounds like exactly the right setup for someone like me.

Speaking of the exactly right, this also feels — for me at least — like exactly the right time to join Microsoft. The company has been through several distinct eras over the decades, but this feels like the era of re-commitment to the planet. To customers delightful experiences, to employees a great environment, and to the natural world, a smaller and eventually negative carbon footprint.

It’s a new year, and I couldn’t be any more here for it!

One of our generation’s greatest and most influential designers hasn’t slowed down as much as he has simply shifted focus. A great example of splitting your life into chapters and knowing when it’s time to explore the next one.

As someone who competed against Figma for a couple of years, I want to talk briefly about what makes them so hard to catch, and why I think Adobe ultimately decided they would never beat them:

Performance.

Even though I have spent over 20 years in the design industry working directly on consumer products, I never fully appreciated the importance of performance until working in the design tools industry.

Most digital consumer products are used in short bursts over a long period of time. Think about the Amazon app on your phone. You open it maybe once a week, peck around for what you need, hit Buy Now, and you’re on your way. If there is a two-second lag between purchasing and getting your confirmation screen, you don’t even think twice about it. Even in the case of an outright error, you just shake your head, hit reload, and things are usually fixed.

With professional production tools though — whether design, engineering, or otherwise — full-time craftspeople spend almost every hour of every work-week inside of your software. Every time something goes even remotely astray, it is noticed. Putting aside catastrophic stuff like data loss, even things like cursor lag, screen flicker, progress bars, and scroll/zoom performance are tiny paper cuts that form into pools of blood by the end of each day.

Figma did a lot of things right over the ten (yes, ten!) years they’ve worked on the product, but one thing they did that no one else has been able to replicate is meet and in some cases exceed native app performance inside of a web browser.

Nothing Figma has accomplished in the marketplace would be possible without this, and it is the thing that competitors have struggled the hardest to replicate. When you build software using native code, you get a lot of stuff for free. Need a scrolling list? Apple, Microsoft, and Google have multiple pre-built components you can use. Need to draw one semi-transparent shape on top of another? The system already knows how to render that. Need to optimize it all for speed? Most of that work has already been done.

Inside of a browser though, the work is rarely done for you. Even in instances where someone has already built a component, it’s often too slow or glitchy to use in a professional development environment. So what did Figma do about this? Over the course of several years, they:

• Built their own components and architecture painstakingly from scratch and never settled for “good enough”
• Worked with organizations like WebKit and Chromium to improve web browsers themselves (the benefits of which go beyond Figma)
• Detailed out in the open what they were doing and how

It was this last one that really made me see how wide the moat was for the first time. Normally companies keep their secret sauce secret. After all, why would you want to give your competitors any information that might help them compete? But a company who routinely publishes information that is useful to competitors? That is some confident shit right there. It reminded me of a tweet I can’t find from several years ago:

“The design tool war is already over, but no one knows it yet.”

Fast forward a few years and everyone who has tried to match Figma’s all-around performance has fallen short. Private companies like Sketch and InVision. Public companies like Adobe. It’s not for lack of effort by hundreds of incredibly smart people. It’s just really frickin’ hard. Combine that with the fact that Figma is a moving target who is now building entirely new capabilities, and you can see why Adobe decided this wasn’t just a move they wanted to make… it was a move they had to make.

… which brings us back to a lot of the reaction we are seeing on Design Twitter today.

I don’t think people mind the abstract concept of Figma being acquired by another company nearly as much as they mind the very real threat of Figma losing what makes it so special in the first place: focusing maniacally on performance, thinking differently, and optimizing for user experience above all else. The backlash is an expression of how a lot of people feel Adobe has done in those categories over the last decade.

If I’m Adobe, I am printing out as many Tweets from today as I can, making a book out of them, and then doing this:

After that, I’m letting Figma lay the tracks for the next decade of this industry and rallying the thousands of talented people at my own company to rethink how the entire organization builds software. Within the next several years, it’s going to be possible to go from idea in the morning, to prototype in the afternoon, to working code in the evening… and the company who can do that most thoughtfully is going to be one of the most important companies in the world.

Our first reflex was to head over to the Herman Miller site and see what colors were available. These chairs come in plastic or fiberglass, but you really want the fiberglass ones because they just look so much nicer. These fiberglass models went out of production in 1989 because of environmental concerns about the manufacturing process, but they returned in a more eco-friendly formulation in 2013. Herman Miller does not offer these new fiberglass models in many colors, but to our delight, they listed a Seafoam Green that looked great. We placed an order for six of them and thought we’d be on our way.

Unfortunately, after many months of delays (damn you, supply chain!), the chairs were still nowhere on the horizon, so we did what we should have done all along: look for vintage ones in good condition. After searching a bunch of different auction sites, we found Eames.com, a marketplace of used Eames items. They have a pretty good selection, and they happened to have six chairs in an Olive Green color that we liked even more than the Seafoam Green.

Their site does not list the United States as a shippable destination, but they agreed to ship us some chairs for about $400 each including shipping, which was quite a bit less than they would have cost from Herman Miller.

I thought I had lucked out until they sent me detailed photos of each chair, which revealed some scuffing and a few significant scratches. This is to be expected in any item that is 50 years old, of course, but I really wanted these things in primo shape.

After consulting with some friends, some people on Twitter, and some articles on the internet, I became confident enough I could restore these easily to almost new condition, so I placed the order.

Amazingly, the chairs showed up all the way from the U.K. in less than a week, and I got to work immediately.

Step one was buying a couple of sanding sponges. I bought a 320 grit and a 220 grit. I started with the 320 because I was worried about scratching the chairs, but that concern proved unfounded. It was a little too fine for the job. You really need to remove some fiberglass when you’re doing this, so 220 ended up being the right strength. If your job is really gnar, you could probably go a little rougher even.

Anyway, I took the chair shells outside to the patio, and put one on a towel to work on it. I used thick rubber gloves to make sure no fiberglass debris came in contact with my skin, but other than that, just a small bucket of water to wet the sponge was the only other tool. Some people also recommend wearing a mask, but I was outside and not doing any electric or dry sanding, so it didn’t seem necessary.

The process of removing even the deepest scratches was surprisingly quick. I probably spent 10 or 15 minutes per chair, depending on severity. It’s hard to tell if you’re completely done until the chairs dry back up, so I hosed them off, let them dry overnight, and then gave each one a touch-up sanding the next day, followed by another hose off and air dry overnight.

It’s really amazing how self-healing fiberglass is. You think you are just scratching the shit out of it when you’re sanding it, but then when you’re done, it all blends right in automatically. And that’s even before the next step, which is when the chairs really come to life.

The final step was wiping on some Penetrol. Penetrol is an oil-based paint additive that happens to restore fiberglass to brand new condition. It’s really amazing, and you should be able to get it at your local hardware store. I used one of these Swedish dishcloths we had laying around, but you can use any number of other lint-free cloths for the application.

This step was even easier than the sanding, with each chair taking no longer than a minute or two to cover in Penetrol. Once they are coated, that’s when your jaw really drops. Not only does the shine look spectacular, but it also further hides any imperfections that may have remained on the surface.

After letting the Penetrol dry for 24 hours, I inspected each chair. I probably would have been happy with just that one coat, but I did see a couple of areas where I had perhaps applied a tad too much, causing a bit of a run. Not a big deal, but I decided to just give them all another coat to smooth out the runs and also provide more protection.

One coat and 24 hours later and they were looking perfect! The last step was to assemble the new dowel legs. This proved to be the most difficult part of the whole project as each chair took about a half hour to assemble. Granted, I was watching football as I was working, so perhaps a focused person could have gotten it done quicker.

Anyway, below is the end result! We’re so happy with how these turned out that we are now looking for two more in another color for the ends of the dining room table. It’s great to, a) get something vintage, b) save money, and c) re-use an existing object rather than cause a new one to be manufactured. Win-win-win.

I encourage you to go on your own Eames restoration journey. It’s not very difficult, and the extra effort you put in will make you appreciate them even more.

I remember vividly the first time I saw Flash on a computer screen. It was 1997, I was finishing up college, and I had managed to teach myself enough HTML to think about pivoting from print design to interactive design as a career.

Web design, at the time, was a clumsy beast. Most web sites were essentially Times New Roman black text on a grey background with an occasional low-quality image here and there. The “design” part was often just figuring out how to best organize information hierarchies so users could feel their way around.

Once we got bored of basic HTML (there was no CSS at the time), we started doing unholy things with images. We’d set entire pages in Photoshop, slice our layouts into grids of smaller images, and then reassemble everything into a clickable mess. These were dark times.

My college, having invented PINE, was considered “on the front edge” of the internet at the time. Here’s is what our site looked like back then:

Even the most beautifully designed sites felt a bit lifeless, and once someone came up with a new layout that worked well, everyone would just ape it. To make matters worse, every new advancement in methods required more convoluted hacking to display correctly across Netscape, Internet Explorer, and every other fringe browser in use at the time. It was a total mess.

Here is the first version of Zeldman.com I could find, from 1998. Amazing for the era, and holds up impressively in a nostalgic, cyber-Americana sort of way, but you can see how limited we were by screen widths, color palettes, and layout technologies.

Then one day in 1997, I clicked on a link to Kanwa Nagafuji’s Image Dive site and the whole trajectory of web design changed for me. It looked like nothing I had ever seen in a web browser. A beautiful, dynamic interface, driven by anti-aliased Helvetica type and buttery smooth vector animation? And the whole thing loaded instantly on a dial-up connection with nothing suspicious to install? What was this sorcery? Sadly, I can’t find any representation of the site online anymore, but imagine the difference in going not just from black-and-white TV to color TV, but from newspaper to television.

Nagafuji’s work was such a huge, unexpected leap from everything that came before it that I had to figure out how it was done. A quick View Source later revealed an object/embed tag pointing to a file that ended in “.swf”. A few AltaVista searches later led me to the website of Macromedia, makers of ShockWave Flash (“SWF”), the technology that powered this amazing site.

I downloaded a trial version and was blown away at the editing interface. Instead of a shotgun marriage of Photoshop, HTML, browser hacks, and a bunch of other stuff that felt more like assembly than design, here was a single interface to lay out text, shapes, images, and buttons, and animate everything together into an interactive experience! It was magic.

After mucking around in the Flash editor (version 2 at the time) for a few hours, I did what every self-respecting web designer would do and immediately set out to find other cool stuff to copy. Over the course of the next several months and years I would find such gems as:

Yugop from Yugo Nakamura
Once Upon a Forest and Praystation from Joshua Davis
Nose Pilot by Alex Sacui
Natzke.com by Eric Natzke
Presstube by James Paterson
Gabocorp from Gabo Mendoza
John Mark Sorum by WDDG
2Advanced by Eric Jordan
NRG Design by Peter Van Den Wyngaert
The Hoover Vacuum Site by Fred Flade

… and of course, everything by Hillman Curtis (Rest in Peace)

(Sadly, much of this work is hard to relive due to Flash already being disabled in many browsers. I’ve tried to point to video demos where possible, but you can also try your luck with the Ruffle plug-in.)

From there, a bunch of us new designers set out to learn more about animation, type, scripting, and everything else that put you at the vanguard of the profession in those days. Flash was the first technology that showed us we could be great.

My initial effort was mdavidson.com, a rudimentary personal site that was the precursor to Mike Industries:

From there, I would move on to design Flash sites and features for ESPN, Disney, K2, The New York Rangers, and dozens of other organizations, never matching the quality of the masters listed above, but always breaking new ground in one way or another.

Other fun projects I collaborated on with my friend Danny Mavromatis included a virtual observation deck for the Space Needle, an interactive on-demand SportsCenter, and a Disney movies-on-demand service fully 20 years ahead of Disney+! All in Flash.

Perhaps the thing that gives me the most joy though is something we built and gave away for free: sIFR. What started as our brute-force attempt to use Akzidenz Grotesk for headlines on the front page of ESPN, turned into a more elegant implementation by Shaun Inman, which then turned into a scalable solution by Mark Wubben and me. We poured hundreds of hours into sIFR not to make any money but just to advance the state of typography on the web.

Over the next several years, sIFR was used to display rich type on tens of thousands of web sites. Although it relied on Flash, it was standards-compliant and accessible in its implementation, so it was the preferred choice for rich type until Typekit came along in 2009 and obviated the need for it.

All of this is to say, the role Flash played in helping transition the web from its awkward teenage years to a more mature adulthood is one I will always appreciate. And we haven’t even talked about its role in game development.

When discussing the life and death of Flash, people often point to Steve Jobs’ “Thoughts on Flash” as the moment things turned south for it. Worse yet, the idea that “Steve Jobs killed Flash”. I don’t think either of those things is actually true.

Flash, from the very beginning, was a transitional technology. It was a language that compiled into a binary executable. This made it consistent and performant, but was in conflict with how most of the web works. It was designed for a desktop world which wasn’t compatible with the emerging mobile web. Perhaps most importantly, it was developed by a single company. This allowed it to evolve more quickly for awhile, but goes against the very spirit of the entire internet. Long-term, we never want single companies — no matter who they may be — controlling the very building blocks of the web. The internet is a marketplace of technologies loosely tied together, each living and dying in rhythm with the utility it provides.

Most technology is transitional if your window is long enough. Cassette tapes showed us that taking our music with us was possible. Tapes served their purpose until compact discs and then MP3s came along. Then they took their rightful place in history alongside other evolutionary technologies. Flash showed us where we could go, without ever promising that it would be the long-term solution once we got there.

So here lies Flash. Granddaddy of the rich, interactive internet. Inspiration for tens of thousands of careers in design and gaming. Loved by fans, reviled by enemies, but forever remembered for pushing us further down this windy road of interactive design, lighting the path for generations to come.

RIP Flash. 1996-2020.

If you feel so moved, pour one out for our old friend in the comment section below.

It took an article I almost didn’t publish and tens of thousands of people saying they were creeped out, but Superhuman admitted they were wrong and reduced the danger that their surveillance pixels introduce. Good on Rahul Vohra and team for that.

I will say, however, that I’m a little surprised how quickly some people are rolling over and giving Superhuman credit for fixing a problem that they didn’t actually fix. From tech press articles implying that the company quickly closed all of its privacy issues, to friends sending me nice notes, I don’t think people are paying close enough attention here. This is not “Mission Accomplished” for ethical product design or privacy — at all.

I noticed two people — Walt Mossberg and Josh Constine — who spoke out immediately with the exact thoughts I had in my head.

1/ This is a good *first* step. Better than doing nothing. But it’s not enough. I read the full blog post. It makes no mention of disabling tracking how *often* the recipient opens the email. It’s also full of the rationalization that secret tracking is ok in “business” software. https://t.co/c0PbCRLgdp

— Walt Mossberg (@waltmossberg) July 3, 2019

I appreciate Superhuman’s changes, but the problem is recipients don’t know they’re tracked, and it’s still not going to warn them https://t.co/GPfUYVkBMs

— Josh Constine (@JoshConstine) July 3, 2019

Let’s take a look at how Superhuman explains their changes. Rahul correctly lays out four of the criticisms leveled at Superhuman’s read receipts:

1. “Location data could be used in nefarious ways.”
2. “Read statuses are on by default.”
3. “Recipients of emails cannot opt out.”
4. “Superhuman users cannot disable remote image loading.”

However, he also omits the core criticism: Recipients of Superhuman emails do not know their actions are being tracked or sent back to senders.

Rahul then details the five ways they plan to address those concerns:

1. “We have stopped logging location information for new email, effective immediately.”
2. “We are releasing new app versions today that no longer show location information.”
3. “We are deleting all historical location data from our apps.”
4. “We are keeping the read status feature, but turning it off by default. Users who want it will have to explicitly turn it on.”
5. “We are prioritizing building an option to disable remote image loading.”

The first three apply only to the first criticism about location, but fine. All good moves. Bravo.

The fourth addresses the concern about teaching customers to surveil by default but also establishes that Superhuman is keeping the feature working almost exactly as-is, with the exception of not collecting or displaying actual locations. I’ve spoken with several people about how they interpreted Rahul’s post on this particular detail. Some believed the whole log of timestamped read events was going away and were happy about that. Others read it the way Walt, Josh, and I did: you can still see exactly when and how many times someone has opened your email, complete with multiple timestamps — you just can’t see the location anymore. That, to me, is not sufficient. “A little less creepy” is still creepy.

Also worth noting, “turning receipts off by default” does nothing to educate customers about the undisclosed surveillance they are enabling if they flip that switch. If they’ve used read receipts at all in the past, they will probably assume it works just like Outlook. At the very least, Superhuman should display a message when you flip that switch saying something like “by turning on Read Receipts, you are monitoring your recipients’ actions without their knowledge or permission. Are you sure you want to do this?”

Rahul’s fifth and final fix is also good in that they now realize pixel spying is a threat that they need to protect their own users from. This introduces a moral paradox, however: if the technology you are using on others is something you need to protect your own users from, then why are you using it on others in the first place? These are all questions I’ve asked Rahul publicly in this series of tweets, which I’m still waiting for a response on, four days later:

Thanks for responding, @rahulvohra. I appreciate that you apologized and not in a “sorry *if* you were offended” sort of way. Before I respond, I have a few questions though.

#1: I couldn’t tell from the post if you are still collecting and reporting a log of timestamped (more) https://t.co/oQG5UH8uOE

— Mike Davidson (@mikeindustries) July 3, 2019

Ask yourself, even under this new system, whether you would ever not feel creeped out by someone saying:

“I’ve noticed you’ve opened my email four times, including last night, and even five minutes ago… and you haven’t responded yet.”

What if someone in your family said that? What if your ex said that? What if someone who had threatened you in the past said that? How about someone you didn’t even know? How about your boss?

It would be creepy enough for someone to actually say that to you, but even if they kept their mouth shut, they still know when you are looking at their email, and you don’t even know that they know. All because of these tracking pixels, which Superhuman has decided to continue using.

The message that sender-controlled read receipts send is “I’m watching you, I’ve been watching you, and you didn’t even know it”. Can you imagine ever saying that to someone, in any context, and having it go well?

I cannot. And the reason is that it communicates not only that you don’t trust me, but that I (the recipient) can’t trust you. It also implies that I’m doing something wrong by not emailing you back. As Ray Ozzie says, mess with people’s expectations at your own risk:

As early as Fernando Flores's Coordinator, we learned to view email as the distributed implementation of a very specific set of social contracts.

Once any social tool is ubiquitously-embraced, breach those contracts at your peril & the community's peril.https://t.co/dyV1oyLbFm

— ray ozzie (@rozzie) July 3, 2019

“I’m always watching you” is exactly the expectation that sender-controlled read receipts set. It’s how they work. And it’s the reason people don’t (and likely won’t) disclose that they’re using them.

Above all else, I want to know if people feel safe with this implementation. It doesn’t matter if I feel safe or if Rahul feels safe. Do women feel safe? Do people who have been creeped on over work email feel safe? Do people who have been harassed by salespeople feel safe? These are questions I would love for Rahul and team to investigate. You can probably start with someone like Cindy Southworth (hat tip: @amac) or many of the women, like Tracy Chou, who chimed in on the thread:

my summary of this is essentially: wow, fuck this flagrant violation of my privacy. https://t.co/VL10ySzotS

— Tracy Chou (@triketora) July 2, 2019

To Superhuman’s tremendous credit, they appear to have a pretty diverse team. Out of 30 people, I count 10 women and a variety of ethnicities. In Bay Area tech, that usually takes intentionality. Well done on that. It’s hard to believe, then, that not a single person — employee or customer — ever brought up how creepy the display of timestamps and read statuses are. Maybe someone internally did but the culture was not psychologically safe enough to bring it up and advocate against it. I’m just speculating. I don’t actually know. As Derek Powazek said:

In all those funders, engineers, designers, coders … NO ONE ever expressed surprise that the sender of an email could see a location every time a recipient opened it? When they were implementing it, no one said anything? Ever? I just don’t believe it.

— Derek Powazek (@fraying) July 4, 2019

Turns out, there seems to have been plenty of feedback, at least as far back as October 2018. Here is a Tweet from Elies Campo, formerly of WhatsApp and now working at Telegram (both known for their attention to privacy):

Last October I barely made it past @Superhuman’s onboarding & cancelled shortly after. I was shocked about their tracking & privacy values. I sent them feedback about the tracking features. They replied that they where considering a “stealth mode” . https://t.co/JzwijG5VYd pic.twitter.com/Qtak7t3k3d

— Elies Campo (@elies) July 5, 2019

Read the words from Superhuman “Delight Team” employee Cameron Wiese. He says explicitly says “I agree” and says he thinks Superhuman should turn images off to avoid triggering read receipts and “having your privacy violated”.

Cameron is no longer at the company. I have no reason to believe that is related to this, but it’s proof that Superhuman’s own very small team knew about this a long time ago and decided to do nothing about it.

I began to wonder why, so I started reading up on Rahul. I haven’t followed his career so I wanted to read some things he’d written or said to get a better picture of how he thinks about products. The first thing I came across was this article entitled How Superhuman Built an Engine to Find Product/Market Fit. It’s really well-written and full of a lot of great wisdom from Rahul that can help other entrepreneurs. Stuff I have never thought about for sure. In particular, the bit about zeroing in on the question “how disappointed would you be if you could no longer use this product” is great. It’s kind of an inverse NPS. Really good stuff. There’s one part of the article that may, however, reveal what led to this situation Superhuman now finds itself in: Rahul talks about how he explicitly ignores feedback from people who don’t already love his product. You can read it yourself inside that article or listen to it from his own voice in this interview at the 17:50 mark. Please get the full context from the material provided, but here’s the quote:

“You take the users who most love your product and turn those into an HXC (high-expectation customer), and you use those to narrow the market. And what I mean by that is, deliberately ignore the responses from customers who don’t fit that archetype of people who love your product.”

Bingo.

There is already a huge survivorship bias problem whenever you survey existing customers (which is why people like Elies and me aren’t even represented in these surveys), but doing things the way Rahul describes is like some sort of “devotional bias” on top of the existing survivorship bias.

I will say this: if you were skeptical of Superhuman’s commitment to privacy and safety after reading the last article, you should probably be even more skeptical after these changes. The company’s efforts demonstrate a desire to tamp down liability and damage to their brand, but they do not show an understanding of the core problem: you should not build software that surreptitiously collects data on people in a way that would surprise and frighten them. Superhuman needs to realize that the people their customers send emails to aren’t “externalities”. They are people. And they deserve not to be spied on by software they don’t even know about and never signed up to use. This was an opportunity for Superhuman to internalize what it means to respect privacy, and model behavior for the next generation of companies by doing just that. Instead, they have done little more than the minimum.

I want to quickly detour into a few other issues unearthed by the conversation last week, and then we’ll get back to Superhuman.

First and foremost, it’s important to understand how dissatisfying it is that I happen to be the one who was able to break through on this issue. I am not the internet’s ombudsman or a beacon of morality. For that, I would turn to someone like danah boyd or Anil Dash, who are always a step ahead in thinking about unintended consequences of technology. Second, to my knowledge, I have never been stalked or abused. I am not a victim speaking out. I’m just another white guy of moderately impeachable character who got on my privileged soapbox and said something.

There are several reasons I was able to do this:

• Because of my background and the way I look, I don’t have to worry about getting discredited or blackballed.
• Despite tweeting stuff like this, I have a decent size following on Twitter.
• I got extremely lucky twice in tech, so I’m secure enough financially and career-wise to where I don’t have to give a shit what the technology and venture capital world think of me.
• I haven’t personally wielded the sort of granular tracking technology I am railing against.
• I took the time to write a proper argument in long-form, litigating issues and not people.

Without all five of those things aligning, I think this whole thing wouldn’t have registered a blip. Furthermore, the first four of those things are about who I am and not what I wrote. Think about how frustrating this is for all of the people in the world who have something important they want to bring to light but are only able to do number five. This happens every day, and we miss a lot of it.

Conversely, I will also say that there are a lot of people in the world who have either all or some of the first four taken care of and instead take the easy route by tweeting out some thought-terminating-cliches (hat tip: Kristy Tillman), and then moving on to the next thing they feel like tweeting. If you have an argument to make, put in the work.

Along these lines, it’s been interesting to see who has reacted (and how) to my original article. If you search for who has linked to it on Twitter, you have to scroll through more than 50 posts before you find a single detractor. I didn’t research any further, and I could be biased by how Twitter displays search results, but my gut is that this is at least a 95%/5% situation, if not higher. To anyone who thinks “everyone knows this stuff is going on”, this is a death blow to that theory. “Everyone” in ad tech might know about email surveillance, but the great majority of people in the world do not… and those are the people you are either signing up to be honest with or signing up to deceive. As Upton Sinclair said:

“It is difficult to get a man to understand something, when his salary depends on his not understanding it.”

In this case, the statement refers to getting people who surreptitiously track others to understand that those being tracked do not know they are being tracked nor want to be tracked and that it is a violation of their privacy.

It’s also been interesting to see who has not weighed in. That includes a lot of people on both sides of this issue, including most of Superhuman’s VCs and 120 well-placed angel investors. I have, however, gotten DMs from some very prominent people in the investment community expressing solidarity but unwilling to say anything publicly. I’ve also gotten similar messages from people involved in the creation of Outlook and other tools that have had to wrangle these sorts of issues. I’ve also heard from entrepreneurs who have been specifically told by investors not to engage in discussions like these because it may limit their ability to fundraise in the future.

To those who have spoken out publicly or messaged me privately, thank you!

Conversely, there are also probably people on the other side of the issue who haven’t spoken up because they don’t want to look like jerks. This issue can really make you look like a jerk quite easily, so it’s sometimes easier to just let everyone else tell on themselves instead, like this guy from Founders Fund:

Exactly the caring, benevolent way the venture capital world would love to be represented, I’m sure.

While we’re on the topic of Twitter, I should mention that I’m generally not a fan of having public, free-for-all debates about heated subjects on the platform. I think the format often turns us into the worst versions of ourselves, expelling incomplete thoughts in such staccato bursts that we are often talking past each other and to the larger audience we are trying to impress. Twitter at its best exposes us to wonderful things we’ve never seen before. But Twitter at its worst is just bad performance art. I feel bad that Rahul and team had to absorb the tens of thousands of Tweets directed at them last week. But at the same time, I also feel like they had advance warning several months ago from myself and surely others that what Superhuman is doing is not right.

Being on the inside of this whole clandestine web of intrigue for a few days has made me think twice about this tech ecosystem of ours and what sorts of behaviors we are enabling with it. How many VCs and powerful people hate what Superhuman is doing with people’s privacy but won’t say anything because they aren’t sure if another company in their portfolio does something similarly sketchy with data? How many won’t say anything because they are concerned about their relationship with Andreessen Horowitz?

This episode has also made me take stock of whether there’s anything in my own life which is collecting data it doesn’t need to collect. Someone on Twitter brought up the fact that I use Mailchimp to send out newsletters. That’s a good place to start. A few years ago, I enlisted Mailchimp to automate newsletter creation for me. I wanted to give people an easy way receive an email every time I wrote a new post. That’s about two or three times a year. Mailchimp makes this so easy that since installing it, I’ve never once logged into the service. I didn’t even know what, if any, data they were collecting aside from the number of subscribers I had. Turns out, they can collect a lot more data than I am comfortable with. Thankfully you can disable substantially all of it, which I have done. It bothers me that these services are choosing to collect all of this data for people who don’t even need it or want it. It turns people into “unwitting data collectors”.

It reminds me of the early days of Android when developers immediately asked for every single permission they could get from you. Now the conventional wisdom is to only ask for what you need, when you need it. It makes things slightly better that people are at least opting into these newsletters, but to use my same test from the original post, there is no way they know how much data is being collected on them. That said, I’m generally not moved by straw man arguments that attempt to paint bulk newsletter analytics with the same brush as email surveillance. News organizations are well within their rights to employ the former while criticizing the latter.

The second thing someone asked me on Twitter is whether the company I work for uses tracking pixels anywhere. I’m not in sales or I.T. so I have to look into how different people use analytics over here, but I imagine there are a variety of ways. I’m going to be proposing an explicit policy against the sort of thing described in this article this week and I don’t expect that will be controversial.

Ok, so back to Superhuman.

We are left now in a better state than we were last week. The threat level has decreased. But I am still left wondering, why is Superhuman taking this feature — which clearly creeps people out — and doing barely more than the minimum to make it less creepy? Rahul said exactly why in his post:

“If one of us creates something new, and that innovation becomes popular, then market dynamics will pull us all in that direction. This is how we ended up with location tracking inside of Superhuman, Mixmax, Yesware, Streak, and many others.”

Rahul is not wrong. But that is not how the greatest innovators think. The reason Slack is now an $18b company whose software is loved by millions of customers is that Stewart Butterfield created a new workplace communications tool based on how he thinks workplace communications should work. Stewart and team looked at what tools people were currently using, invented a new service full of things that seemed good, and left out everything that seemed bad. Are there any deceptive, creepy, or harmful features that exist in Slack because they already exist in other products? Not that I can think of. Someone please tell me if I’m wrong. Heck, Slack doesn’t even have read receipts! And it would be easy to design them ethically within Slack if they wanted to.

Either Rahul thinks email apps should be able to spy on recipients’ behavior without their knowledge or permission, or he doesn’t think they should — but he’s doing it anyway because other bad actors do. Neither of these represents the standard we should hold our entrepreneurs to… especially those we point to as models for great design and great leadership.

The other thing that’s been bugging me is that Superhuman’s other co-founder, Vivek Sodera, has openly compared Superhuman to Apple. See this thread:

Notwithstanding the fact that I believe Google could extinguish Superhuman’s entire existence with the flip of an API access switch, it strains credulity to see how the decision to surreptitiously collect behavioral data on unsuspecting users is Apple-like at all. Vivek is talking here about whether they will license their data out in the future, but still… if you are going to say you are like Apple, then you should at least try and act like Apple. Do I think Apple would ever insert invisible tracking pixels into emails so senders could monitor the actions of recipients without their knowledge or permission? Not in a million years. Do you?

To test my assumptions about how Apple, and in particular Steve Jobs, might approach a problem like this, I asked the only person I know who has worked directly for Jobs, across several decades: Mike Slade. Mike is the founder of ESPN.com, the original product manager of Excel, and worked directly alongside Jobs at both NeXT and Apple. Here’s what Mike said to me:

“Steve was the most consumer-first person I’ve ever worked with. If he didn’t like what the consumer was going to experience, he changed it. This functionality would’ve definitely creeped him out and he would’ve never implemented something as creepy as this.” — Mike Slade

Sadly, we’ll never know for sure, but this makes sense. Jobs used to talk a lot about the importance of taste in product development. That is exactly the concept that is missing here.

You just raised $36m so you could build a product for the long term. You think tracking pixels in emails are even going to be around in a few years? Differentiate yourselves from your competitors by giving a shit about privacy. Think Different.

To show you how this might work, I’ve taken the liberty of redesigning your sales pitch for you. Here is how you currently describe Read Receipts on your front page:

Now here’s how it would look if you decided to take a stand on privacy and protect people from both tracking and being tracked:

I feel like I am doing an unusual amount of free work here. THIS is the sort of morality I want to see in enterprise software. It’s funny, one of the things people like to talk about is how the iPhone kicked off “the consumerization of enterprise software”. Meaning, because the iPhone set the bar so high for how consumers experience digital products, all enterprise software eventually rose to meet this bar. If we can now expect our enterprise apps to look and feel as nice as our consumer apps, why can’t we also expect them to behave as nice?

To harken back again to the tao of Steve: “Design is how it works.”

Alright, that’s all I have on the subject of Superhuman specifically. Take my advice or leave it. It’s your company.

A couple of more things before I go. One of the valid criticisms of my article from last week is that I didn’t call out any of the various other companies that enable email spying. This is fair. I frankly didn’t know about most of them, since I don’t use email tracking myself. For instance, I had heard of Mixmax but thought it was just an extension which let people book time on your calendar (it does this too). I am more than happy to name the names of every company who does this. From Rahul’s post, that looks like Mixmax, Yesware, Streak, Mailtrack, and HubSpot (whose founder is a Superhuman investor). There are probably others too. To all of you: what you are enabling is bad and you should feel bad about enabling it. None of you pitch yourself as a well-designed email client so you avoided attention in my first post, but isn’t there a way for you to operate your business without enabling your customers to spy on their customers? Mixmax, you have that useful calendar thing. Is that not enough? HubSpot, Streak, and Yesware, you offer a bunch of services that are unrelated to this. Mailtrack, welp… this seems like pretty much all you do from what I can tell.

Is this stuff even useful in a material way? If you send someone an email and they don’t respond, you can either let it go or reach out again. Does knowing whether someone read it really change what you’re going to do? On top of that, aren’t we already in a world where you’re getting false positives and false negatives from that data? If I have images off (which, for the love of god, everyone should at this point), you’re seeing that I haven’t read email, when maybe I have. If Gmail or something else is proxying my email images, you might be seeing that I have read email, when maybe I haven’t. The data may be “usually right” because most normals don’t pay attention to this stuff, but how can you be sure? You can’t. This seems like a case of very low value data being collected and distributed in a potentially very harmful way.

To wrap things up, I want to address the final issue brought up from the first article: what are the big three mail platforms (Apple, Microsoft, and Google) doing to protect us? The answer seems to be “some, but not enough”. All three allow users to disable images, but none make that a default. I could make the argument that email should be text-only by default, but I don’t think that’s realistic given the sorts of emails people subscribe to these days (real estate listings, deals of the day, etc). Accepting this, it seems like email providers should use the same sort of filtering they use to keep us safe from malware. If an email with a Mailtrack pixel comes in, for instance, strip it. You’d have to maintain a growing list of these things as they mutate across IP addresses, but it would send a strong signal to the industry that this sort of stuff is on the outs.

Google and possibly also Microsoft could also stop this stuff from the other side: disallow extensions which provide this functionality. You don’t have to kill most of these companies’ entire businesses. Just specifically disallow this behavior.

Because neither of these solutions stops companies from using large, visible, legit images in their emails to provide the same tracking abilities, the big three should also proxy and cache remote images in email whenever they can. I believe Google already does some sort of this, but it’s unclear to me exactly how obfuscatory it is. I don’t believe either Microsoft or Apple does any of this yet.

This is almost certainly a case of “if you think the solution is easy, you don’t understand the problem”, so I recognize there’s a lot of implementation complexity I’m missing here. I guess I would just like to see all of these companies do everything they can to protect the majority of the world, who — unless they were paying attention last week — still doesn’t know they are naked with their curtains open for all people using spyware pixels to see.

Perhaps if the big platforms aren’t able to sufficiently protect people, the last resort is the law. I’m not a COPPA or GDPR expert, but it seems crazy that collecting information via a website about someone under 13 without parental consent is illegal, but providing software that can automatically track that child’s movements when they open an email is not.

This whole thing may already be advancing through the legal system, as just a few days ago, the British Information Commissioner’s Office issued guidelines requiring consent and transparency for email tracking pixels. If ethics can’t keep companies from doing these sorts of things, maybe fines can.

It seems to me we’ve still got a lot of work to do here to keep people safe. Until that work is done, the best way to stay safe is to follow the same two pieces of advice from my previous article:

1. Don’t use Superhuman yourself. They have not given a date for when they will protect you from other people’s tracking pixels and they have not shown a proper appreciation for privacy. Remember, when Superhuman says “you can turn it off”, that only means you can stop sending your own tracking pixels out.
2. Turn off remote image loading in whatever email client you use. You may also want to consider using an always-on VPN to keep your location from ever being revealed.

Thank you for reading this. Stay safe out there.

It is disappointing then that one of the most hyped new email clients, Superhuman, has decided to embed hidden tracking pixels inside of the emails its customers send out. Superhuman calls this feature “Read Receipts” and turns it on by default for its customers, without the consent of its recipients. You’ve heard the term “Read Receipts” before, so you have most likely been conditioned to believe it’s a simple “Read/Unread” status that people can opt out of. With Superhuman, it is not. If I send you an email using Superhuman (no matter what email client you use), and you open it 9 times, this is what I see:

That’s right. A running log of every single time you have opened my email, including your location when you opened it. Before we continue, ask yourself if you expect this information to be collected on you and relayed back to your parent, your child, your spouse, your co-worker, a salesperson, an ex, a random stranger, or a stalker every time you read an email. Although some one-to-many email blasting software has used similar technologies to track open rates, the answer is no; most people don’t expect this. People reasonably expect that when — and especially where — they read their email is their own business.

When I initially tweeted about this last week, the tweet was faved by a wide variety of people, including current and former employees and CEOs of companies ranging from Facebook, to Apple, to Twitter:

It was also met critically by several Superhuman users, as well as some Superhuman investors (who never disclosed that they were investors, even in past, private conversations with me). I want to talk about this issue because I think it’s instructive to how we build products and companies with a sense of ethics and responsibility. I think what Superhuman is doing here demonstrates a lack of regard for both.

First, a few caveats:

1. I was invited into the Superhuman service several months ago. I began their onboarding process, was excited to try using it as my primary email client, and bailed out the moment I found out about this spyware functionality.
2. Nothing in this post evaluates other things about the Superhuman service. I’m not here to tell you it isn’t fast, isn’t good looking, or doesn’t save you time. I suspect it is all of those things, in fact. So if your response to this post is “BUT I LIKE IT!”, I believe you that there are things to like about it. That is not the subject of this post.
3. I know people whom I consider to be ethical people who use workplace software that embeds tracking pixels in emails.

Second, I want to talk about why this particular issue is so important. Not why privacy is important; we are all already learning that the hard way. Rather, why making ethical decisions at the earliest stages of your company is important.

When a company first forms, there are no norms or principles guiding how its people should make decisions. It’s basically just what’s in the founders’ heads. With each decision a company makes, its “decision genome” is established and subsequently hardened. You’ve decided in your first month that you’re only going to hire engineers from Top 10 engineering schools? That’s now part of your genome and will determine the composition of your company. You’ve decided to forgo extra profits by keeping your prices low for consumers? That’s now part of your genome. You’ve decided to employ a single dark pattern to trick users into adding more things to their shopping cart? Part of your genome.

The reason this matters is that what may seem like small decisions early on become the basis for many more decisions down the road. These decisions affect your ethical trajectory as a company. Let’s use the dark pattern example. Maybe the shopping cart thing was pretty minor and you were able to rationalize it internally in a variety of ways, including the fact that the extra item in the user’s cart was inexpensive and provided value (like a product warranty, for instance). Down the road, when employees want to employ more dark patterns, here is how the conversation would go:

Greg: “Hey, we aren’t getting enough people to opt-into our mailing list when they sign up. Can we try maybe unchecking that box by default but using language such that leaving it unchecked opts people in?”

Desi: “Wouldn’t we be intentionally deceiving users if we did that?”

Greg: “Uhhhh, we already add things to your shopping cart that you don’t even ask for!”

Desi: “True. This seems like less of a big deal than that. I guess I’m OK with it.”

If you’ve never worked at a tech company before, this is how things go. When faced with making a product decision that is even mildly uncomfortable, employees often first look towards expressed company principles like “Always put the customer first”, but the next thing they look for is precedent. What other decisions have we made that look like this one? Designers do this. Engineers do this. Product managers do this. Executives do this. It’s an easy way to inform your current decision, and it’s also an easy way to cover your ass. Imagine the above decision was made by a product manager, and later on the company was called out publicly on it. The CEO or Head of Product marches over to the product manager and says “what were you thinking here?!?” The product manager needs only to point to the shopping cart behavior in order to let him or herself off the hook.

The point here is that companies decide early on what sort of companies they will end up being. The company they may want to be is often written in things like “core values” that are displayed in lunch rooms and employee handbooks, but the company they will be is a product of the actual decisions they make — especially the tough decisions.

So back to Superhuman. Here we have a company that professes to create a better email experience mainly through better design and engineering. So far so good! Those who know me know that I would be among the first people to sign up for something like this and also among the most vocal to evangelize it. Heck, I love a certain showerhead so much that I:

1. Have bought one for every shower in every place I’ve lived for the past several years.
2. Bought one for every member of my leadership team at Twitter for Christmas.
3. Used to keep a brand new spare one in my trunk to give away to friends every time the subject came up.
4. Turned the damn showerhead emoji on Twitter into a Flipside. Actually, my team did this, but as an homage to my love for it.

In other words, when I see great design, I proactively try to spread it as far and wide as possible.

What I see in Superhuman though is a company that has mistaken taking advantage of people for good design. They’ve identified a feature that provides value to some of their customers (i.e. seeing if someone has opened your email yet) and they’ve trampled the privacy of every single person they send email to in order to achieve that. Superhuman never asks the person on the other end if they are OK with sending a read receipt (complete with timestamp and geolocation). Superhuman never offers a way to opt out. Just as troublingly, Superhuman teaches its user to surveil by default. I imagine many users sign up for this, see the feature, and say to themselves “Cool! Read receipts! I guess that’s one of the things my $30 a month buys me.”

When products are introduced into the market with behaviors like this, customers are trained to think they are not just legal but also ethical. They don’t always take the next step and ask themselves “wait, should I be doing this?” It’s kind of like if you walked by someone’s window at night and saw them naked. You could do one of two things: a) look away and get out of there, realizing you saw something that person wouldn’t want you to see, or b) keep staring, because if they really didn’t want anyone to see them, they should have closed their blinds. It’s two ways of looking at the world, and Superhuman is not just allowing for option B but actively causing it to happen. It’s almost as if Superhuman is aiming a motion-sensitive camera outside people’s windows and sending alerts when there is motion. It’s automated and designed to capture info when your family, your friend, your co-worker, or your victim is not aware. You may think “victim” is too harsh of a word to use here, but remember, we aren’t talking about you. We are talking about anyone who might use Superhuman.

Even though most of the feedback I’ve gotten about raising this issue has been supportive, here is a collection of replies I’ve gotten on Twitter, so we can address them all in one place:

1. “Email clients have done this for years. Even Apple does this with iMessage.” — Multiple People. This argument is naive at best and disingenuous at worst. Superhuman’s competitors are Apple Mail, Gmail, and Outlook. Exactly zero of those companies insert a tracking pixel into their emails. Furthermore, both Outlook and iMessage use Read Receipts that are turned off by default and controlled completely by the receiving user. In other words, when you buy a new iPhone or start using Outlook, no one requesting an Outlook or iMessage read receipt can receive one without your explicit permission. Furthermore, even if you do turn those on, it’s a simple one-time receipt… not a log of times and geolocations every time the recipient views the message. Both Microsoft and Apple — as well as other messaging platforms like Twitter, LinkedIn, and Signal — have designed their read receipts in an ethical way. Superhuman has not. Also ask yourself what the backlash would be if a company somehow retrofit these spying capabilities on top of iMessage. What if every single time you viewed someone’s text message, your phone sent a timestamp and location to back to the sender, creating a map of your movements? There’s a reason Apple doesn’t allow this.
2. “Other tools like MailChimp, PersistIQ, SendGrid, and MailTrack do this.” — Multiple People. Superhuman is an e-mail client, much like Gmail, Outlook, and Apple Mail. It is not mail automation software. Mass mailing companies, for the most part, use that technology to track open rates and also to *stop* sending out emails to people who haven’t opened them in months. There are indeed some sales-enablement companies that use this technology to track individual opens, and I find that just as creepy. The main point here is: just because technology is being used unethically by others does not mean you should use it unethically yourself. Harmful pesticides have also been around for years. That doesn’t mean you should use them yourself.

3. 

   Where to start with this one from Gary Sheynkman. You, the sender, do not get to decide how I, the receiver, respond to you. Not returning your email right away is not passive-aggressive. It’s often just being busy or prioritizing. As pointed out by Erica, being “left on read” can send unintended hurtful messages. Furthermore, in the workplace, this can be used as a tool to monitor or coerce around-the-clock work.

4. 

   This is from Nick Abouzeid, a Superhuman investor (who did not disclose that, but I got it by going to his website) from the aptly named “Shrug Capital”. This comment gets to the crux of the ethics question we are talking about here. When you are making software, you can either say “lets exploit everything in the world that can act to our benefit” or you can say “lets build something that’s great for the world”. This person looks at all people who use email as potential people to exploit. How many people use email? 2 billion or so? How many of those have images turned off? Probably a tiny percentage. And how many are expecting that every time they open an email from a friend, their friend gets notified with their geolocation? I would guess almost zero. So what this person is essentially saying is that since most people leave their curtains open at night, it’s ethical for the company he funds to film what goes on inside. Furthermore, Superhuman doesn’t even let its own customers turn images off. So merely by using Superhuman, you are vulnerable to the exact same spying that Superhuman enables you to do to others. He is right about one thing though: because of spyware-foisting companies like Superhuman, you should unfortunately turn off all image loading in your email client.

When you start to think about all of the ways Superhuman can be used to violate privacy, you really wonder why The New York Times spent 1,200 words on a tongue-bath that doesn’t even talk meaningfully about privacy issues at all. We don’t need journalism to tell us where venture capitalists are putting other people’s money. We need it to examine the ramifications of the technology we are pushing into the world and in what ways it might shift the Overton Window for Ethics in either helpful or hurtful ways.

There are some bad people out there, so what are some bad things that people can do with technology like this? Here are just a few:

• An ex-boyfriend is a Superhuman user who pens a desperate email. Subject: “I’ve been thinking about us”. He sends it to his former partner. She reads it when she gets to work in Downtown Los Angeles at 9am. She reads it again before dinner with friends in Pasadena at 7pm. She reads it again at home in Santa Monica at 1am. Over the weekend, she takes a trip to New York and reads it again. Twice. She decides not to answer the email, because her ex has stalked her in the past and she doesn’t want to communicate any further. But because of the tracking pixel, her email is always communicating, and it’s sharing info she does not want to send and doesn’t even know she is sending. She didn’t reply, but her ex still knows she read his email five times, including most likely in her bed. And he knows she took a trip to New York.
• A pedophile uses Superhuman to send your child an email. Subject: “Ten Tips to Get Great at Minecraft”. Your child keeps the email in their inbox and refers back to it often over the course of a year. Sometimes when they are at home in Vermont. Sometimes when they are at school in New Hampshire. Sometimes when they are with their grandparents in Massachusetts. Every time your child opens the email, that person knows generally where they are (or specifically, if they have other info to triangulate against).
• Superhuman decides they can make more money by supplementing their subscription fees with data licensing agreements. Maybe they decide to leave out data from paying Superhuman customers but they include location history from every single person you’ve ever emailed, because they have no contract with those people. Location maps with timestamps, other insights about things like working hours and locations, device types, and whatever else they collect. That data is then used to target those people in a variety of ways. If Superhuman is truly willing to commit to never license any data to anyone for any reason, they should be able to clearly say so right now. But they probably won’t, because they want to keep their options open.

I understand wanting to cover a new product, but an outfit as respected as the Times needs to go deeper on this stuff. Heck, I’m already at 4000 words — on a single subject — and I just wrote this on a whim over the weekend.

Even though I wish companies didn’t make the sorts of product decisions Superhuman has made, I’m glad they are at least showing their cards early (and appear to stand by them) so I can avoid their service. Not just on principle but because I have no reason to trust them with any of my data. Remember that they require full access to your Gmail in order to do their thing. Fast forward a year or two and I can see them licensing location data either from their own customers or their non-customers to a third-party for any number of distasteful purposes. They say they have a privacy policy that forbids this, but I don’t read their policy that way at all. It allows and even specifies all sorts of things they can do. Here’s an excerpt (emphasis mine):

HOW WE USE TRACKING TECHNOLOGY TO COLLECT INFORMATION

We automatically collect usage information when you visit our Website or use the Service through the use of tracking technologies, including tracking pixels and similar technology (collectively, “Tracking Technologies”). We may use the data collected through Tracking Technologies to: (a) remember information so that you will not have to re-enter it the next time you visit the Website or use the Service; (b) provide and monitor the effectiveness of our Service; (c) provide functionality of the Service including read receipts; (d) monitor and collect analytics data using third-party tools like Google Analytics in order to help measure traffic and usage trends for the Service; (e) diagnose or fix technology problems; and (f) otherwise to plan for and enhance our Service.

The form of this paragraph is very familiar to lawyers. Specify some stuff that sounds mundane and then leave yourself all sorts of escape routes. Item F is essentially a universal license to do whatever they want (i.e. “We’ve ‘enhanced’ our service by using your location data in a new way!”). Now, I’m not saying this privacy policy is out of the ordinary at all. I’m just saying there is nothing about Superhuman’s Terms of Service that prevent them from making further decisions that violate your privacy in the future. Not to mention, companies can change their policy at any time. When you use a product, you need to trust the people who are building it — not the documents their attorneys create. And finally, once again, because of this spyware pixel, most of the people they are collecting information on aren’t even Superhuman customers and never even signed up for this policy.

So what would I do if I were Rahul Vohra, Superhuman’s CEO?

The first thing I’d do is apologize and remove this functionality for everyone. You don’t need to take out a front page ad in the Times. Just own the mistake and disable the feature unless and until you can design it in an ethical way. Don’t keep it up for a year while you work on it. Take it down. This would show responsibility and regard for doing the right thing. A sign of an honorable company is when it is willing to learn, take responsibility, and improve.

Next I would recalibrate how important it is to even offer Read Receipts. Superhuman seems to be doing just fine in terms of customer satisfaction. I know some people like this functionality but does the success of Superhuman’s business depend on it? I would guess not.

Third, if Superhuman really cared about protecting the privacy of its users, they would actually provide the exact opposite of this feature. Protect all Superhuman users from emails loaded with surveillance pixels, do it by default, and never embed them in their own outbound emails. There are a few browser extensions that do this decently while keeping most benign images turned on, but this would be a great feature to have baked into an email client. Use this moment as a turning point to honor all of the other work going on at the company and turn this negative into a positive.

Finally, if I still didn’t agree that tracking the geolocations and reading behavior of unwitting people was deceptive, I’d wear it on my sleeve:

A good ethics test for investors, employees & users of Superhuman. If you use SH’s default spyware behavior, you should be willing to change your signature to this:

– Sent via Superhuman. Every time you open this email, the time and your location are sent to me.

Who’s first?

— Mike Davidson (@mikeindustries) June 27, 2019

A lot of Superhuman’s customers — and I assume most of its employees and execs — use the “Sent with Superhuman” signature already. If you are so sure that automatically receiving a recipient’s geolocation every time they read one of your emails is OK, you should be OK with telling them that when you send them an email. In fact, since it’s a feature you are proud of, you should be more than happy to market it so clearly, right? (In case you’re wondering, exactly zero people stepped up to do this. Not Rahul Vohra, no employees, no investors, and no customers.)

Before I close, I want to talk about blame.

I don’t automatically blame Superhuman designers for this, because I don’t know if they fought for it or against it. Likewise, I don’t blame Superhuman engineers, product managers, or anyone else, for the same reason. For all I know, this was something the team pushed back strenuously on and lost. If so, thanks for fighting the good fight, and you should have your equity grant doubled for trying to do the right thing. If not, this whole article also applies to you. The only person I know for sure shares some or all of the responsibility for this is Superhuman’s CEO, Rahul Vohra. He is the only employee I’ve seen actively defend it (although perhaps others have), and it’s safe to say he was intimately involved in its development.

This is also important: I do not know Rahul. I don’t know if he’s nice, mean, a good person, a bad person, likes his steak medium rare or well-done with ketchup, or anything else about him. I make no value judgements about him as a human, and as with most other humans, I try to assume the best about him. This article is about a very specific decision of his that I find to be dangerous and wrong. Whenever I see something like this, I always give the benefit of the doubt and assume the person simply doesn’t realize the downside consequences of their decision. I brought this up several months ago and the company did nothing to address it. I brought it up again last week and still nothing. In light of that, and in light of some of the responses from investors defending Superhuman’s surveillance behavior, I felt justified writing a proper piece about it.

Finally, if you are a user wanting to protect yourself against automatically having your email behavior and geolocation sent to people who use Superhuman and other surveillance tools, you need to do both of these things:

1. Don’t use Superhuman yourself. As mentioned earlier, Superhuman leaves you unprotected from spying because they don’t allow you to block these spyware pixels. It also turns you into an unwitting information collection machine, aggregating info on every single person you send email to. Possibly including your parents, children, partners, and friends.
2. Turn off remote image loading in whatever email client you use. Almost every client these days allows you to do this, with the strange exception of Gmail for iOS. If you are a Gmail user, I recommend switching to something like Outlook on your iPhone. It’s free and allows you to use your existing Gmail account. No migration necessary. Apple Mail is fine too.

So to sum up — whether you are an email provider or an email user — don’t surveil, and don’t allow yourself to be surveilled. I hope Superhuman does the right thing and decides to help stop this problem instead of trying to normalize it.

A wonderful state-of-the-union for the design industry as we move out of the age of attention hijacking towards a mindset that puts users’ health and happiness first. Great writing from Fabricio Teixeira and Caio Braga.

The comfort of our design jobs, especially in Silicon Valley, has, in many ways, limited our power to advocate for the right thing. We are comfortable in our expensive chairs, busy pleasing our internal stakeholders and pretending we can keep our responsibilities as citizens out of our work, and the impact of our work out of our personal lives. For a long time, we even ignored harassment issues in our offices.

Some people survive on talent, vision, persistence, or a host of other superpowers, but to make up for what I lack, I’ve always been a tool nerd.

Whether it was messing around with Print Shop on a Commodore 64 when I was 10, teaching myself Photoshop 3.0 in my teens, or learning HTML via “HoT MetaL” while most designers were only doing print work, aggressively learning the tools of the near future has been one of the only consistencies of my career.

For that, and other reasons I’ll explain below, I’m incredibly excited to announce that after a refreshing two-year sabbatical from work, I’m joining one of my favorite companies — InVision — to head up Partnerships & Community!

We are entering a golden age of product design tools right now, and I’ve seen first-hand what introducing great prototyping and collaboration software like InVision can do within a company. It breaks down barriers between different groups and gets everyone thinking about user experience. It demystifies what product design actually is. It replaces 50-slide presentations and exhaustive spec documents with quick-to-create working demos that everyone can hold in their hand. More showing, less telling.

As my friend at Google, Darren Delaye once told me, there’s no better way to communicate a product idea than pulling out your phone and saying “Hey, wanna see something cool?”

When Clark and I first started talking about working together a little while ago — and about this role in particular — I’ll admit I had some question marks. In particular:

1. InVision is a fully distributed company with no offices.

A lot of people probably view this as a positive, but the thing I miss most about my last job at Twitter is being around my team and all of the other great people there. I’m an extrovert, and even if I have to brave open-office plans and commute times, being around teammates has always been important to me. Video calls, in fact, have always felt like a burden because face-to-face is the default mode of communication at most companies.

I had a really tough time evaluating what working remotely would be like, but in talking to Clark, Stephen, Aarron, David, Hilary and others at the company, everyone says exactly the same thing: it sounds like it’s going to be awkward, but after a couple of months, you never want to work any other way again. It’s great to be able to spend the first few hours of your day working from your couch, then go on a run whenever the weather clears up, and then spend the rest of the day working from your patio or your local coffee shop… all without ever getting stuck in traffic! InVision even gives you a $100 coffee & tea credit every month to encourage you to explore new surroundings.

It seems that when everyone is remote, the working dynamic changes. You aren’t sitting in a room with 10 people and then figuring out a way to dial in your poor London teammate with a choppy internet connection. You aren’t keying off of all of the social cues in the conference room and losing nuance from your one teammate in India. When everyone’s in the same boat — even if it’s a metaphorical boat held together by thousands of miles of fiber — something apparently changes.

I am now looking forward to this social experiment in telepresence, even if it means I have to remove the Zuck Tape covering my laptop’s camera. I think InVision is correct that this will be the new normal for a lot of companies over the next decade, and if you are a designer who doesn’t want to (or can’t) live in San Francisco, this seems like good news for you too. Either way, I will report back in a few weeks to let you know how it’s going.

2. I’m used to working on consumer products, as opposed to products that help people make other products.

Whether at ESPN, Newsvine, NBC, or Twitter, most of my career has been spent building things that consumers use. In evaluating each of those opportunities, I’ve tried to ask myself “what kind of impact is this going to have on the world?” This opportunity is a bit different because InVision doesn’t really touch consumers directly, but rather, it touches the designers, engineers, PMs, researchers, writers and others responsible for making the products that touch consumers. In that sense, it’s hard to trace the social impact of your work because you aren’t even aware of all the products that are being built with it. Having used InVision regularly myself, however, I am confident of three things: it tends to raise the profile of design work inside of companies, it facilitates a more inclusive product development process, and it ultimately helps create better user experiences. All great things.

You may say to yourself “but I like another tool better!” That’s totally fine. Any of the options available today are light years better than what we did ten years ago when we emailed JPEGs back and forth. I happen to think InVision is the best set of tools out there for entire organizations to use (with more products and capabilities on the way), but a world in which designers have several great options to choose from is a world I want to live in. Competition breeds excellence, and there is a lot more work for all of us to do.

One last note on tools and their impact: even though it was over ten years ago now and it started as a fun little hack, sIFR is still one of the projects I’ve enjoyed working on the most. None of us ever made a penny on it, or even tried to, but watching that little tool help people beautify typography across the web was really fulfilling. Seeing the need for it slowly disappear with the advent of TypeKit and proper web fonts was just as satisfying. It’s great to work on things that inch the world forward and make other people’s work better, and I look forward to doing that again in whatever ways I can at InVision.

3. The output of my work will, for the most part, not be directly within the product.

The role I’m taking on at InVision is not within Product, Engineering, or Design, so I won’t actually be working on the products themselves (but you can be sure I’ll be in people’s grills with ideas and feedback! ).

Instead, I’ll be doing the following:

• Working with hundreds of companies and design teams around the world listening to how they currently develop products, how they want to develop products, and how we might be able to help.
• Keeping an eye on interesting new products and teams in the design software space and working on acquisitions when it makes sense.
• Designing and executing product integrations that bring the functionality of InVision to other platforms, and vice versa. InVision is already integrated with the largest platforms that drive digital product success, including those from Slack, Atlassian, Dropbox, and Microsoft, but there are still a lot more nodes in the digital product ecosystem to connect. The goal is to make the workflow of product design as seamless as possible, no matter what assortment of platforms a team is using.
• Emceeing InVision’s Design Leadership Forum, which hosts private events for design leaders from around the world. Its goal is to advance the practice of design leadership by creating a community where leaders can learn from one another.
• Further building out InVision’s programs within the traditional and continuing education spaces.
• Working on inclusive ways to bring the design community together, both online and off.

It’s a different sort of role for me, but at the same time, I’ve actually been doing a decent amount of exactly this sort of thing during my time off.

In some ways, this seems like the perfect opportunity at the perfect company right now, and yet, in other ways, I’ve never done anything like it! One thing that feels palpable already, however, is that this is a company driven by producing great experiences — for its customers, its partners, and its employees — and for that reason, I already feel at home (also, I *am* at home).

I’m not sure what my hiring plan is just yet, but if I’ve ever worked with you in the past or if you see anything you like in the 71 positions posted here, feel free to reach out to me directly! If you’re interested in helping power the next decade of digital product development from the comfort of wherever you choose to work, now is a great time to be joining InVision.

You may hear things like “Design is very important to us” or “Design has a seat at the table” or even “We’re a design-driven organization”. As a designer evaluating job opportunities, should you take statements like this at face value or might you be able to get a clearer picture by asking questions, reviewing a company’s products, and other investigative means?

Throughout my 20 years in design, I’ve worked in everything from design-hostile to design-driven environments, and I can tell you that succeeding (and being happy) within each requires a different mindset going in.

Before discussing strategies, let’s describe several types of organizations along the design spectrum so as to identify what they look like. Importantly, when I say things like “design-driven”, I don’t necessarily mean Design-ER driven. In other words, it doesn’t mean that designers call all the shots. It just means that great care goes into every detail of the user experience of a company’s products. That level of detail could come from engineers, researchers, execs, or any number of other people (and often does!), but it also tends to correlate with designers playing key roles at the company.

The Spectrum of Design Drivenness

Design Hostile

CRMster is a 20-person startup that develops Customer Relationship Management software. There are 10 salespeople, 8 engineers, 1 CEO, and one contract designer. Most product decisions are made by the CEO and one salesperson. The rest of the team just builds whatever they are told to build. If a designer or engineer brings up concerns about a product feature, they are told to just stick to the program and produce what the CEO asked for as quickly as possible.

Design Ignorant

GitBusy is a 5-person startup building a new way to sync files across computers. All five team members are back-end engineers and they have spent their first year building out core functionality. The product is a bear to use, but it’s starting to work reliably. They keep making minor usability improvements but they’ve never thought to hire someone full-time who specializes in user experience.

Design Agnostic

MegaloBank is a Fortune 100 financial institution that employs over 10,000 people. They employ plenty of designers and design-minded people around the company, but mainly as support for other teams. If you asked the CEO of the company what his or her designers did, the response would be something like “I think they make our logos and business cards.” Important product design work does get done at Design Agnostic companies, but the people who do it just aren’t looked at as core talent the same way as sales, engineering, or marketing might be. Their compensation unfortunately reflects this.

Design Interested

CellYou is a 20-year old wireless carrier, employing over 10,000 people. They are feeling the heat from their competitors and have just embarked on a high-profile effort to redesign their product line so it is much easier to use. They don’t have all the right people in place yet, they aren’t walking the walk in terms of a user-centered product development process, and they still pay their designers a lot less than PMs and engineers, but they are starting to talk about things in the right way and starting to recognize the value of design. It may take a few years, but they are moving in the right direction. A lot of companies who say they are Design Driven are actually in this category instead.

Design Driven

HyperBowl is a 500 person company that makes a versatile kitchen appliance which can cook hundreds of different foods. It started as one chef and one engineer, but even from the very beginning, there was a relentless focus on building product prototypes and iterating them rigorously based on how they performed with users. New products are developed only when they can perform important jobs for users. At HyperBowl, there are 10 full-time designers but everyone in the company considers user experience to be one of their most important job functions. Many decisions are made with data and research, but there’s room for subjectivity, taste, and long-term vision as well.

Those are the five archetypes you’ll generally run across in the market today.

The Challenges of Each

As a designer, you may immediately think to yourself “I only want to work for a Design Driven organization!” If your goal is to join an environment that immediately affords you the ability to practice great design, then this is a natural choice. Take Apple, for instance. Whether you are fresh out of school or a 20-year design veteran, as a new employee at Apple, you won’t need to spend any of your time convincing anyone of the value of design. You will be paid roughly on par with engineers, expected to help lead the product development process, and likely do some of the best work of your career. I have a bunch of friends who have worked at Apple, and one of them summed it up best: it’s like career rocket fuel. You may find the work/life balance unsustainable in the long term, but your time there will be unencumbered by any fights about the value of design.

A lot of companies (even some of the largest and most successful in the world) will tell you they are Design Driven, but they are actually one level away, in the Design Interested category. You should always maintain a healthy suspicion about this, in fact. Like our cellular carrier above, they are in the midst of a positive transformation, but they just aren’t there yet. What this means for you as a designer (or researcher, or front-end engineer) is that not that you won’t be respected to listened to, but rather that part of your job will be to move the company from its old way of doing things to a new way of doing things. This involves a lot more than just your technical skills. It involves the patience to work in suboptimal conditions and the willingness to help lead the process of becoming more Design Driven. Thankfully since the company is already Design Interested, the “what”, “when”, and “why” have already been taken care of for you, but the “who” and “how” are still open questions. You are part of the who that will help determine the how!

A large number of companies in the world fit into the next category: Design Agnostic. These companies have found a way to turn a profit without paying as much attention to user experience as they could or should. Sometimes when you find a market need that is strong enough, you can get away with selling a “merely adequate” solution. Other times, your business is specifically aimed at gathering the most profit for the least amount of work possible. Joining a company like this requires you to be cool with one of two things. Either you are content to produce “good enough” work in a system that doesn’t value you as core talent, or you are interested in taking on the much larger challenge of turning your company into more of a Design Driven organization. To be clear, thousands of designers have no problem with the former, and I have no problem with that career choice. In fact, sometimes it’s all that exists, depending on geography, experience, economy health and other factors. The latter, however, is one of the hardest and most substantial things you could ever accomplish in your career. Imagine being the person (or small group of people in this case) who got Google to care about design? If you find yourself thinking about joining one of the many Design Agnostic companies in the world today, don’t think of it necessarily as “settling”. Think of it as an opportunity to redesign the entire product development factory within. If you succeed, they’ll carry you around on a rickshaw.

The next category, Design Ignorant companies, are actually a bit easier to make an impact in than Design Agnostic companies. That is because Design Ignorant companies haven’t normalized the role of design yet. They simply haven’t experienced it yet. In fact, you may even be treated more as a hero upon your arrival at a Design Ignorant company than a Design Driven company. Going back to our example above with the file-syncing startup, imagine how much immediate impact you’ll make when you optimize the sign-up flow, improve the product copy, and flatten the navigation. If you do your job well, you’ll be appreciated right from the start. Unfortunately, you probably won’t be paid well, but Design Driven companies are usually the only companies that pay designers what they are worth at this point. This is changing for the better, but it is a multi-year process.

Finally, at the end of the spectrum, are Design Hostile companies. Think of the challenges and rewards here much the same as Design Agnostic companies except these companies have already decided that your craft is but a necessary evil for them. This by far the least desirable company type for people in our field to work at, but hey, there are plenty of cases of these sorts of companies turning around eventually. In fact, I would venture to say that using my definitions here, there are plenty of Design Agnostic companies today that ten years ago may have been Design Hostile. If you find yourself entertaining an offer from a company like this, you really need to determine how pliable they are with regard to how they view design and whether you’re ok with that.

Strategies to Know What You’re Getting Into

There are four ways to determine what sort of company you’re thinking about working at: looking, listening, asking, and verifying. You should do all four.

Looking

Thinking about working for that popular, growing auto insurance company in town? Start looking at their products. Does their visual identity seem professionally executed? How usable is their app or website? Go ahead and actually sign up if you can. Was the process reasonably well designed? Nothing is ever perfect, but often times, just spending an hour or two with a company’s product will give you a feel for how much they care about details. And when I say “details”, I don’t just mean how buttery smooth is the animation but also how smooth is the “Forgot Password” process?

Take notes as you go, in case you end up interviewing there. It’s always good to have firsthand knowledge and constructive criticism ready for when someone asks you what changes you’d make to their product (do this tactfully though, as you don’t know what factors went into a given product decision).

Make sure to also look at competitors’ offerings. While the auto insurance site may not be as modern as your favorite social networking site, maybe it’s head and shoulders better designed than all other auto insurance sites. In an industry that perhaps moves slowly in terms of technology upgrades, maybe this company is moving 10x faster than its competitors. That would be a pretty good sign.

Listening

When recruiters or employees of the auto insurance company try to pitch their company to you, they will usually do so in a way that portrays their company in as positive of a light as possible. In other words, although it does happen, you’ll rarely hear a prospective employer tell you “design is an absolute mess here”, even though it very well might be. Instead, listen for coded language. Things like:

“Over the last year, design has become a real priority.” Why? What happened before that?

“Design has a seat at the table now.” Cool, why now? What problems occurred before that?

“We’re looking to bring some fresh design blood into the company.” What are you doing with all of this blood? What problems with the current staff are you trying to solve?

Every statement should be examined for possible hidden meaning. By the way, these people are just doing their jobs. When I recruit designers, I also try to accentuate the positive. It’s your job as a candidate though, to dig deeper. Especially since your prospective employer will be digging deeply into your work as well.

Asking

In addition to responding to statements like the coded ones above, there are some good universal questions you can ask on your own:

“What is the attrition rate of designers at your company?” This should ideally be low.

“Are designers paid on par with engineers and PMs? If not, how close are they?” Pay should ideally be close or equal. Don’t be surprised if you get some bad answers or non-answers here, depending on company.

“Who does the Head of Design report to?” The CEO is always the best answer, but a great COO, CTO, CPO, GM is fine too.

“When someone needs to break a rule here, what is the process?” If they need help, perhaps give an example of a rule you’ve needed to break in the past, and ask how they’d handle it.

“What is the one negative or challenging thing about working here that no one is telling me right now and I will only find out after I start?” It’s a bad sign if they don’t have an answer.

“When someone has an idea for something they want to build, what is the process of getting approval and then building it?” Ask for a lot of detail here, right down to prototyping, user testing, and release.

“How do decisions get made when Design, Engineering, and PM can’t agree on something?” You’ll have to judge for yourself whether you like the answer.

“Tell me about a time when a product design was made subjectively or in the face of opposing data.” You might need to ask a particular person to get a good answer here, but if the answer is “never”, that’s indicative of an overly rigid decision making process.

“How are designers judged and promoted here? Is it different than Engineering and PM?” Ideally there is a thoughtful, well-articulated process in place that rewards behaviors and not just outcomes. If you’re judged solely on the metrics you move, that’s a bad sign.

It’s up to you to ask these questions with tact and at the appropriate times, but you are well within your rights to ask them. In fact, it’s a bit reckless not to.

Verifying

Do you know anyone who has worked for this company, past or present? If so, ask to get coffee with them. Depending on how secret your candidacy is, you may need to keep your questions general, but there is a lot to be gleaned from employees who aren’t trying to pitch you on anything. Start by asking what their overall experience was like. Would they recommend the company to a friend as a good place to work? Dig, if you can, into some of the things the recruiter or hiring manager told you.

“I’ve heard the decision making process is pretty egalitarian. Design, PM, and Eng all take part in that process.”

To which you might hear:

“Yep. It’s a great system. I always felt like an equal partner.”

Or:

“Haha. I am dying of laughter right now. If you’re in the room at all, it will just be to tell you what decision has been made without you.”

Importantly, when you hear things like this, you need to get a feel for whether those conditions still exist. Did this employee work there three months ago or three years ago? Is the person who enacted that decision-making process still even there? Because of the passage of time, you may get both false positive and false negative results from ex-employees. It’s ok though… it’s just a data point for you.

Also, at massive, sprawling companies, you may get different answers depending on which department’s employees you talk to. Maybe the auto insurance company’s consumer product is built in San Francisco and its broker product is built in Seattle. Maybe the working environment in San Francisco sucks but it’s great in Seattle.

So… by looking, listening, asking, and verifying, you can get a pretty accurate idea of what sort of company you are thinking about joining.

Fixing the Product Development Process, One Company at a Time

One of the reasons I wrote this piece was that I read a Tweet from a well-followed person in San Francisco talking about how the best companies in the world are all design-driven now. I get what he was trying to say, and I think that directionally, more companies are design-driven now than ever, but the vast majority of boots-on-the-ground designers in the world know how much work is still left to do. They also know that just because some high-profile companies have figured out that design is important, that doesn’t mean their own company in Seattle, or Omaha, or Bangalore has. A rising tide lifts all boats, but this is more of a slow motion wave you need to stay upright long enough to ride.

Finally, I also think it’s important to highlight the value of helping upgrade your own company’s product development processes. Designers love talking about the actual product design work they’ve done in the form of visual artifacts and launched services. Just as valuable, however, is the work that went into reshaping the processes that made these products possible. PMs make their own PM-centric contributions to what product development processes look like, and engineers do the same with their own lens. By adding your own perspective as a designer and improving the product development process at your own company, you’re accomplishing something you may not even get to accomplish at a place like Apple… and that, is something to be proud of.

(This post also available on Medium.)

After 11 years and 7 months in service, Newsvine, a participatory news site I launched with four friends on March 1st, 2006 was officially sunsetted by NBCNews.

Although I’ve been away from the company and the service for five years now, today brings back a rush of memories and some perspective on how the problems Newsvine set out to solve over a decade ago are actually the opposite of the problems that most need solving today.

In 2005, I found myself five years into a stint at Disney, wondering what was next for news. We owned ESPN, ABCNews, and several other media properties, but most of the fresh new takes on news seemed to be coming from non-traditional sources. Neither Twitter nor the iPhone had been invented yet, and Facebook was still just a campus dating site, but blogs were sprouting up by the thousands and sites like Digg and Slashdot were becoming popular destinations.

There seemed to be this growing bifurcation between mainstream media and citizen journalism. Mainstream newsrooms didn’t want to share their platform with amateur writers, and a lot of amateur writers grew more distrustful of mainstream media. Our big idea with Newsvine was to license the same Associated Press feed of professional reporting that made up the majority of what you’d see on a site like CNN.com, publish it faster than any other site in the world, and enlist citizens from around the world to create original, paid journalism around and alongside it… and open up every single piece of content for threaded discussion as well.

In other words:

CNN = AP Wire stories + Professional Journalism
Newsvine = AP Wire stories + Citizen Journalism + Discussion

… and we could do it all with a staff of under 10 people.

We didn’t know for sure if it was going to work, but the day we decided we’d be happy to have tried it even if it failed was the day we ended up quitting our jobs (incidentally, if you are thinking about leaving your job for a new risky thing, this is the acid test I recommend).

We spent about 6 months getting the company off the ground and the service into public beta, and it wasn’t long until extraordinary acts of journalism began appearing. Chris Thomas, one of our most prolific users, broke news of the Virginia Tech shootings on Newsvine before it appeared anywhere else. Jerry Firman, a 70 year old Newsviner from Ohio, got his name on the ballot for Congress and documented the process of running for office. Corey Spring, a student at Ohio State, scored an original interview with Dave Chappelle.

The design, tech, and operational work associated with growing Newsvine were fairly straightforward, but the one thing that seemed to get more and more difficult as the site grew was moderating and cultivating the community. Your first 1000 users are easy. People are just happy to be there. Then when you get to 10,000 you have a few fights here and there but nothing unmanageable. Even at 100,000, a small team of thoughtful people can stay on top of things. But when you hit 1 million, 10 million, and beyond, the community becomes much less intimate and more volatile.

Such was the case when we were acquired in 2007 by MSNBC.com (now NBCNews.com). Our site was already decently big but MSNBC’s was many times bigger… about 45 million people at the time. The post-acquisition work was twofold: 1) continue growing Newsvine as a standalone property, 2) use our technology to add registration, profiles, discussion threads, and other features to MSNBC.com. We also ended up powering all of the company’s blogs and some other things.

I ended up staying at MSNBC for about five years, and I would say the results of the experiment were mixed overall. On the upside, we provided technology that helped launch new editorial brands quickly and connect journalists to their audiences, but on the downside, “community” at that scale can be very messy. Additionally, with the eventual rise of Twitter and Facebook, Newsvine never grew to those usage levels. MSNBC.com was a great parent throughout though, and I have nothing but love for the people I worked with.

It’s interesting to compare Newsvine (and sites like it) to the now wildly successful fortunes of Facebook and Twitter. Newsvine at its core was a news site with a social network wrapped around it. Facebook at its core is a social network with news (and photos, and events, etc) wrapped around it. Twitter is probably structured more like Facebook in this regard as well, but its biggest challenge, in my opinion, has always been a lack of commitment to building those real-life social connections into the service.

When we look at how the average person’s news and media diet has changed over the last decade or so, we can trace it directly back to the way these and other modern organizations have begun feeding us our news. Up until 10 or 15 years ago, we essentially drank a protein shake full of news. A good amount of fruits and vegetables, some grains, some dairy, some tofu, and then a little bit of sugar, all blended together. Maybe it wasn’t the tastiest thing in the world but it kept us healthy and reasonably informed. Then, with cable news we created a fruit-only shake for half the population and a vegetable-only shake for the other half. Then with internet news, we deconstructed the shake entirely and let you pick your ingredients, often to your own detriment. And finally, with peer-reinforced, social news networks, we’ve given you the illusion of a balanced diet, but it’s often packed with sugar, carcinogens, and other harmful substances without you ever knowing. And it all tastes great!

As someone who has created Newsvine, worked at Twitter, and had many discussions with people at Facebook, I can tell you that this sort of effect was never “part of the business plan”. However, maximizing engagement was and still is, and that has led to a world in which what appears on people’s screens is what is most likely to keep one’s attention, as opposed to what is actually most important to know and understand.

The solutions to these problems will not come easy. They aren’t as simple as banning some jerk from Twitter or improving bot detection on Facebook. We’ve trained people to get their news and information from the cookie jar, and since we now know exactly what that world looks like, we must begin the job of untraining them… or at least engineering a healthy cookie.

We probably got a lot of things wrong at Newsvine, but one thing I still feel we got absolutely right is our longstanding tagline:

Get Smarter Here.

That’s really the only promise we ever wanted the service to fulfill.

After 800,000 articles, 65 million comments, 11+ years, thousands of new friendships, and at least one marriage and child from the site that I’m aware of, I’m confident it has fulfilled its mission for at least some who roamed its jungles.

(Special thanks to the entire Newsvine community. Without the dedicated efforts of all of you, we would have never had this special corner of the internet to write, meet new people, and have our perspectives changed. Thanks also to my partners Calvin, Mark, Lance, Josh, Tom, Tyler, Sally, Luke, Todd, Bobby, Dave, Arun, Jim, Mike, Brenda, Carl, Charlie, Rex, and everyone else at MSNBC.com for making this all possible. Also, extra special thanks to Nick, one of our investors, for introducing me to my wife, who I would have never met were it not for this little chance we took. And finally, thanks to my wife who helped get me through everything back then and since.)

(This post also available on Medium.)

If you were intensely motivated, you’d pick up a phone book, call Pepsi headquarters long-distance, hope to get routed to anyone close to the Crystal Pepsi “user experience” division, and then come up with some sort of feedback smart enough to justify the amount of effort you just exerted. Or maybe you’d handwrite a letter and mail it off to a possibly unmonitored mailbox.

Barely anybody did this, and the end result was that a lot of products, including Crystal Pepsi, failed without their creators knowing quickly enough why people were rejecting them.

But now we have Twitter! And evvvvvverybody has feedback.

On evvvvvverything.

Always.

The downside is, when feedback is so easy to provide, often less consideration goes into it, making it less valuable to designers, PMs, and engineers as they try to address problems.

“Oh you mean I only have this little text field for the longer novel I could easily write about these terrible changes of yours? Ok then… I’ll just type a provocative sentence and be on my way!”

Feedback is a gift, and if you are lucky enough to work on a product that people care enough to give feedback on, you should take it very seriously. However, as professionals in product design and development, giving good feedback is also a responsibility of ours. We should be the best in the world at it! Unfortunately, we aren’t always.

Instead of concentrating on providing feedback that will help our colleagues improve their work, we often spend our time thinking of remarks that will make us look or feel smarter.

“What can I pack into 140 characters that is so insightful or funny or provocative that a bunch of people will fave and retweet it?”

I know I’ve certainly done this before. While it’s a free country (for now), and people can say whatever they want, as professionals who work in product development, we owe our colleagues more than that. We owe them civility by default and an honest effort at being helpful. In order to be helpful, we need to think a little deeper and provide feedback that reflects our understanding of how products get built.

To help us all get better at this, I came up with a framework to better guide our critiques of the changes, redesigns, and launches of our colleagues. Each level in this framework reflects a deeper level of empathy and understanding which ostensibly leads to more helpful feedback.

Level -1: Troll

Intentionally demean others or their work.

Example: “This is awful and whoever prioritized or produced this work is an idiot.”

Appropriate Usage: None. There is no good reason to ever do this. This type of statement does nothing to solve any problems and insults human beings you’ve never met, who are working under circumstances you are blind to. Type out a Tweet if you must, delete it without publishing, and then move on.

Prerequisites: Being naive, inconsiderate, or both

Level 0: React

Put into words the way something makes you feel.

Example: “Something about the navigation just feels wrong.”

Appropriate Usage: This one is tricky because for laypeople, it’s ok, but for professionals, we need to do better. In other words, I don’t begrudge an average user for not being able to express why they don’t like something about the navigation. That’s their feeling, and if there is the bandwidth for it, designers and researchers can further explore that in a conversation. As professionals though, we should keep statements like this to ourselves until we can more constructively communicate why we feel the way we do.

Prerequisites: None

Level 1: Identify

Point out an actual issue that you have encountered or worry others may encounter.

Example: “The navigation is more cumbersome because the page I use most — my Profile — is now two taps away.”

Appropriate Usage: This is the baseline level of feedback we should expect from ourselves as professionals. It doesn’t really delve into solutions or inquire about what tradeoffs might be involved, but it crisply communicates a problem so that designers on the other side can consider it and respond/remedy appropriately. It can also be more subjective things like a certain shade of green not going well with a certain shade of purple, but it needs to be articulated as specifically as possible.

Prerequisites: None

Level 2: Understand

See something that looks sub-optimal and begin an inquiry to figure out why.

Example: “The Profile page is much harder for me to get to right now. What advantages are achieved with this tradeoff?”

Appropriate Usage: Now we’re getting somewhere. Product design is a series of tradeoffs. Simplicity vs extensibility. Speed vs power. Whimsy vs seriousness. When designers move something around, it is usually with an objective in mind. That objective could even be to intentionally get you to do something less frequently. Designers, however, often misfire and don’t fully appreciate what they are giving up for what they are gaining. Worse yet, sometimes they may be gaining nothing at all. It is more than fair to ask these questions. It is helpful! Often in the course of giving Level 2 feedback like this, you may in fact cause the designer to think of this tradeoff for the first time, since they may not have even realized the magnitude of the downside.

Prerequisites: Observations from Level 1

Level 3: Propose

Identify a problem, understand its cause, and propose a possible remedy.

Example: “Ah. I understand it makes sense for my Profile to live with its related navigation items and they won’t all fit down below. Have you considered making the nav customizable so people can swap in their Profile if they want?”

Appropriate Usage: Level 3 feedback must be given in the right way, because if not phrased properly, it can come off more like Level -1. For instance, “Why the hell wouldn’t you just make this all customizable you idiots?!?!” is a lot different than the example text. That sort of phrasing is not only inconsiderate, but it also assumes the issue is simple… and they rarely are. The trick with Level 3 and Level 2 feedback is an exchange of empathy. You want designers to have empathy for you and the problems you are encountering with their product, and you also want to have empathy for the designers and the problems they are trying to navigate through, whether it be deadlines, business goals, conflicts between stakeholders, resource constraints, or any number of other things that we always try and keep behind the curtain. To be clear, average users do not and should not care about these things… all they know is the experience that has been put in front of them. However, as professionals we know better than to discount the complexity of what goes into our work. A good rule of thumb is: if a problem seems simple to you, you probably don’t fully understand it. You certainly might, but you probably don’t, and therefore, you should treat your critiques as investigations or explorations and not conclusions.

Prerequisites: Observations from Level 1 and findings from Level 2

You may notice a pattern here in that as you go up each level, more thinking is required. You may also notice that a lot of the most valuable feedback you can give may not actually fit in a single Tweet. It may require you to first reach out and get more information before you form an opinion. It may even inspire you to write a short article expanding your critiquelets into well-formed, helpful observations, which not only end up making it into the product, but may end up getting you a great job somewhere.

If all of this sounds like a lot of effort, think of the hundreds of hours of someone else’s work we so brazenly and publicly try to cook down to a single sentence. While we can’t avoid the barrage of hot takes from people who have never even heard of product design, we can absolutely commit ourselves to being as helpful as possible when we critique the work of our colleagues.

Thanks to Marc Hemeon, Mike Rundle, and Kory Westerhold for reading and improving drafts of this.

(This post also available on Medium.)