Results from the Google AdSense Experiment

Well it’s been a week since I threw a Google ad into the header region of this site and the results are in. With just under 50,000 page views, my earnings were just under $100.

It was sort of a low traffic week with not a lot of posting, linkage, or commenting activity, but even so, $400 a month isn’t enough for me show that huge ad to everyone who comes to this site.

But still… it’s more than a few pennies so I wanted to come up with a solution which would let me continue to show the ad in some cases and not show it in other cases. I could frequency-cap it but that would change the design in the middle of a user session (not good). I could only show it a few days a week, but that would be a similarly inconsistent experience.

In the end, I came up with a happy solution: show the ad to everyone except subscribers to this blog. That way, people who come here often and have more than a passing interest in what is written here will get a nice, clean ad-free experience, while those who arrive here via searches for MySpace Layouts, sIFR, or other linkage will see the ad and in some cases help monetize it. This will likely result in a small revenue decline, but that’s of little concern to me.

Currently, I’m achieving this conditional ad serving by setting a three month cookie every time someone clicks on a link from my RSS feed. As long as that cookie is around, you should see no ads. I’m also considering extending the ad-free experience to every visitor who doesn’t come in with a referrer. That way, people who have Mike Industries bookmarked won’t see ads either.

That’s it. Happy Thanksgiving!

45 comments on “Results from the Google AdSense Experiment”. Leave your own?
  1. Calvin Tang says:

    I still see the ad, even though I came in via Bloglines. Does this mean you haven’t implemented the new cookie yet? Mmm..mmmm…. cookies.

  2. shorty114 says:

    Isn’t it against the Google AdSense terms of service or whatever to publicly disclose your AdSense earnings? Or maybe I’m thinking of something else…

  3. Mike D. says:

    Calvin: Try again. I think Bloglines uses the GUID field as the clickthrough URL, so I just fixed the GUID field to use the correct cookie-setting URL.

    shorty114: Good catch. I have deleted the full data table and am now just posting two approximate data points (page views and revenue). Hopefully that’s enough obfuscation.

  4. Alex says:

    This is a brilliant idea. It didn’t work for me from bloglines either, though.

  5. Sonny says:

    Great idea. I still see ads on every post although I subscribed to your feed and come from Google Reader. I also tried to visit through bookmark (del.icio.us add-on on Firefox), the ads still showed up.

  6. Ulises says:

    i came in through my firefox live feed, i don’t see the ad.

    Saludos!

  7. Keith says:

    Another one arriving via Bloglines, and the ad is still there, Mike. The only cookies I see set from your domain are those for Mint and the site style.

  8. Clay Smith says:

    It works from Vienna on a Mac. Fantastic idea, by the way.

  9. I just don’t see the ad period. All I see is a big black box where it is supposed to be.

  10. Joshua says:

    What about just adding an ID to your body tag, that way us regular readers can just add a CSS rule to our browser stylesheet.

    My ad blocker blocks it out anyway, but could be useful to some.

  11. Tom W.M. says:

    For me, in Google Reader, every post since “Low Involvement Fantasy Football” popped up again as unread–duplicates. I guess that this is why Atom feeds have GUIDs independent from the clickthrough link URL. I’ve just switched. (On a more positive note, I don’t see the ad.)

  12. For those of you coming in off bloglines, remember that the old page will still be available as unread. You might have seen two posts from Mike today that seemed to be the same (I know I did). Basically, the next article should fix the cookie problem.

  13. Mat says:

    Well, now i’ve subscribed to your feed just so I don’t see the ad. Marketing strategy? :) $400 a month is a nice average, it’s worth keeping unless people really start complaining about it.

  14. Jason Beaird says:

    Great idea Mike. I hadn’t thought to use RSS traffic to determine whether or not to show ads. Back in April I set up Adsense to show to all visitors from search engines. I just checked the referrer for “q=” or “p=” which catches most of the search engine crowd. I don’t feel as bad hitting those people up as I do the non-referrals and links from other blogs.

  15. Jeff Wheeler says:

    Funny you do that. I have been switching from Mac to Linux lately, and still have all my RSS feeds in NNW. So, when I saw this post, I typed in your site’s address on my other machine, and still saw the ad (of course).

  16. Caleb says:

    Concerning the legality of posting the click count and earnings:

    Xooglers ran into some issues before. The comments on that page also provide some insight.

    Jason Edelmen was hit as well by originally posting this. You’re at least on the fence but I’m not sure if posting the 50k clicks and earnings is enough to be expecting words with Google.

  17. Daniel says:

    Seems like an extremely fair way to go… remember, mate, this is your blog. You can do with it as you’d like, and making some spare change from it doesn’t compromise its quality at all in my mind. I’m subscribed (using Safari’s built-in RSS reader) and the ads aren’t showing up for me — but even if they were, I wouldn’t really care. As it is, though, making the ads show only for random, one-time visitors is a nice way of taking care of things.

  18. Bramus! says:

    Very cool, when viewing MI in IE (my not so default browser) the ad is shown, in FX it isn’t. Kudos!

  19. Devon Shaw says:

    Cookie obviously works great in Safari. Came in straight through the built-in RSS and nothing at the top.

  20. PanMan says:

    In bloglines, I now have 2 links:
    https://mikeindustries.com/blog/archive/2006/11/results-from-the-google-adsense-experiment
    and
    https://mikeindustries.com/feeder/?FeederAction=clicked&feed=Main+%28RSS+2.0%29&seed=http%3A%2F%2Fmikeindustries.com%2Fblog%2Farchive%2F2006%2F11%2Fresults-from-the-google-adsense-experiment&seed_title=Results+from+the+Google+AdSense+Experiment
    The second one doesn’t show ads, the first one does. But it can be that it’s just that the cookie is now set, or so. But I do use bloglines on different PC’s, so the cookie won’t be set on all. Can’t you just include something like ?add=no in your RSS feed?
    Anyway, great solution. I know RusselBeattie did something similar, when his blog was still active.

  21. Dave says:

    Ah nice $100 in a week is not bad money Mike.
    Although im sure newsvine is keeping you well ;)

  22. Stu says:

    I dunno if you don’t want to understand that, but THE AD IS PLACED WRONG. Everyone with a little knowledge about Adsense will tell you that. You could earn more if this ad wasn’t a stupid large thing sitting at the top of every page.

  23. I came in via the second Bloglines link and it worked okay.

    Two things though:

    1. You may want to consider playing with that ad Mike. I get less page views but more revenue then you do.
    2. Can we opt back into the ads? Call me crazy, but contextual advertising kind of makes sense to me and I do find the ads useful sometimes.
  24. Martin says:

    Just wondering.. Out of 50.000 views.. How many are “subscribers/none-refferes”?

  25. claus says:

    did i miss something: you are making sure no one from the past will be able to use your postings without infringing your copyright, or why is it you changed your footer to “1000 B.C. to 2006 A.D.”? ;-)

  26. Stephen says:

    I remember a few years back when I first tried adsense, and every click seemed to yield between $1-2. Then shortly before the IPO, this turned into mere pennies per click. Now the stock is shooting past the $500 mark. Bastards.

  27. Mike D. says:

    Everyone: Yeah, Jonathan is right. If you happened to click on the feed item in Bloglines (or Google Reader) before I fixed the GUID, you may still see the ad. The most recent post should fix it though.

    Josh: Done! All of my body tags are now ID’d with “mikeindustries”. I think the fact that you make site-specific tweaks via the body id attribute clearly makes you AB-normal though. :)

    Bramus: It’s possible the cookie isn’t get set correctly in IE… I should check that. Never tested in IE. :)

    PanMan: The newest copy in Bloglines should not contain two URLs. You’re probably (hopefully) looking at the pre-updated item.

    Stu: THE AD IS NOT PLACED WRONG. It is placed where I want it placed… away from the rest of the design of the page. You left a rude comment here earlier, something to the effect of “you’re dumb for putting the ad at the top of the page… even a little kid knows better”. The fact of the matter is that I can place 10 different ads in any portion of the page I want. But my concern is not maximizing revenue. It’s just adding a little bit of revenue without touching the rest of the design. I may eventually experiment with different placements — especially now that I have them turned off for subscribers — but I’m happy with the top of the page for now.

    Mike: Opting back into ads! Wow! I’ll see what I can do.

    Martin: I’d estimate that out of the 50,000 page views, about 20% came from RSS. Could be wrong though.

    claus: Yeah, I modified the footer a couple of months ago so time travelers would know they cannot steal my content and take it with them back to the past. :)

  28. Chad Edge says:

    Mike, I agree with Mike P (opting in). I added google ads the same damn day (I feel like johnny come lately now) as you, and the first ad I saw on my own page was one I had to click on (nested select tree’s with all/some/none selection).

    Also, I’m not an RSS user, but you and 2nd avenue (as well as curious office) are my top first URL’s in my daily routine.

  29. Not sure if others have suggested this, but why not consider running other types of ads than Google Ads?

    You might, for example, ask to be included in Coudal’s “The Deck”, which would likely give you higher quality ads than the crap-shoot Google gives you. However, one downside is you have no control over the format of the ads (they’re all rectangular).

    Or you could roll your own ad service – sponsorships, etc. This requires a bit more committment, but it might end up being a service to both your readers and your wallet. (Just make sure you use rel nofollow – sponsored links get pretty shady when people start selling pagerank)

    If you did your own ad service, you could really get them targeted. Serve up ads for font founderies to people who come looking for sIFR, or serve up ads for addiction counseling for those who come looking for MySpace layouts :)

    My point is, Google Ads are a crap shoot at best. With a bit of effort, you could have far more relevant ads, which leads to happier users and more money.

  30. The Bloglines fix appeared to work.

    For those of us without your very capable skill, can you explain how you look at the referrer? Is it part of your blogging platform or apache or what?

    Thanks again for the answer in advance.

  31. Mike D. says:

    David: I have all URLs in my feeds pointing to a PHP file which does a redirect to the actual URL. When redirecting, I simply set a cookie in PHP which says “subscriber=true”. I never actually check the referrer.

  32. John says:

    An ad on a personal website should never trump the masthead of the site unless the owner is looking at only the monetary aspect. It looks like something from Geocities. It would be better placing it off to the right side so it is not the first thing people see when the page loads.

    Just as an aside, and nothing against your site in particular, I never click ads. Never. I know what I like and when I want it. I don’t like being sold or marketed to in any way. I stopped getting the junk mail in my snail mail and I never watch commercials on TV. I’m one of those people who find advertising anathema.

    Word of mouth always is the single best way to generate business. The Yellow Pages is the only adverting I can stomach since it’s not being thrown at me without my consent. I must consult it.

  33. John says:

    As an aside, referrer logging is easy to beat using Opera or even Firefox. One could come in from anywhere, referred or not, and if that setting is there in the browser, the ad would not appear based on the non-referrer condition.

  34. Mike D. says:

    John: I’m not checking your referrer, so that isn’t relevant. I’m checking to see if you clicked over from a feed (not by checking the referrer). Regardless, if people want to manually set the cookie in order to hide the ad, that doesn’t bother me either.

  35. Adam says:

    Yahoo! Publisher brings more dough in, I have heard.

  36. Ryan says:

    It worked fine, all I did was hit “add the Live Bookmark” on Firefox and loaded the site, ad was gone.

    Nonetheless, no adds for RSS subscribers. Capital idea, ol’ sport.

  37. Brad says:

    I know mike doesn’t want to hear this, but I’m saying it anyway:

    The ad position really bugs me, too. I can’t scroll past it fast enough. I’d suggest at the end of an article, for the ol’ “ok, I’ve read this, now what’s next?” approach.

  38. Matt says:

    I also think you’d be a good addition to the deck.

  39. I think that makes total sense. I was doing traffic analysis on a site that receives 100K+ visitors a year. Our results indicated that loyal visitors knew the layout and visually blocked the ads. However, flighty traffic like search visitors were incredibly likely to click on ads because they’re just blowin’ through anyway.

  40. Devon says:

    My question is…how much money is it generating now that you’re only showing it to non-subscribers? Even more specifically, does the amount it generates, at least pay for the costs of running the site?

  41. Mike D. says:

    Devon,

    It hasn’t dipped a whole lot since I removed the ad for RSS subscribers. And yeah, it costs me $7.95 a month to run the site so it more than makes up for that.

  42. Mind if I steal this idea? It’s ingenious!

  43. Leon says:

    Interesting.

    “Stu: THE AD IS NOT PLACED WRONG. It is placed where I want it placed… away from the rest of the design of the page.”

    It’s the first thing you see so how can that be “away from the rest of the design of the page” it is integral to the design now it’s on the page.

    “But my concern is not maximizing revenue. It’s just adding a little bit of revenue without touching the rest of the design.”

    It’s all about revenue, regardless of how much, otherwise why have it at all, introducing ads to a popular site, when previously it had none, could have a negative effect, it now intrudes and spoils the design of the site and certainly not “without touching the rest of the design” and calling it an ‘experiment’ could be a little misleading, are your readers guinea pigs to make you money ?

    It obviously doesn’t effect the content of the site but it isn’t pretty.

    Either have the add or not.

    Thanks for an interesting read.

    Regards

    Leon

  44. Mike D. says:

    It’s the first thing you see so how can that be “away from the rest of the design of the page” it is integral to the design now it’s on the page.

    Exactly as it sounds. It’s not touching any element on the page. It is thus, away from the rest of the design.

  45. […] I got this idea while experimenting with Google AdSense several months ago. I only wanted to show the ads to casual passers-by (perhaps coming in through search engines) and […]

Leave a Reply

Your email address will not be published.

Subscribe by Email

... or use RSS