Google Adsense Spy
Google Adsense Spy was supposed to work like this. You’d start typing, text would appear on the page, and Adsense advertisements would adjust to the content and magically update below the content, all without any page refreshing.
I created the Google Adsense Spy because I thought it would help me understand my writing in respect to Adsense. Likewise, I thought the tool would help me understand how Google looks at content, which is something I’m very curious about.
Let me be clear. In no way was I trying to game the system. Instead, I was running an experiment. If it is possible to write two paragraphs that have the same meaning but the advertising is better, why not understand it? Also, in the universe of Adsense, isn’t it interesting to think about writing around advertising versus advertising wrapping around content? And, who’s watching the watchers?
Google has far more power than any single writer or content producer. I think it makes sense to investigate how to leverage that power. I’d rather ride the lightning than have it strike me. That is, if we know the nature of Adsense then we can rethink our content. We can consider how content and advertising actually form a symbiotic relationship. As it stands now, Google Adsense is a parasite on our content.
Interestingly, the parasite is far more powerful than any single content producing host. Should content producers ask more questions about the advertising and the advertiser? I think the answer is obvious. I’m an investigator of sorts. Google Adsense Spy was meant to help. Unfortunately, it failed.
I didn’t expect failure or I wouldn’t have built the tool. But it did fail. The reason is pretty simple. My impression was that Google scanned a page and delivered targeted advertising. That is only partly true. Google does scan a page but it doesn’t seem to happen instantly. That was my backbreaking assumption:
“From the day you start offering ads, the AdSense robot visits on a regular basis, reading through your pages with ads. The robot takes a look at the words you use, the frequency with which you use them, even some of your page structure and formatting (for example, bigger fonts usually signify something important). Then Google uses all this info to figure out which ads your readers will warm to.”
So, it is not possible for my Google Adsense Spy to work because the page has not been created and indexed. Therefore, Adsense can’t be smart about the content the instant is created. The ads will not match the text.
With all this said, I hope you give Google Adsense Spy a quick try. It is fun to think about how it could have worked. Type in some text, wait about eight seconds, and watch the Adsense ads flicker. If it worked as expected, the ads would have updated in light of the content you typed. No dice, but still fun in theory, right?
April 17th, 2006 at 11:51 am
I agree that it could be a nice tool from purely scientific reasons - but I do not like it from an editorial perspective. The risk of “content contamination” is too high (even if it in not intended).
April 17th, 2006 at 1:24 pm
I was more interested in the experiment and the pure science. I would not really alter my writing to get better advertising but I’d still like to understand what’s going on, just for fun.
I can imagine that one audience that would care about this: SEOs. Some SEOs might wish to dig deeper into Google and also their own content. Anything to get that special edge.
April 17th, 2006 at 3:20 pm
Even though you might not alter your writing to get better advertising, I’m sure there are others out there who would. For instance, at InvestorGeeks we are trying to post articles about more popular topics in an effort to increase our search traffic. The end effect of this is hopefully more people clicking on our ads. I guess we can still avoid an evil title since we’re also reaching more people with our writing, which could be seen as a noble effort.
Still, I can imagine scenarios of people catering their content in such a way as to increase ad revenue without necessarily increasing readership. For example, some keywords have higher costs than others. Websites built around Real Estate and Search Engine Optimization might make more money than ones about potty training your cat or making cookies. It’s just a small step from choosing a page’s topic to making sure specific ads appear on your site.
Now, is this really a productive way to make more money? I would guess no. Remember, Google wants you to make the most money possible as well. They gain when you get ad clicks too. As a content developer (interested in maximizing ad revenue), I would write to bring as many people to the site as possible and leave the ad selection up to Google. That means writing clearly about interesting topics. It may not seem like it, but Google really does have the most to gain by showing the best ads on your site. Leave it to them.
John, while were on the topic, you could stand to gain by updating your AdSense layouts. I’ve been part of two very successful conversions this weekend alone. Of course maximizing your AdSense revenue may not be your first priority with the site, but if it is feel free to send me an email if you want some help selecting and placing ads.
April 17th, 2006 at 3:25 pm
As an experiment it is very interesting. There are probably many things we can learn from Google in understanding what content is about. Getting information in an automated fashion is something we in the usability community is very poor at.
We look at varies ways categorize content, either trough traditional linear categorization, or trough much more complex approaches like tagging and faceted navigation - but we have failed to find a usable way to automate this. And, this failure is an obstacle in itself to good usability (nobody want to focus on categorization - we want to focus on creating content).
In this sense your Google Adsense Spy is very interesting. I could help us to get a better understanding of automating content processing - in many scenarios (including those who do not include ads).
April 17th, 2006 at 4:34 pm
Jason,
Very well written, very clear posting. Thanks for taking the time to share your thoughts.
Regarding the use of Adsense and the best layouts, I’m very interested. Did you happed to see the article I wrote for Apogee?
Is Google Advertising Evil?
http://www.apogeehk.com/articles/Is_Google_Advertising_Evil.html
Of course, this article is more about placement and blending than the type of advertisement. Then again, I’m interested in all aspects of the discussion. I’ll be in touch. And at the same time, if it makes sense, post your response here.
April 17th, 2006 at 4:42 pm
Thomas,
I agree with many of your comments. I’m often torn between thinking that automatic mechanisms are good versus bad or useful versus substandard.
However, upon reflection, I don’t think that the trick is making something automatic or not. That’s the wrong question. Instead, it is knowing *when* to make something automatic or not. Everything from there is implementation, either human or computational.
Related: Computers versus Humans
http://www.theopavlidis.com/comphumans/comphuman.htm
Related: Humans vs. Computers
http://www.wired.com/wired/archive/10.08/start.html?pg=11
April 18th, 2006 at 12:46 pm
John, I’ve always been a fan of your writing. After your call for more people to respond to your articles, I realized how useful this site could be if there was more of a discussion around the stuff you post. So I’ve been replying to your site the past few days to encourage this sort of thing. Let’s hope it works.
I did see your post for Apogee, and I had seen Google’s diagram before. It’s very interesting. People initial though that placing ads in a way to disguise them into the site was “evil” since you might be tricking people into clicking on the ads. And Google obviously (well maybe not) doesn’t want you to accidentally click on an ad. However, the reason it’s become more acceptable for people to disguise their ads (through placement or color choice) is because it’s generally accepted now that people don’t accidentally *click* the ad; they just accidentally *read* the ad. If after they’ve read it they decide to click on it, everyone wins.
So some real quick pointers for increasing you clic-kthrough on the site:
- Remove the borders on the sidebar ads on the homepage. Also, make the background the same as the rest of the sidebar.
- Another thing you can do is add an eye-catching image right above the ad-list. We use a quirky image of a bull in a field at www.investorgeeks.com.
- Add inline medium 300×250 ads in the middle of your articles. Make sure to use the same color you use for links for the ad headers.(https://www.google.com/adsense/inline) You can see how we use it at InvestorGeeks here . This is a bit annoying for readers, however these are the most lucrative forms of ads. The reason (I believe) this works so well is because users eyes are just unavoidably drawn to the ad headers when they look so much like links. We’ve been trained to scan a page for links. It’s possible that users will eventually be trained away from looking at these ad blocks like they’ve been trained away from looking at flashy banner ads, but that is probably a little ways away. Since we’ve implemented these inline ads at InvestorGeeks our ad revenue has tripled.
John, I’ve always been a fan of your writing. After your call for more people to respond to your articles, I realized how useful this site could be if there was more of a discussion around the stuff you post. So I’ve been replying to your site the past few days to encourage this sort of thing. Let’s hope it works.
I did see your post for Apogee, and I had seen Google’s diagram before. It’s very interesting. People initial though that placing ads in a way to disguise them into the site was “evil” since you might be tricking people into clicking on the ads. And Google obviously (well maybe not) doesn’t want you to accidentally click on an ad. However, the reason it’s become more acceptable for people to disguise their ads (through placement or color choice) is because it’s generally accepted now that people don’t accidentally *click* the ad; they just accidentally *read* the ad. If after they’ve read it they decide to click on it, everyone wins.
So some real quick pointers for increasing you clic-kthrough on the site:
- Remove the borders on the sidebar ads on the homepage. Also, make the background the same as the rest of the sidebar.
- Another thing you can do is add an eye-catching image right above the ad-list. We use a quirky image of a bull in a field at www.investorgeeks.com.
- Add inline medium 300×250 ads in the middle of your articles. Make sure to use the same color you use for links for the ad headers.(https://www.google.com/adsense/inline) You can see how we use it at InvestorGeeks here . This is a bit annoying for readers, however these are the most lucrative forms of ads. The reason (I believe) this works so well is because users eyes are just unavoidably drawn to the ad headers when they look so much like links. We’ve been trained to scan a page for links. It’s possible that users will eventually be trained away from looking at these ad blocks like they’ve been trained away from looking at flashy banner ads, but that is probably a little ways away. Since we’ve implemented these inline ads at InvestorGeeks our ad revenue has tripled.
We used the “Adsense-Deluxe” plugin for WordPress (http://www.acmetech.com/blog/adsense-deluxe/) to add the ads to our posts. My colleague set it up, but it sounded pretty simple. The one draw back is that to include the ads in old posts, you’ll need to go back and ad a comment to each one.
Good luck.
April 19th, 2006 at 9:31 am
Oops. Put two copies in that last post. Sorry about that.