AdSense Experiment Results

Last week, I announced I would be conducting a little AdSense experiment to gain insight into how often different kinds of visitors click on ads. Today, I will present the results.

The core motivation for the experiment was to test the prevalent assumption that people coming from search engines are the ones most likely to click on ads. To this end, I used a little piece of JS that would present different ads to people based on how they arrived on this site – either directly, from a search engine, from another kind of external site, or from a different page on the same site. All the ads were visually identical but were tracked separately.

The Numbers Are In

As it turns out, search engine visitors are not the ones most likely to click on ads – when different types of traffic are ranked by the click-through ratio (CTR), search visitors actually fall into the second place. The most eager ad-clickers are the people entering the page via an internal link. Here’s a comparison chart :

My apologies for hiding the units. AdSense prohibits one from publicly disclosing their CTR and related figures.

While this might not be obvious from the chart, the actual CTR disparity between internal, external and search traffic is very low (less than one percent). However, direct visitors – e.g. people accessing the site via a bookmark or clicking a link in an email – are significantly less likely to click on ads.

CTR alone doesn’t give us the full picture – it tells us how many clicks we can expect to get if we display our ads to a certain type of visitor, but not what those clicks are worth. To judge the financial value of a visitor we need to consider CPM :

Again, the specific numbers have been hidden to conform with the AdSense ToS.

The same trend is even more pronounced here – people clicking on internal links also deliver the most valuable AdSense clicks, people arriving from search engines come second, and external & direct visitors are, on average, the least likely to bring in oodles of cash.

Conclusions

Over the course of the experiment, the pages that had the ad-varying script installed received about 23 000 impressions and over 1500 AdSense clicks. I feel that those numbers are big enough to call the results statistically significant.

Given that, and how close the CTR values of search, external and internal visitors are for this site, it wouldn’t really make sense to show ads only to one of those groups. However, I can safely stop showing AdSense to people who visit the site via bookmarks and not lose anything much, income-wise.

Now the question that’s probably on your mind is : “How does this translate to my site?” Should you show ads to everyone, or just a certain kind of visitor? Now that I’ve demonstrated that the folk wisdom of only showing ads to search visitors is not necessarily valid, how are we to maximize our financial fortunes?

It depends ๐Ÿ™‚

I’m not going to substitute one piece of dubiously specific advice for another. Instead, the take-away of this exercise is :

Don’t take AdSense micro-optimization advice on trust. Test it, and use what works for you!

Related posts :

8 Responses to “AdSense Experiment Results”

  1. adbox says:

    Great post! Any plans to extend study?

  2. White Shadow says:

    Thanks ๐Ÿ™‚ As for extended studies, do you have anything specific in mind?

  3. adbox says:

    I’d like to see if the 1% holds up over the span of a year. Maybe a revist of subject every couple of months?

  4. White Shadow says:

    That might work. I suppose I’ll leave the ad-variation script running and see if/how the stats change over time.

  5. […] Experiment : The Follow-Up document.write(get_top_adsense_ad()); Way back in March, I conducted a little AdSense experiment to see which type of visitors (direct, coming from internal or external pages, or arriving from […]

  6. Bruce says:

    Great post! Any plans to extend study?

  7. White Shadow says:

    Thanks ๐Ÿ™‚ The ad variation script is still running, and I’ll probably do another follow-up in a month or so.

  8. Mathijs says:

    Thanx! Super post! Heads up from Holland!!

Leave a Reply