About three and a half months ago Google’s Adsense team decided to venture out into the murky world of reader profiling. Prior to this Adsense had targeted content and based on what your particular page was about, would suitably deliver targeted ads to match.

Whilst this works great for written content sites, sites that contained other forms of multimedia content (audio, video, photos) for example were often not targeted as effectively as they could be. What would tend to happen is you’d either wind up serving up a small number of ads over and over again, showing irrelevant ads or the annoying public service announcements.

Interest-based targeting (which is really just a fancy name for reader profiling) changes the focus from the content and instead starts to keep track of what interests visitors to your site have.

This information is based on previous visits to other categorised sites and long term browsing habits. Whilst this might be good news for non written content sites as they have the potential to gain more relevant advertising, written content sites won’t necessarily share this improvement.

Here’s why.

1. Internet surfers are fickle.

Unless you’re one of those weird people that use the internet for one thing and one thing only, you’re probably going to be hard pressed remembering what it was you were searching for yesterday, let alone 10 minutes ago.

Don’t believe me? Go and open up your browsers history and see if you can remember half the stuff you searched for today. If that’s not challenging enough set it to show the last 7 days browsing history and see what you can remember.

A typical internet user burns through that many searches it’s pretty impractical to group them into user-based categories and hope they still haven’t found what they were looking for a day, or even a few hours after they stopped looking.

2. Time-delay relevancy

Personally when I’m scanning a webpage I focus things that are relevant to what I’m looking for at that particular time. Eg. if I came to your site looking for information on lightbulbs and a bicycle ad was prominently featured it’d have to be a damn good ad to detract my attention away from your aeroplane content, despite me being an avid cyclist.

Right then and there I’m not necessarily interest in bicycle deals, my mind is focused on lightbulbs.

Meanwhile an ad on say lightbulb sales would be much more likely to catch my eye as it fits in with the content I’m accessing and you’re more likely to get a click.

3. Topic familiarity

Most people who visit my blog are looking up specific information in regards to something. For this reason I try to have complementary advertising that might enhance their reading experience.

To use the cyclist example again, I’m an avid cyclist and as such am pretty fixed in terms of where I buy things and what I read in regards to cycling. Most people with well established interests are the same.

Because people are creatures of habit, capturing the attention of someone who is well versed in a particular topic is much more difficult. I am far less likely to click on a cycling ad to learn more information about a particular aspect of cycling, then if the ad was about something I’d just searched on but knew a lot less about.

For this reason although it might be nice statistic wise to group readers into interest-based categories, the ads you display when this is in effect have to be that more engaging to attract interest and compliment your content.


Taking those three points into consideration and in my own testing over the last few months I’ve concluded that for written content based sites, Adsense’s interest-based ads aren’t really that effective.

If however you find that there simply aren’t enough relevant ads for whatever it is you’re writing about it’s certainly a much better idea to show interest-based ads then hit and miss non-relevant ads. The same applies for photo, video and audio dependant sites.

Adsense’s interest-based ads feature is something I’ve been thinking about on and off for a few weeks now as I’ve weighed the pros and cons and tried to better understand reader engagements with advertisements. Hopefully my thoughts give some other publishers out there some ideas on what may or may not be working if they’ve noticed a drop in revenue over the last few months.