I got an interesting email from Jonathan Fields over the weekend (I subscribe to his list) that offered a review of a new site/service along with the usual affiliate link. Except, this wasn't simply "the usual affiliate link." Jonathan, a relentless experimenter, chose to post two different links--one clearly labeled as an affiliate link (Jonathan gets a commission), and one leading to the same site that was clearly labeled as not an affiliate link (i.e., Jonathan gets nothing if you sign up using that link.) I was fascinated by this approach, and was curious to know his results. Well, today, Jonathan posted that X% of clickers chose the affiliate link, while Y% chose the non-affiliate link (sorry, dear readers, but X and Y are spoilers. Read his post!)
Now, I am not in the business of affiliate marketing (and, for the record, my links are never affiliate links--not that there is anything wrong with them.) And you might be tempted to read the results of Jonathan's experiment and draw some conclusions. However, I'll point out two things: first, readers of Jonathan's blog (and, in my case, his email newsletter) are on average more predisposed to support Jonathan than not, so the percentage of "affiliate clickers" has to be weighted by that. But my main point is this: when I read Jonathan's interesting experiment, my first question was not which link would be more popular.
My first question was whether or not the total number of clicks to both links exceeded his typical article clickthrough rate on a single link, affiliate or not.
I'll make an admission here: I clicked on the non-affiliate link, but only because I knew I wouldn't be buying the service he linked to. But I clicked. And I have to wonder--does the clearly labeled presence of affiliate AND non-affiliate links increase the likelihood that either link will be clicked? In other words, is there a significant percentage of people who refuse to click on affiliate links because they question the motives of the author, or for some other reason, and does the availability of an "out" increase the probability of action?
To put it another way, some might look on the "Y%" who clicked the non-affiliate link as some kind of referendum on the relationship the author has with those possibly peripheral or otherwise unengaged readers. But did that "Y%" actually lead to more clicks than an article that only contained affiliate links? Or an article that contained (illegally) only affiliate links that were not identified as such?
I wonder if what Jonathan has actually learned here might not be more profound. I hope he'll share that data with us--because if the disclosure of bias, no matter how slight, actually leads to increased conversion, I think we'll all be happier.