Well it’s been 32 days since I started the Adsense rich media ads experiment on my highest earning ad block.
I’m using data from June 6 to today, July 8. The 2 or 3 days the server was hacked will be included because both ads would have been affected the same, and earnings were very low on those days anyway so any statistical noise will affect data very little.
So here are the results.
I’d rather not disclose exact dollar figures, but I’ve disclosed the important data for each ad unit. As you can see, they are almost identical. The key numbers to look at are RPM – or revenue per 1000 visitors for each ad type. Over the last 32 days, the revenue for text-only ads has been almost exactly the same as the revenue for rich media ads. The difference is only about 2% – and it’s actually the text ads which are earning slightly more! So it would seem at first glance that google’s advice to switch to text-and-image ads is, at least for me, bad advice.
But what about my concern that since one ad block is new and one is months old, this will distort the data? Well let’s look at just the last 1/4 segment of the data – RPM for the last 7 days.
Just taking the last 7 days into account, we are in fact seeing a significant difference. Now instead of a -2% difference, it’s a positive 25% difference showing up in the data when comparing the two side-by-side.
It’s worth mentioning that overall CPM is lower for both ad units for the past 7 days because of enormous referral traffic from Reddit and other sites. This always lowers my CPM while doubling or tripling overall traffic – an overall boon for my statistics.
Adsense Experiment Conclusion
Sorry, but there isn’t one – at least not yet. The data seems to be displaying 2 different messages, which to me means we need to gather more data! So I’m going to leave this experiment running for at least another 30 days.
Soon – I’ll run the numbers on the banner ad I put up mid-june and see whether I should remove it, or test a different ad block.