Here at Thing Towers, we've been running an "accidental" experiment on the impact of Google+ profile pictures in search results. The results are pretty spectacular! That teeny 44x44 thumbnail appears to be capable of having a dramatic impact on Google traffic.

What we're talking about is the author snippet you can see in search results, when you're on Google+ and follow the steps published by Google. Here's a lovely chap who'll help you chop up your fruit:

Example Google authorship snippet

We all like graphs, so let's start with the important one:

Graph of Google visitors

Eek! That's a 135% increase in visitors from Google after changing a picture. Here's the timeline of events:

  • On 23rd of February, Google author snippets started appearing in search results for the site. The image for this Google+ profile was, frankly awful. But we didn't expect anything dramatic. It's only a tiny thumbnail, right?
  • Traffic levels plunged, so on 1st March, we deleted the profile image to see what would happen
  • Deciding enough was enough, of 12th March, we uploaded a rather splendid picture for the profile image
  • Traffic immediately starting rising to previous levels

The site in question is a hobby project, so we were prepared to abuse it for this experiment. Cause and effect is a tricky beast, and unfortunately, in order to protect the guilty, we can't reveal the actual pictures involved. However, we can confirm that:

  • There were no onsite changes to the site during the period
  • As far as we're aware, there was nothing interesting happening off-site
  • Webmaster Tools is not up to date with Google's own CTR data, which we'll share when we have it. At this point, Google reports minor decreases in impressions/average positions

Our initial conclusions are:

  • Having the verified author snippet could have a dramatic impact on your Google traffic (both positive and negative!)
  • Getting the right picture may be extremely important
  • We should run some more scientifically-sound tests of the impact of various images and opting in/out of the verified author snippets

We know this isn't the most scientific of tests, but we thought it worth sharing. Anyone else prepared to massacre some of their Google traffic in the interests of science? :)

We await your questions and comments with open arms!