Robert_Charlton - 10:21 pm on Sep 5, 2012 (gmt 0)

Mod's note: I'm not mentioning one piece of key vocabulary searched here, in order not to skew search results more than we have to. Chances are we may skew Google's behavioral test as it is.

Exploring further... I could reproduce the same four results shown in the first screen capture, for the k***** word search, by searching [manufacturing] first, checking in Wikipedia as if looking for the k word, backing out and then searching for the k-word. I got the same four results shown in the screen shot, but I also got various Universal results... news etc... thrown in. Still, just four organic results.

The additional 17,700,000 results do show on subsequent pages, so in that sense Google is preserving engagement with the bulk of its serp real estate, and I perhaps misspoke regarding that.

Also, to note, I could not reproduce these results in situations where I was signed in order to retain a cookie to turn off Search History. I could reproduce them only where I wasn't signed in to a Google account and (paradoxically) search history was not disabled. I could see in the url that Google was most definitely tracking some sort of history... and I assume the cookies were also being used as long as I didn't exit the browser.

The url appears to carry a long and convoluted history, and abbreviations for certain keywords suggest that the tracking was pre-planned... I don't think they make up those abbreviations on the fly.

This not-signed-in browser is set to flush Google cookies when it closes, though, and I haven't gone to the lengths of changing those settings to see how long-lived the history effect is. I leave that to others to try and report back. It would be an interesting test about whether it extends beyond the immediate session. Ditto for the subsequent search shown in screen capture two.

Regarding the Adwords at the bottom... all searches for the k-word are currently displaying Adwords at the bottom, so this may be a constant factor in this test. Chances are Google wants to compare apples with apples as much as it can, and keeping Adwords out of the picture as much as possible probably clarifies user response to organic... and also probably later gives Google some extra data points.