The Google Instant Preview Issue Isn’t Resolved

You Won the Battle, Lost the War

We’ve been discussing Google Instant Previews (here and here) and just a few of the concerns they introduce for businesses on the web. The outcry caused when it was discovered that the Google Web Preview Agent was executing the Google Analytics Javascript and logging visits – and thereby affecting pageviews, bounce rate, etc. – has now been silenced as Google has announced that the issue has been “resolved” and that visits from the Google Web Preview Agent will not be logged in Google Analytics. Which means that either the agent is no longer executing the Google Analytics Javascript or the agent is being filtered by Google Analytics. The implications for tools like SiteCatalyst and CoreMetrics is still up in the air.

This has caused a great deal of celebration on twitter. Certainly, none of us want our data sullied, and if you need accurate pageview counts to report to advertisers, or if your site’s performance is being judged at the executive level based on bounce rate, then such a sudden change is alarming.

However, if the success of your business’s online efforts rely in any way on organic search results and your efforts utilize digital measurement for strategy and reporting, then the community demands to remove the opportunity to track Google Instant Previews is disconcerting and disappointing.

There is now no aspect of Google Instant Previews that is directly measurable, and that change was spearheaded by the analytics and digital measurement community. This is backwards. Arguing for less data around major changes by Google that impact your ability to compete on the web is short-sighted.

Google Instant Previews are functionally miniature landing pages and we need to treat them as such. This is a strategic opportunity missed, not a victory.

There’s a lot of valuable data that could come from knowing more about how searchers interact with your Google Instant Preview. What if:

  • Views of the preview of a particular page double but visits only increased by 5%?
  • Preview Views site-wide suddenly cut in half? What would you do? Did you just drop to page 2 for relevant search terms?
  • Pages on your site don’t render well in Google Instant Preview, how will you convince your Executive and Design teams to budget for redesign if you can’t demonstrate a rough size of the exposure?
  • A page that previously had negligible views of the preview suddenly started trending sharply upward? Would that signify an opportunity?
  • You’re pushing for more extensive organizational adoption of digital measurement within your business? Would the ability to measure and act on high-visibility changes in the search space bolster your case?
  • You could serve optimized content to the Google Web Preview Agent? It’s a user agent like any other. What if you could serve content for searchers that made previews more engaging, more relevant, and clearer?

Ultimately, Google Instant Previews will affect how searchers find you. What we need is better measurement, not less measurement. Could there have been a better implementation with more warning? Yes. Do some of these things belong in Webmaster Tools, not Google Analytics? Yes. But is removing the ability to measure previews in any fashion a resolution? No.

This entry was posted in experience optimization, metrics & analytics, search. Bookmark the permalink. Follow any comments here with the RSS feed for this post. Both comments and trackbacks are currently closed.


  1. Posted November 23, 2010 at 11:49 am | Permalink

    Totally agree. Not having data is hardly ever the solution to anything.

    However, having the data “in the wrong place” (thus wrecking the data we were were already collecting) was certainly not working either.

    What I like to think about is that there are probably a whole new set of metrics and techniques to measure them that could arise from the previews.

    (Tongue in cheek- maybe like the speed of the users cursor in px/s as it flashes over the magnifying glass and triggers a 40ms preview as they are racing to get to a competitors listing that is below yours)

    More seriously, I hope that Google takes some time to think about how it could make this data available.

    On another note, I found it interesting that it took almost a full two weeks before this “story” became really main stream (just in time for it to be “fixed”). Maybe the nature of our traffic was different than others, but I noticed it right off- it was 10% of my visits. I would have thought it would have been days, not weeks before there was more coverage of all this.

  2. Posted November 23, 2010 at 12:03 pm | Permalink

    My gut tells me that this reaction is symptomatic of a splintered community around digital marketing. Witness an exchange between @VaBeachKevin and @usujason yesterday about URL parameters and how differently SEO and analytics people regard them (spoiler: the upshot was “screw SEO,” which was retweeted by at least one analytics pro).

    If folks aren’t in the habit of thinking holistically about online marketing, then it’s an understandable, if disappointing, reaction to cry foul at the dirtying of their data rather than see the strategic opportunity the data could have presented.

    Great post.