• http://www.grokdotcom.com/2007/04/03/grokcast-googles-tom-leung-on-the-google-website-optimizer/ Jeffrey Eisenberg

    It is not cloaking. You can hear from Tom Leung, the Google product manager himself in his interview with Bryan Eisenberg at http://www.grokdotcom.com/2007/04/03/grokcast-googles-tom-leung-on-the-google-website-optimizer/

  • http://www.traffick.com AndrewGoodman

    Call it cloaking or not, the system has enough flexibility in it that it makes certain deceptive tricks possible, to be sure. I’m not sure most of those tricks will rank you higher in search engines, however. Whatever got the “original page” to rank high *should* allow it to continue to rank high while you test new versions.

  • http://www.grokdotcom.com/googlewebsiteoptimizer/ Jeffrey Eisenberg

    Andrew, Tom Leung explains how it works in the interview. I’d be curious on how it might be gamed in the future. BTW, we’ve also published 7 Freebies to go along with Google Website Optimizer. You can get them here http://www.grokdotcom.com/googlewebsiteoptimizer/

  • http://searchengineland.com Danny Sullivan

    So I listened to the first part where cloaking, if it was to come up, should have. Jeffrey, please tell me if it was beyond the first five minutes. That’s as far as I listened fully. I then skipped through the rest and feel fairly confident I didn’t miss a discussion of rankings elsewhere.

    In the part where rankings were covered, nothing was said about cloaking. Nothing. Tom responded instead generally to questions about ranking:

    “Using the website optimizer in and of itself will not effect your Google search engine ranking. The original content will still be there and still be indexed as if you weren’t testing. However, if you use the test and find ways to make for a better customer experience and then you implement those changes, then you should have every reason to believe that will be reflected in a positive way for your rankings. but using the tool in and of itself won’t change your rankings one way or another.”

    OK — and with respect to Tom — he absolutely cannot guarantee your rankings will not be impacted up or down. In fact, the “every reason to believe” statement that making the page more usable according to the tool should improve rankings is probably an overstatement. Sure, search engines tried to model ranking algorithms to favor pages that people like. But that’s not guaranteed, and that type of statement makes it sound like the tool also has some magic Google rank improvement tips to offer.

    In addition, Tom works for Google but not for the search quality team. Even if he did say it is not cloaking, unless he specifically says the search quality teams has said it isn’t cloaking, the jury remains out. It’s that team that decides.

    I like what the new tool offers. Heck, I like Tom. I also like Offermatica. And I’m not trying to scare people away from these tools. But they’ve been sitting on the gray area about whether they are cloaking. With Google now offering its own tool to everyone, the company needs to clarify the cloaking situation right now. Otherwise, you’re going to see many people take these tools as tactic approval that many types of content swapping can be allowed.

  • http://www.grokdotcom.com/googlewebsiteoptimizer/ Jeffrey Eisenberg

    Danny, I listened to it again and I’m sorry. You are correct. The interview is a few days old and I thought it had more detail.

  • http://www.mattcutts.com/blog/ Matt Cutts

    Danny, I met with the WSO team in advance to make sure that using WSO wouldn’t cause issues; it’s clearly helpful for people to be able to do UI tests with things like WSO or Offermatica.

    I’ve had a “best practices for doing UI testing” post rolling around in my brain for a while, and I’ll try to work with a few people here to get that out. But in the mean time, I wouldn’t consider using these tools a problem at all.

  • http://SiteTuners.com Tim Ash – SiteTuners

    There is a fundamental issue underlying the cloaking discussion: what spiders need to get your page ranked well is not what visitors want to experience when they get to the page. There will always be this inherent tension.

    Google prefers content-rich static-text pages of a particular word count range. Anyone who does not write pages to this editorial style will suffer.

    Unfortunately testing has shown repeatedly that people don’t “read” when they look at web pages. So the banner of “what’s best for the visitor” that Google trots out frequently to justify its policies just does not fly. For many landing pages that we have tested, a stripped-down stark version of the page works best. All of the clutter and detailed information is often removed so that the user can focus on their task with minimal distractions.

    Google’s Tom Leung is correct when he says that testing should not affect your page rank. This is because most testing technology uses client-side JavaScript (or something similar) that the spider never sees. But that is not the point. The fear among website designers is that if a competitor tattles on you, any content that is seen by during landing page optimization tests by users but is not read by the spider will be considered cloaking.

    As a Google Website Optimizer Authorized Consultant http://SiteTuners.com has had discussions with Google about this. The line in the sand is still shifting. Until we hear something definitive, our policy is to leave the majority of the text on the page intact so that the “theme” of it is preserved.

    We have developed a number of specific techniques to change the overall appearance and improve the openness of the design. These techniques still leave the text on the page while changing its visual emphasis for the end user. This gives our clients a defense against cloaking allegations since the spider and visitor both see essentially the same text content.