Posted April 4, 2007 9:40 am by with 7 comments

Tweet about this on TwitterShare on LinkedInShare on Google+Share on FacebookBuffer this page

I’d really wanted to spend more time looking at the full launch of Google Optimizer (announced last October), but sadly have been tied-up with clients this week.

Instead, ClickZ does the heavy lifting and provides a great overview of Google’s full deployment of the multi- variate/variant/variable (who knows what the correct useage is) product that allows web site owners to test the effectiveness of various page items.

“The main problem we’re trying to solve is to get people out of the dark ages in terms of how they develop pages,” Tom Leung, Google’s product manager for Website Optimizer, told ClickZ News. “All too often, they’ll just put a page together and maybe the designer will do a few mock-ups, and they’ll point to the one they feel is going to be the best one.” Leung says the goal is to help marketers convert their Web sites into a “living laboratory” and expressed the hope it will improve the usability of the Internet overall. Sites will, theoretically, provide more of what online consumers are looking for.

Just five companies have been named trusted consultants for the product, including our good friends (and sponsors) ROI Revolution.

  • We are another of the partners. Could we let your readers know that we are offering seven free resources to support Google Website Optimizer at ?

  • Dean

    We have employed Google Website Optimizer and found one huge limitation. In order to get statistically significant results from GWO you need ALOT of traffic to the test pages. As more variations are introduced that number grows exponetially.

    GWO is great in theory, but in practice only useful for high traffic pages.

  • Jeffrey, can I invoice you for that link? 😉

  • Andy,

    Sure thing. We can do a CPA deal. You can have 50% from all the sales of free resources 😉 I wanted to answer Dean’s point. The way Google Website Optimizer works is that each variation you create needs a minimum of 5 conversion and 5 non-conversions. The more variations you have and the poorer you convert the more traffic you need. Then of course one variation needs to be significantly better than the rest to show any confidence levels. My suggestions is to test fewer sections and with only a handful of variations at a time.

  • Dean, That’s true. Or else you need a longer time to gather the data for a test. It’s not a limitation of the tool but of testing in general.

  • Dean

    Agreed that it is a testing limitation not a tool limitation….but a limitation none the less.

    “Yes I can tell you which page will work better…I’ll get back to you in 3 months” 😛

  • The official Google Website Optimizer group just changed status from announcement only to general discussion.

    If you want to ask a question or share your experiences, it’s a great place to start.