Testing for the Masses

Yesterday, Google quietly unveiled its newest product. It doesn’t have the glamour of last week’s billion dollar purchase of YouTube. In fact it has nothing do with video, recent acquisitions, or Web 2.0. Google’s new product, dubbed Website Optimizer, might sound more like the prevalent pop-under carrying a similar name, but unlike that ad, this one has the potential to not only help Google advertisers but Google make more money too. The soft launch, in beta and accepting applicants, came not from news story or press release but from a posting on Google’s Inside AdWords blog. Google’s Website Optimizer doesn’t make a page load faster; it’s not that type of optimizer. It’s better.

Google Website Optimizer is perhaps the most significant product release since Google decided to throw the web analytics industry into chaos by not just purchasing upstart Urchin but making the company’s web tracking tools available for free. Where Urchin helps tell about the people and behavior of those people to one’s site, the Website Optimizer has the potential to make the most of those visiting. It helps advertisers create different versions of their page to test and pick the one that yields the greatest lift in conversion. And again, similar to Urchin, Google will offer the service for free, where today with other companies it costs money.

Here are some selected sentences from the Inside AdWords blog: “Looking to increase the number of conversions you’re getting on your site?” “We know that this can be difficult to test and we want to help you out.” “…we’ll be testing a new tool called the Website Optimizer that can help you find out which content will convert best on your site. Whether you define a conversion as a purchase or a newsletter sign-up, Website Optimizer allows you to experiment with different headlines, copy, and images on your site in order to find out which combination results in the most conversions. You can use this tool on your landing page or any page that represents a conversion.

At the end of each experiment, graphical reports show which version of your landing page users liked best, as measured by which variation had the highest conversion rate. So, if you’re interested in increasing conversions, we think you’ll find the Website Optimizer useful.” The post contains an interesting point regarding the much maligned Quality Score. It says, “Using Website Optimizer to experiment with your landing page does not have any impact on your Quality Score, so long as you maintain the existing default landing page for the Ad Group. Once you make a change to your site based on the results, however, the Quality Score might change as with any other changes to your landing page. That said, if a change is good for your users, it is probably good for your Quality Score too.”

The last sentence, “…if a change is good for your users, it is probably good for your Quality Score too” deserves another look. The Website Optimizer helps increase conversion rates. As any experienced direct marketer will tell you, especially one with online lead generation experience, what helps to increase conversion rates often seems to run at odds with the Quality Score. Anyone at a network has probably come across this scenario. A client will want to run a campaign; they show their landing page, but it has no chance of succeeding. It has too much content, too many links, overall too many ways for the user to get distracted.

Successful landing pages contain less content and navigation for a reason. It converts better, and the quality of the conversion does not decrease as a result of narrowing the scope and mass on the page. The conversion rate versus Quality Score is a huge issue that has caused a lot more work, and from my discussions with those in the field, reduced budgets as a result. It will be interesting to see what happens as the Google base (not that Google Base) starts to play with their conversion rate. It has the potential though to be a win-win. With the tool advertisers will become more effective, and as a result, be able to spend more on Google.

It cannot be stressed enough though, just how important this rollout could be. Landing page optimization is one of the key pieces to advertising success, and this applies especially for search. A/B testing can accomplish a lot, but many in our space have advanced beyond that to do multi-variate testing, e.g. testing three headlines, two pictures, and two sub headlines without having to create the twelve-plus separate pages that such a test would normally require. All of the sudden, just about anyone can leverage something that once belonged only to a few. If employed properly, it’s easy to think that Website Optimizer can become Google Revenue Optimizer and lift revenues easily 5%. That might not sound like much, but 5% of several billion dollars is a very nice sum.