February 7, 2012
Last October, I wrote about the importance of running targeted tests to discover what experiences resonate most with your key audience segments. The goal? To deliver the right shopping experience (consisting of information, design, flow, and other elements) to each visitor—every time.
I still believe this, but what I've learned in conversations since is that many website optimization specialists disagree. Advocating a different school of thought, this group promotes un-targeted testing with the goal of finding the One Perfect Page—a single end state characterized by maximum page performance—beyond which incremental improvements are impossible. And although this group might have Aristotle on its side, the myth of the perfect page is ultimately rooted in the mistaken belief that any one page can be everything for everyone.
So let's take another look at why targeted testing is the best approach to website optimization.
Online marketing experts Bryan Eisenberg, Jeffrey Eisenberg, and Lisa Davis cite four personality types in discussing the factors that motivate visitors to make purchases. Each personality type is represented by a question the marketer must answer, along with specific tactics for doing so:
The key takeaway is that no matter how specialized or niche your business is, shoppers still exhibit large differences in terms of what advances, stifles, or ultimately reverses their likelihood to purchase. Said another way, website optimization isn't about understanding how the elements on a page interact with each other, it's about how visitors interact with your page. As a result, to focus too narrowly on discovering the One Perfect Page is ultimately to obsess over the wrong things, and miss the point that you don't have one perfect customer.
The problem (so the critics say) is that no one arrives on your site wearing a sign that says, "Humanist." However, you probably know much more than you think. For example, whether a visitor came from organic or paid search, the query he or she typed, and the search result/ad clicked can reveal a great deal about that visitor's likelihood to purchase as well as the information most likely to lead him or her to do so. Similarly, whether a visitor is new vs. returning or a first-time vs. habitual buyer, all help reveal the visitor's needs, desires, and even fears.
For example, all things equal, visitors referred by Yelp.com may be more interested in product ratings and reviews (and those of your company) than visitors referred by Ebates, who are likely shopping for the best deal possible. It's the marketer's task to act on these data points to deliver the most relevant website experience possible—and with tools that deliver on the promise of one-tag/one-time integration, doing so has never been easier.
Thus, we come full circle. The true problem with the myth of the perfect page is not that it once had merit. (Sorry guys, it never did.) It's that it was the product of an era in which targeted tests were difficult to execute. Each test—in fact, every element of each test—required new and different code to run, making the idea of multiple experiences impractical. The one perfect page, therefore, always was an imperfect solution driven by the technological possibilities (or rather, limitations) of the era in which the first-generation testing solutions were born.
As the technology that supports website optimization continues to evolve, approaches to testing and targeting must evolve with them. When, instead, you're always prepared to fight the last war, you'll always be defeated by more agile and savvy competitors.
References: Davis, L., Eisenberg, B., & Eisenberg, J. (2006). Call to Action: Secret Formulas to Improve Online Results, 52-60.