ellipsis flag icon-blogicon-check icon-comments icon-email icon-error icon-facebook icon-follow-comment icon-googleicon-hamburger icon-imedia-blog icon-imediaicon-instagramicon-left-arrow icon-linked-in icon-linked icon-linkedin icon-multi-page-view icon-person icon-print icon-right-arrow icon-save icon-searchicon-share-arrow icon-single-page-view icon-tag icon-twitter icon-unfollow icon-upload icon-valid icon-video-play icon-views icon-website icon-youtubelogo-imedia-white logo-imedia logo-mediaWhite review-star thumbs_down thumbs_up

Improve the odds on your targeting guesses

Improve the odds on your targeting guesses Jamie Roche

When I studied for the SAT test in high school, I was taught to guess. I know; it struck me as funny at the time as well. The teacher stood in front of the class and told us that the most important advantage she could teach us was how to guess smarter. As much as we knew, and as much as we could pack into our heads using flash cards and practice tests, there were going to be questions to which we just wouldn't know the answer. In these cases the next best thing to knowing was knowing how to guess.

The logic goes like this: If we could eliminate one possible answer we would increase our odds of guessing correctly from 20 percent (1 in 5) to 25 percent (1 in 4). If we could eliminate two, our odds jump to 33 percent (1 in 3). Assuming four to 10 guesses in the test, this would have a substantial effect on our scores.

What if instead of eliminating options, we could fill in two ovals and get full credit if either were correct? Our chances of guessing correctly START at 40 percent (2 out of 5). If we can eliminate two options and make two guesses our chances of guessing correctly jump to better than even, 66 percent (2 out of 3). Trying two things where there is no penalty significantly increases your odds. Call it the "Rule of 2."

The "Rule of 2" applies directly to behavioral targeting. Behavioral targeting makes good sense. If we can easily add, remove or change content to make the experience more relevant to a specific population, results should improve. But, as much as we know about our user populations and their preferences, some (and usually quite a bit) of what we try is based on guessing. And as we know from our testing background, guesses, even those made by experienced marketers, are often wrong. Fortunately, in online marketing, it is easy to place two bets and improve our odds.

Rapid behavioral targeting page development
A few days ago we built a custom landing page for Offermatica to handle traffic from visitors who demonstrate an interest in behavioral targeting. We did this to capitalize on a surge in interest coming from AOL's purchase of TACODA and OMMA's wildly oversubscribed behavioral targeting show. 

The start of our process was pretty much the same as everyone's:

  1. Agree on target metrics: We agreed that we would measure time on site for traffic originating from keywords and phrases that had to do with behavioral targeting and for visitors who clicked on behavioral targeting content on the site.

  2. Decide between creating a new landing page and using an existing page with customized content: We decided to use existing home page with custom content. 

  3. Select a primary use case: We identified two that were important: prospects researching behavioral targeting and analysts researching Offermatica's offering. Prospects were selected as use case #1.

  4. Build a wire frame: We then needed to decide what content would stay as is, what content would get altered and what fresh content needed to be created. In this case, we decided to switch in a new top image and headline, add a new main content paragraph, and switch out the "careers" content with a list of behavioral targeting articles from our blogs.

  5. Build new page: We then built the new page based on our best guesses about what would drive the most time on site. Call it V1a.

And this is where the "Rule of 2" comes in. After creating the V1a based on our assumptions, we built V1b by identifying every place where there was internal debate about what would work better and then substituting a "b" version.

For example, there was debate about the headline and about the contents of the main copy. So we had the people who thought a different version would work better create the new headline and copy.

Building the V1a page took less than a day and creating the V1b added less than one hour. Within one day, the new page with a solid competitive "b" option was launched.

Twenty-four hours later, we learned that targeting behavioral targeting content to people who demonstrated interest in behavioral targeting created a significant boost in time on site, and that the version of the body copy that said, "Conversions Rarely Happen With a Single Click" performed over 100 percent better than "Behavioral Targeting Sells More," which, of course, surprised us.

Behavioral targeting has been sparking a renaissance of personalization, the recently moribund darling of the late 1990s. And it should.

Different populations respond to different things. As I learn every time I try on the robe in a moderately priced hotel, one size does not, in fact, fit all. But behavioral targeting is a new thing for most marketers, and it does require extra work. Fortunately when you couple simple testing with the "Rule of 2," the additional work is minimal and the improved results and learning are huge.

Jamie Roche is president of Offermatica. .

Jamie Roche is President and Co-Founder of Offermatica – the leading provider of on–demand marketing services, including testing and landing page optimization, that allow marketers to maximize revenue from their online advertising spend.  Also,...

View full biography


to leave comments.