Hitmetrix - User behavior analytics & recording

Landing pages gone wrong

Forget the hustle and bustle of downtown New York; the endless blanket of buildings outside of the city center is what impresses me most. In most cities, a taxi ride to the airport or a train trip to its final destination reveals brick complexes that house thousands of tenants, as well as a billboard or two. My first thought is frequently, “Who are all of these people, these consumers?” This thought is so overwhelming that I shift to thought number two, usually about the billboard advertising. “Is this really targeted to me, the gal on the way to the airport?” Needless to say, I sometimes feel like I’ve tripped across advertising aimed at someone else. If only there were a big back-button on the Long Island Expressway.

There’s only one thing worse than looking at a poorly targeted ad, and that is clicking through to a landing page that just doesn’t jibe with the ad. Despite all the advances in ad targeting and landing-page optimization, not every marketer employs the tools available. For those who deal in this trade, it can be quite amusing to deduce the strategy behind ad copy or a landing page. So when a colleague asked me to review a campaign for an extremely competitive market, I willingly took her up on the challenge.

Now, if the ads were for toothpaste or luxury handbags, this would be amusing, at most. But these ads aimed to sway a higher involvement decision: to support or not support John McCain in the 2008 Presidential election. One quick look at the landing page’s URL, which ended in .com/landing, and I knew I was in for some serious sleuthing.

A Google search for the “McCain landing page” revealed another landing page, with the ending .com/landing/b3/htm. Now this was getting fun. If there were a “b3” landing page, clearly there would be others. And indeed there were. Live landing pages labeled a1, a2, a4 and b4 exposed four very different messages. 

I turned to Jonathan Mendez of RAMP Digital to answer a few burning questions about John McCain’s landing-page strategy and, specifically, if such blunt URL structures were a normal practice. “This is quite normal and common with A/B testing,” said Mendez, “but in a hyper-competitive election, it’s just lazy and unacceptable.”

So what should McCain’s team have done?

“What’s almost as amazing to me is that they are doing basic landing page A/B testing at this point and not using one of the advanced testing and optimization tools,” commented Mendez. So while we gave McCain kudos for the tactic, there were demerit points for not staying up on technology — a popular meme with the politics and technology crowd. According to Mendez, the site’s source code points to an older version of HBX. “At this point of the election cycle, you would expect them to be doing multivariate testing and/or targeted content delivery,” assessed Mendez.

Mendez did offer this advice to marketers embarking on landing-page testing: Even if basic testing is chosen, parameters are totally open and URL strings are easily disguised, so that alternate test pages are not so obvious. On the other hand, he points out that most testing scenarios don’t need to mask the test URLs. “If you A/B test with Javascript, you only have a single URL.”

Related Posts