When Rand Blair was 12 years old he was assigned his first research paper—a daunt­ing task for any 7th grader. A quick trip to the local library turned up his first major stum­bling block: there were so many books. So a paper about arche­ol­ogy quickly became one on horoscopes—and when that was thwarted, it was on to gyne­col­ogy. Why? Dur­ing his first trip he became so over­whelmed that, when he saw a book on astrol­ogy right by the entrance, he grabbed it and ran ver­sus car­ry­ing out some level of due dili­gence and seek­ing out the book that would have helped his paper along. On the sec­ond go he made a com­mit­ment to dig deeper and find the rel­e­vant infor­ma­tion he needed to write a proper paper, but impa­tience set in and, after see­ing a cue (the suf­fix “-ology”), he snagged a book and ran. That book? Gyne­col­ogy Insights.

It’s a funny, mem­o­rable tale that our younger selves can relate to, but it’s also some­thing Rand sees hap­pen­ing in the dig­i­tal mar­ket­ing uni­verse today.  I caught up with Rand, a prin­ci­pal con­sul­tant with Adobe Pro­fes­sional Ser­vices, and he noted that like the hur­ried, impa­tient, well-meaning 12-year-old, even sophis­ti­cated mar­keters can become over­whelmed when faced with the vast amounts of data avail­able today. It should be a good prob­lem to have. We have so much infor­ma­tion because of a seem­ingly end­less sup­ply of sources, count­less plat­forms, and real-time met­rics com­ing at us minute by minute. But, in real­ity, it’s a lot like that library, where find­ing exactly what we need to write that arche­ol­ogy paper—or carry out data-driven decisioning—can be extremely dif­fi­cult with­out some level of cura­tion and dis­cern. And even when you have the back­ground, the resources, and the knowl­edge to drill down to that level, see­ing that much infor­ma­tion at your fin­ger­tips can make you second-guess your­self and your moti­va­tions, to the point that a book on gynecology—or, in the case of the mar­keter, super­flu­ous and poten­tially mis­lead­ing directives—winds up dri­ving next steps. The end result can be wrong turns, mis­steps, inef­fi­cien­cies, and, pos­si­bly, a paper on astro­log­i­cal signs and repro­duc­tive systems.

You may laugh and think this is funny … but what’s inter­est­ing about this story is that many mar­keters are using this exact same approach to opti­mize their web­site,” explained Rand. “They’re sur­rounded by data, they’re sur­rounded by books. But they’re grab­bing the test ideas that are eas­i­est, they’re grab­bing the test ideas that are clos­est to the door … they’re shoot­ing from the hip because there’s so much urgency to get that 10-page paper done.”

So how can data be used in a way that inspires trust and con­fi­dence and dri­ves mean­ing­ful, data-driven deci­sion­ing, test­ing, and opti­miza­tion? This is a ques­tion that has frus­trated many of Rand’s clients. To that end, dur­ing Adobe Sum­mit 2014, Rand, in col­lab­o­ra­tion with the Fox Net­works Group, gave a run­down of the three ques­tions to ask that, when fleshed out, can help steer mar­keters to where they need to be look­ing. Likely, it’s not clos­est to the door.

  • Where are the 800-pound goril­las? Where can you dive deeper? Per­haps entry rate or other refer­ral sources? Focus may be on the wrong ques­tions or less mean­ing­ful data—page views or lower per­form­ing channels—which seem like nat­ural heavy hit­ters, but may not be. “Opti­mize those tem­plates and you win,” explained Rand.
  • Which pages are cash cows? What are the vol­ume dri­vers on your site that will have an imme­di­ate, mea­sur­able impact? With lim­ited time, these should be the core focus points, at least at first. It’s the Pareto Prin­ci­pal at play—80 per­cent of the wealth comes from 20 per­cent of the pop­u­la­tion. That 20 per­cent applies to your site, most likely. And those pages are your cash cows. Rank by rev­enue and dive in. Just because you envi­sioned a cer­tain page mov­ing the nee­dle doesn’t mean it actu­ally does.
  • Do I have any low per­form­ers? Sim­ply, what page or pages aren’t pulling their weight? Maybe they’re high-volume per­form­ers, but they just don’t get it done. They could be high poten­tial pages with bounce rate problems—which sug­gests “holes in your bucket.” Drill down to more cal­cu­lated met­rics such as rev­enue per page view (ver­sus just rev­enue) or exit rate. Chances are, those are good indi­ca­tors of gap­ing holes or, sim­ply, they show what’s hap­pen­ing big pic­ture. From here, pulling this analy­sis to entire cat­e­gories or chan­nels can effec­tively steer areas of focus.

What do his clients need to do after ask­ing them­selves these ques­tions, accord­ing to Rand?

  1. Use the data to drive insights.
  2. Use those insights to drive hypotheses.
  3. Use the hypothe­ses to drive iter­a­tive tests with advanced segmentation.

Accord­ing to Rand, dig­i­tal mar­keters have long wanted the abil­ity to make their ana­lyt­ics data action­able, and to opti­mize with­out hav­ing to antic­i­pate all the suc­cess met­rics and audi­ences they want to report on before run­ning tests. They want to dig into test results, fur­ther seg­ment to find that gold, and ana­lyze suc­cess ret­ro­spec­tively. To this end, Adobe Tar­get users have started lever­ag­ing Adobe Ana­lyt­ics as their report­ing source. Be sure to check out this video for more info on the ben­e­fits of analytics-enhanced reporting.

We know where to go to get what we need, just like the young Rand. But once we’re there, star­ing down all that data, how do we know if our results, although accu­rate, are actu­ally mean­ing­ful to our busi­ness? With­out a strong sense of what moves the nee­dle and what’s clut­ter, mar­keters can find them­selves in a sim­i­lar distress-laden sit­u­a­tion, grasp­ing for what­ever book—or metric—is within arm’s reach.

Con­tinue the conversation—or ask more about being an analytics-driven orga­ni­za­tion— with @randblair on Twitter.