In part one of this blog series, we talked about the impor­tance of data but also the chal­lenges of hav­ing so much of it. All this raises a good ques­tion: If a volume-based approach toward data and ana­lyt­ics is inap­pro­pri­ate, then why do we keep pro­duc­ing pro­grams that accu­mu­late more and more data?


You’ve def­i­nitely heard the words Big Data thrown about, and you’ve likely heard peo­ple say we’ve now entered the era of Big Data. Dra­matic pro­nounce­ments aside, those state­ments are right. We now accu­mu­late a lot more user infor­ma­tion than ever before, and we now accu­mu­late far more user infor­ma­tion than any one per­son could hope to keep up with. If you’re a mar­keter and you try to per­son­ally han­dle all this data on your own, you will quickly short out your ratio­nal brain and end up mak­ing your deci­sions by resort­ing to your own fal­li­ble instincts.

Talk about coun­ter­pro­duc­tive! We brought ana­lyt­ics to mar­ket­ing in the first place so mar­keters could go beyond their ambigu­ous instincts and so they could use hard data to reach pre­dictably suc­cess­ful decisions.

All this makes the Big Data story sound like a bit of a bind.

  • On the one hand, dig­i­tal mar­keters need to use data to make their decisions.
  • On the other hand, dig­i­tal mar­keters need to use an ever-increasing vol­ume of data with­out let­ting that data over­whelm them.

The way out of this poten­tial pit­fall is actu­ally pretty sim­ple: mar­keters need to spend more time exe­cut­ing on all that data and less time col­lect­ing and sim­ply visu­al­iz­ing it.

Exe­cu­tion Over Pure Accumulation

I’d love to put mar­keters who just accu­mu­late unex­e­cuted data on their own spe­cial episode of Hoard­ers. When it comes down to it, build­ing up a moun­tain of untouched data is the IT equiv­a­lent of stuff­ing every last cor­ner of your home with tow­er­ing stacks of unread newspapers—there’s a lot of infor­ma­tion sit­ting there, but it’s not doing you any good!

If you don’t exe­cute on data, then that data is worth­less. When you actu­ally put your data to work, you’re going to reap the ben­e­fits of Big Data, as our clients have proven time and time again.

Lead­ing finan­cial ser­vices firm Citi used Adobe Insight to reduce delin­quent pay­ments by 15 per­cent through tar­get­ing at-risk cus­tomers with spe­cial mar­ket­ing pro­mo­tions push­ing auto­matic pay­ment reminders. Citi also used Insight to reduce cus­tomer ser­vice calls by 20 per­cent through iden­ti­fy­ing and cor­rect­ing web­site aban­don­ment points that lead to unnec­es­sary phone time.

Our clients have also found the more sophis­ti­cated their ana­lyt­ics pack­age, the bet­ter their results. Los Ange­les based Dol­lar Rent A Car opti­mized their Web pres­ence with Adobe Site­Cat­a­lyst and Search­Cen­ter+ (http://​www​.adobe​.com/​p​r​o​d​u​c​t​s​/​s​e​a​r​c​h​c​e​n​t​e​r​.​h​tml) before adopt­ing Insight and Adobe Test & Tar­get to uti­lize their offline data just as effec­tively. Imple­ment­ing this expanded ana­lyt­ics pack­age, Dol­lar Rent A Car pro­duced a 45 perecnt ROI as they increased pro­duc­tiv­ity, reduced ser­vice costs, improved cus­tomer seg­men­ta­tion and per­son­al­iza­tion, and improved pro­mo­tional effectiveness.

The Data Keeps Get­ting Bigger

If you’re going to cre­ate effec­tive mar­ket­ing cam­paigns, you need to do more than just mea­sure and visu­al­ize large quan­ti­ties of user data in order to put up the veneer of help­ful­ness. You need to use an ana­lyt­ics solu­tion that sifts through these ever-expanding met­rics and offers highly tar­geted rec­om­men­da­tions based on its evaluations.

And make no mis­take, the vol­ume of user data being sucked into these ana­lyt­ics solu­tions is grow­ing at an expo­nen­tial rate. If you think it’s hard to keep track of this data on your own right now just, imag­ine where we’ll be in five years.

Over­all, this era of Big Data (and soon to be era of Even Big­ger Data) is a good thing. All this data not only lets mar­keters under­stand their exist­ing users bet­ter than ever, it lets mar­keters per­form pre­vi­ously unheard of feats of pre­dic­tive mar­ket­ing. With the vol­ume of data we pull in and with a sophis­ti­cated solu­tion, you can not only mea­sure what your users do online, you can also dis­cover who they are offline and even what they’re going to do next.

And that’s big.

Mea­sure Every­thing, Look at Next To Nothing

The data required to accu­rately define and pre­dict indi­vid­u­als and their future behav­ior is more than just big … it’s mon­strous.

While there are a lot of bright peo­ple work­ing in the world of dig­i­tal mar­ket­ing, even the bright­est brain can only jug­gle so much data at any given moment. For­get about mon­strous vol­umes of data; the human brain can only han­dle a rel­a­tively small amount of data before it shorts out and loses its ratio­nal decision-making capabilities.

But unlike even the bright­est dig­i­tal mar­keter around, the right solu­tion can han­dle an infi­nite vol­ume of data.

As I’ve said before, machines are just plain bet­ter than humans at sift­ing through a huge amount of data to find rel­e­vance within the mess of met­rics laid out before them. That’s why we cre­ated these machines in the first place, and that’s why we keep mak­ing these machines smarter and more spe­cial­ized to han­dle the sorts of ana­lyt­ics we’re cre­at­ing for the future. After all, as the vol­ume of data we can mea­sure accel­er­ates expo­nen­tially over the com­ing years, the (already poor) human abil­ity to han­dle this data on its own will con­tinue to dimin­ish, and an effec­tive marketer’s reliance on data-sifting pro­grams will increase equally dramatically.

In other words, although smart mar­keters under­stand the impor­tance of uti­liz­ing all the data they can get their hands on, they actu­ally look at very lit­tle of it. The smartest mar­keters pass along as much of this data-sifting, pattern-finding, and rel­e­vance dis­cov­ery to machines as pos­si­ble and only look at a few out­puts. It’s a lot eas­ier to look at and make deci­sions from—a half dozen out­puts ver­sus a tril­lion inputs.

This means the smartest mar­keter in the room isn’t the mar­keter who looks at the most data. In a lot of ways, the smartest mar­keter in the room is the mar­keter who looks at the least data.