Ev­ery two days we gen­er­ate as much data as we did from the dawn of time till 2003. If you were to store all the data gen­er­ated on­line in the last two years on floppy disks, the num­ber of disks needed would reach as far as the earth to the sun (149.6 mil­lion km). But this isn’t triv­ial pur­suit, and I’m sure you are well aware by now of the sheer data vol­ume we are cre­at­ing and giv­ing away.

If all this talk of ‘big data’ sounds daunt­ing, you’re not alone. Count­less mar­keters say they don’t even know where to be­gin. All that data, fly­ing at us in hyper-speed needs to be mined un­til a few key nuggets of in­sights ap­pear.

The ques­tion is how can brands use that in­for­ma­tion in mean­ing­ful ways, with­out be­ing over­whelmed by its crush­ing weight? What are the rules around us­ing big data vs. small data? When should mar­keters be look­ing at fewer data points, rather than more? And is more nec­es­sar­ily bet­ter?

Let’s break things down. First, it’s im­por­tant to keep in mind that not all data is cre­ated equal.

This is the gold stan­dard. It in­cor­po­rates ev­ery­thing you know about your cus­tomers first­hand. This in­cludes me­dia mes­sages they’ve re­sponded to, buy­ing be­hav­iour, size, and de­mo­graph­ics.

The next best thing. Given the na­ture of their business, pub­lish­ers can sup­ply you with more data than you typ­i­cally would have ac­cess to. This in­cludes in­ter­ests, hob­bies, and pref­er­ences.

Now nec­es­sary to over­lay an emo­tional an­gle onto your con­sumer pro­files. It pro­vides ‘say vs. do’ nu­ances and high­lights how your au­di­ence con­nects with one an­other and their in­flu­encers.

You’d strug­gle to have missed the buzz around ‘big data’. But be­fore we dive into it, a def­i­ni­tion is in or­der. Big Data’s size keeps chang­ing. A cou­ple of years ago, a few dozen ter­abytes of data were con­sid­ered big. This vol­ume has now been dwarfed by petabytes of data (1 peta =1,000 ter­abytes). The ex­plo­sion in smart­phone adop­tion meant that many more data points are cre­ated in real-time. That’s the ve­loc­ity el­e­ment of big data.

If you’re just dip­ping your toes in the sea of data, then your en­try point needs to be more gran­u­lar. Web­site vis­its, top per­form­ing con­tent, im­pres­sions, and viewa­bil­ity can be just as valu­able in help­ing you de­rive in­sights. Big data doesn’t re­place smaller datasets, it adds a more uni­ver­sal and real-time layer for more so­phis­ti­cated anal­y­sis.

Here’s how you can make the most of the data you have avail­able:

Gart­ner iden­ti­fies four ap­pli­ca­tions for data.

Small data will typ­i­cally al­low you to con­duct De­scrip­tive and Di­ag­nos­tic an­a­lyt­ics. Hind­sight shows what hap­pened and why. You can then de­rive in­sights for more in­formed de­ci­sion-mak­ing.

This type of data is gath­ered af­ter run­ning a cam­paign. The re­sults are then optimised and cam­paigns run again, po­ten­tially tak­ing months to com­plete. But fast-paced is the un­der­state­ment of our gen­er­a­tion. Con­sumers are on a trend­ing sub­ject one minute jump to an­other just as quickly. Can any­one re­mem­ber how quickly #the­dress broke and then un­broke the in­ter­net? 48 hours, tops.

That’s where big data comes in. The real ben­e­fit does not de­rive merely from its size, but its ve­loc­ity. Get­ting data as it hap­pens and, more im­por­tantly, act­ing upon it rapidly, min­imises the turn­around for cam­paigns and their op­ti­mi­sa­tion. Youtube stars are the lead­ers in this do­main. They al­ways have their fin­ger on the pulse, so as soon as some­thing starts trend­ing, they pounce with con­tent turned around in as lit­tle as a day.

Be­ing able to re­spond rapidly is only the first step that big data sup­ports. Even­tu­ally, we can look be­yond just the trends, to pick out pat­terns that haven’t yet be­gun trend­ing. That’s Pre­dic­tive An­a­lyt­ics.

Us­ing Google data, Kleenex planned a cam­paign around pre-empt­ing flu out­breaks in spe­cific re­gions, based on chang­ing pat­terns in search queries. The mes­sages about tis­sues were de­liv­ered in the right places at ex­actly the right time.

The in­ter­sec­tion of peo­ple, data, and in­tel­li­gent al­go­rithms will have far-reach­ing im­pacts on me­dia ef­fi­ciency. In an om­nichan­nel world, con­sumers are flut­ter­ing from one me­dia de­vice to the other. This has in­tro­duced a twist to cre­ative sto­ry­telling,

Newspapers in English

Newspapers from Bahrain

© PressReader. All rights reserved.