“This is a com­pli­cated ecosys­tem”

In this chat, Garza speaks about how it takes more than tra­di­tional mar­ket­ing to level up in the dig­i­tal game.

The Brand Reporter - - COVERSTORY - By Abid Hus­sain Bar­laskar abid.burlaskar@afaqs.comww

Amidst all of the bus­tle and con­fu­sion around pro­gram­matic ad­ver­tis­ing in the in­dus­try, we caught up with Os­car Garza, global head of me­dia ac­ti­va­tion for Essence, GroupM’s data and mea­sure­ment-driven arm, to un­der­stand what mar­keters in In­dia need to be wary of be­fore step­ping up their dig­i­tal game. “AI does what hu­mans can’t on such a large scale and at a high pace. Pri­mary uses in ad­ver­tis­ing is de­ter­min­ing the im­pact of me­dia ex­po­sure. Essence cur­rently in­gests over four ter­abytes of data a day. That’s the big­gest op­por­tu­nity in ad­ver­tis­ing. We can use it for ad­just­ment in tar­get­ing or to de­ter­mine the im­pact on the out­come,” Garza says. Ex­cerpts:

Garza points out that tech­no­log­i­cally, there’s noth­ing that’s hold­ing In­dia back from par­tic­i­pa­tion in AI/ML in ad­ver­tis­ing, but then it’s largely a tra­di­tional mar­ket. “Ad­ver­tis­ers spend a lot of money on tele­vi­sion and the gran­u­lar in­for­ma­tion re­quired for ma­chine learn­ing doesn’t ex­ist for tra­di­tional chan­nels.

Garza then opens up about the way mar­keters need to equip them­selves for a bet­ter un­der­stand­ing of dig­i­tal ad­ver­tis­ing, to deal with ad fraud and to match the ROI game. “Most peo­ple we en­counter in mar­ket­ing po­si­tions have come out of busi­ness schools and have an MBA and that’s fan­tas­tic. But for the world of dig­i­tal mar­ket­ing, ar­ti­fi­cial in­tel­li­gence - ma­chine learn­ing, you also have to un­der­stand com­puter science to some de­gree, ad­vanced sta­tis­ti­cal meth­ods for un­der­stand­ing the math, telecom­mu­ni­ca­tions, data­base man­age­ment, and un­der­stand­ing how the in­ter­net works,” he ex­plains.

“For de­ter­min­ing the ROI, it’s not suf­fi­cient to put a track­ing pixel on a web­site since un­der­stand­ing the in­ter­ac­tion with the me­dia that led to a sale is a com­plex process. It’s im­por­tant to un­der­stand how an im­pres­sion flows throw mul­ti­ple sys­tems and back to the ad­ver­tiser. If one can as­so­ci­ate a sale with an im­pres­sion or the se­ries of im­pres­sions that led to it, it cre­ates the op­por­tu­nity of a sta­tis­ti­cal anal­y­sis. If ad­ver­tis­ers don’t un­der­stand all of this, they are go­ing to fall vic­tim to fraud,” Garza adds.

What about pro­gram­matic in-hous­ing and the mix to build­ing a dig­i­tal me­dia plan­ning and buy­ing scene in-house?

“Agen­cies should be sell­ing their ser­vice as ad­vice; in our case, it is me­dia buy­ing, which might not be a field of ex­per­tise for a brand and it needs some help with it. The brand, be­fore in-hous­ing, has to un­der­stand the full range of ser­vices that it has to pro­vide for it­self. One of the big­gest mis­takes I have done in the past is fir­ing agen­cies and go­ing in alone. For now, I would sug­gest a hy­brid re­la­tion­ship be­tween the brand and the agency,” Garza says.

“I had been buy­ing pro­gram­matic for nine years. I took the job here to run pro­gram­matic and it’s the first job that has ‘pro­gram­matic’ in the ti­tle. Now, we recog­nise it, it is pop­u­lar enough and a sig­nif­i­cant part of the in­dus­try.”

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.