Can Video ad fraud de­rail pro­gram­matic?

New Zealand Marketing - - Front Page -

“Au­to­mated buy­ing in New Zealand is ad­vanc­ing rapidly, un­der­pinned by the de­sire of brand ad­ver­tis­ers to drive trans­parency and ef­fi­ciency in their ad spend,” said the com­pany’s Australia and New Zealand man­ag­ing direc­tor Sam Smith in a re­lease at the time.

The prob­lem, how­ever, is that pro­gram­matic is of­ten priced in terms of cost per thou­sand views (CPM). And this ob­vi­ously isn’t ideal when statis­tics sug­gest that as much as 25 per­cent of all on­line video views are reg­is­tered fraud­u­lently through bot ac­tiv­ity.

As il­lus­trated by the grow­ing de­mand in the pro­gram­matic mar­ket, this hasn’t

In the case of tele­vi­sion, Nielsen pro­vides data on the per­for­mance of shows; in ra­dio, TNS has un­til re­cently con­ducted a twice-yearly sur­vey on the num­ber of lis­ten­ers that tune in; and the mag­a­zine in­dus­try con­tin­ues to rely on Nielsen and ABC fig­ures for read­er­ship and cir­cu­la­tion, re­spec­tively.

In each of th­ese in­stances, me­dia own­ers sell their ad slots based on in­for­ma­tion that’s pro­vided by the in­de­pen­dent third party. How­ever, as things stand at the mo­ment, the dig­i­tal in­dus­try does not have a pre­ferred sup­plier for on­line rat­ings, mak­ing it chal­leng­ing for yet stopped ad­ver­tis­ers from spend­ing ad dol­lars on pro­gram­matic ad-buy­ing.

As OMD’s An­drew Rein­holds points out: “Ul­ti­mately pro­gram­matic for OMD de­liv­ers such a cost-ef­fec­tive au­di­ence for our clients that even say if 25 per­cent of the clicks were fraud­u­lent, the way we track and op­ti­mise, then the re­main­ing 75 per­cent of the ac­tual clicks are of real value for us and the CPA [cost per ac­tion] de­liv­ered will be well be­low other types of buys.”

And pro­vided that pro­gram­matic play­ers can limit the ex­tent of ad fraud and keep this per­cep­tion of adding value to the client’s of­fer­ing, then it’s likely that the mar­ket will con­tinue to grow. ad­ver­tis­ers to de­ter­mine what the most ac­cu­rate data is.

To ad­dress this prob­lem, the IABNZ last Au­gust re­leased a doc­u­ment that com­pares the dif­fer­ent mea­sure­ment ven­dors avail­able in New Zealand.

“No cur­rent au­di­ence mea­sure­ment sys­tem is per­fect – each has its own strengths and weak­nesses so the Mea­sure­ment Coun­cil helps progress im­prove­ments across multi chan­nel plat­forms and keep New Zealand at the fore­front of best prac­tice,” says the on­line pub­li­ca­tion.

Nielsen, Coms­core, Ex­pe­rian and Ef­fec­tive Mea­sure are all pit­ted against each other in terms of method­ol­ogy, met­rics, pos­i­tives and user in­ter­face, among other things. The IABNZ is not hedg­ing its bets on any spe­cific ven­dor, and in­stead says that its “ap­proach is to pro­vide the in­dus­try with ob­jec­tive in­for­ma­tion on the ex­ist­ing method­olo­gies.”

In do­ing so, it places the onus on ad­ver­tis­ers, me­dia agen­cies and pub­lish­ers to in­cor­po­rate their own mea­sures to en­sure that their re­port­ing is ac­cu­rate.

A me­dia agency yoke

Given me­dia agen­cies’ role in buy­ing the most suit­able slots for their clients, it fol­lows that they are prin­ci­pally re­spon­si­ble for safe­guard­ing against fraud­u­lent ac­tiv­i­ties that may ob­scure the truth be­hind a cam­paign’s per­for­mance. For this rea­son, me­dia agen­cies across New Zealand have con­sol­i­dated their on­line cam­paign track­ing ca­pa­bil­i­ties and skills.

“Bot traf­fic im­pacts ev­ery sin­gle cam­paign on­line,” ad­mits Nick Boul­stridge, the head of dig­i­tal me­dia at Vi­vaKi. And he says his com­pany has re­sponded though Vi­vaKi Ver­i­fied, a pol­icy that sees Vi­vaKi eval­u­ate and score web ad­ver­tis­ing plat­forms in terms of how they mon­i­tor and fil­ter out bot traf­fic and other fraud, as well as their process for vet­ting sites for brand safety and qual­ity ex­e­cu­tions.

“We’ve done a lot of testing on this and found that our whitelist within AoD, our pro­gram­matic plat­form, does a very good job of keep­ing us off this sort of in­ven­tory,” he says.

Sim­i­larly, OMD’s man­ag­ing direc­tor An­drew Rein­holds says that his agency has also in­tro­duced a strat­egy that helps to iden­tify fraud­u­lent ac­tiv­ity.

“When us­ing pro­gram­matic buy­ing we utilise cost-per-click (CPC ) deals the ma­jor­ity of the time, which there­fore sets

Ac­cord­ing to TubeMogul’s Q4 re­sults, New Zealand’s pro­gram­matic video ad­ver­tis­ing mar­ket grew nearly three-fold the fourth quar­ter of 2014, with video ads avail­able for real-time buy­ing jump­ing 210 per­cent to 260 mil­lion auc­tions, from 90 mil­lion in the third quar­ter.

us up to a risk of be­ing charged for clicks that a ‘non-hu­man’ has gen­er­ated,” says Rein­holds. “The way we counter this is by track­ing from click-through to ses­sion gen­er­ated on the web­site and then through to the de­sired ac­tion for the cam­paign.”

Tak­ing th­ese sys­tem­atic steps al­lows OMD to spot anoma­lous trends—such as an un­char­ac­ter­is­ti­cally high bounce rate— that are con­sis­tent with bot ac­tiv­ity.

Although the ap­proach is prov­ing ef­fec­tive, Rein­holds says that the fight against ad fraud isn’t easy due to the lev­els so­phis­ti­ca­tion in­volved.

“It is al­ways go­ing to be dif­fi­cult to cat­e­gor­i­cally iden­tify the vol­ume of ad fraud,” he says. “It’s been go­ing on for years with search and Google rightly, when they iden­tify sus­pected fraud­u­lent ac­tiv­ity, re­im­burse their clients. That said, there could be so much more go­ing on that we don’t know about.”

The greyest area

One area that is prov­ing par­tic­u­larly dif­fi­cult to track is video ad­ver­tis­ing and Boul­stridge be­lieves that this is largely be­cause of the sys­tem cur­rently em­ployed.

“Video re­port­ing lags be­hind due to lim­i­ta­tions within the VAST [video ad serv­ing tem­plate] stan­dard,” he says.

Last year, in an ar­ti­cle pub­lished on AdEx­changer.com writer Ryan Joe ex­plained that if a “video was in­te­grated us­ing only the video ad-serv­ing tem­plate, then mar­keters might as well dump their video ads into a ti­ta­nium box for all the visibility they’ll get.” He went on to say that VAST video place­ments “don’t al­low third­party tech ven­dors or ad­ver­tis­ers to ap­ply viewa­bil­ity tech­nolo­gies,” mak­ing it dif­fi­cult to tell if an ad was viewed by a per­son.

One way to get around this prob­lem is by in­cor­po­rat­ing a video player ad-serv­ing in­ter­face def­i­ni­tion (VPAI D), which would sup­port deeper an­a­lyt­ics of the video ads run­ning via a plat­form.

“Un­for­tu­nately in New Zealand, VPAI D in­ven­tory is quite limited due to the dom­i­nance of YouTube (Google has re­jected the VPAI D stan­dard),” says Boul­stridge. “True viewa­bil­ity and verification is only re­ally pos­si­ble with VPAI D.”

A Google spokesper­son dis­agrees with this state­ment: “Google’s ad ex­change al­lows buy­ers to serve VPAI D ads to pub­lish­ers that opt in.”

And while this might be true, the quick look at TrueView poli­cies for ads served onto it ex­plic­itly states that “VPAI D is not al­lowed on YouTube.” And a spokesper­son from pro­gram­matic ad-buy­ing plat­form TubeMogul says: “Ex­cept in cer­tain cir­cum­stances, YouTube gen­er­ally does not ac­cept the VPAI D for­mat, which lim­its third­party re­port­ing. Gen­er­ally, [TubeMogul] sup­ports pub­lish­ers and sup­ply-side plat­forms tran­si­tion­ing from VAST to the VPAI D stan­dard as it al­lows for more ro­bust mea­sur­a­bil­ity.”

What this means is that ad­ver­tis­ers are more of­ten than not re­liant on the statis­tics pro­vided through Google’s in­ter­face and do not have re­course to in­de­pen­dent verification of statis­tics.

“Ad­ver­tis­ers have full trans­parency on the num­ber and du­ra­tion of views on YouTube, as well as other met­rics,” says the Google spokesper­son. “And us­ing TrueView, ad­ver­tis­ers on YouTube only have to pay when a viewer watches their video for 30 sec­onds, or to the end if the video is shorter.”

But Boul­stridge isn’t con­vinced by this ar­gu­ment. “It is very easy to say that TrueView has de­liv­ered and only charged for the com­pleted view, but was it hu­man? With­out third party track­ing you have to ac­cept what is pro­vided by Google.”

Time for stan­dard­i­s­a­tion?

“Hav­ing a me­dia owner re­spon­si­ble for the track­ing and mea­sure­ment of cam­paign per­for­mance is un­usual,” says Rein­holds. “By de­fault it can cre­ate a per­cep­tion of non­im­par­tial­ity and con­flict of in­ter­est. As with all other me­dia, an in­de­pen­dent third party means that th­ese per­cep­tions are re­moved and ev­ery­one is work­ing on a level play­ing field.”

This is not to say that new me­dia jug­ger­nauts such as Google and Face­book aren’t fight­ing the good fight against fraud­u­lent on­line ac­tiv­i­ties. Both or­gan­i­sa­tions have in­vested heav­ily in tech­nol­ogy that iden­ti­fies ad fraud, and it is, af­ter all, in their favour to have plat­forms that ad­ver­tis­ers trust enough to spend money on. How­ever, there is some­thing icky about giv­ing the pub­lisher ex­clu­sive right to re­port on cam­paign per­for­mance.

So does this mean that it’s time to stan­dard­ise the in­dus­try by in­cor­po­rat­ing third-party track­ing fa­cil­i­tated by an in­de­pen­dent third party?

Sam Smith, the Australia and New Zealand man­ag­ing direc­tor of TubeMogul, be­lieves that stan­dard­i­s­a­tion does have its mer­its but could also in­ad­ver­tently slow down the battle against ad fraud.

“There are sit­u­a­tions where stan­dard­i­s­a­tion by in­dus­try bod­ies is help­ful,” says Smith. “The stan­dards adopted for video viewa­bil­ity adopted around the world ob­vi­ously helped ad­ver­tis­ers have a sin­gle met­ric for whether a viewer had an op­por­tu­nity to see an ad and what that means. That said, there is a wealth of data avail­able to ad­ver­tis­ers in dig­i­tal ad­ver­tis­ing and we should not de­tract from the di­ver­sity of what is avail­able by be­ing too rigid as an in­dus­try.”

The very di­ver­sity of ap­proaches cur­rently be­ing used by in­dus­try play­ers means that fraud­sters are fight­ing on myr­iad fronts—and this makes it more dif­fi­cult for them to game the sys­tem.

How­ever, there is also room for col­lab­o­ra­tion in this frag­mented ap­proach. As Smith ex­plains: “We be­lieve the best ap­proach is to de­velop your own tech­nol­ogy and also work with other lead­ing play­ers in the mar­ket.”

But even if the in­dus­try does col­lab­o­rate and find ways to ef­fec­tively mit­i­gate the prob­lem, there’s no guar­an­tee that ad fraud will dis­ap­pear. In much the same way that sweat­shops are still pro­duc­ing fake prod­ucts, there’ll al­ways be un­scrupu­lous char­ac­ters who find ways to make il­licit prof­its on­line.

Newspapers in English

Newspapers from New Zealand

© PressReader. All rights reserved.