Who is accountable for marketplace quality? ExchangeWire study exposes market’s views on fraud in programmatic advertising

In 2014 programmatic transactions hit the $21bn mark globally, yet a total 84 per cent of respondents said they believe there would be a further “significant” increase in programmatic investment if marketplace quality was no longer an issue.

Desktop display was found to be the weakest link to more than half (54 per cent) of respondents when it comes to poor quality traffic compared to video and mobile. Just under half (44 per cent) of marketers said they regard marketplace quality in video as “very serious”, with 42 per cent viewing it as “somewhat serious”.

Meanwhile only half of respondents said quality is an issue for mobile display, which the report attributed to the fact that fraudulent processes are often harder to carry out in mobile, compounded by the fact mobile CPMs are lower, making them less attractive to fraudsters.

Carat’s global chief digital officer Anthony Rhind, who took part in the qualitative part of the study, called for the media owner and supply side of the market to take accountability and stop flooding the market with bad inventory.

Rhind said: “The industry must take the issue of marketplace quality seriously. I don’t think price reduction is the starting point – we must work with the publishers to increase visibility in order to justify preserving pricing.

“We also need to ensure poor metrics are indicative of non-visible or non-human traffic, as in some instances investigation shows that alarming metrics are in part due to measurement failure.

Nonetheless, the supply side needs to stop pushing invisible, cluttered or fraudulent impressions ot market as these erode the value of all stakeholders. The marketplaces that manage receivables and payables should embrace that position by proactively stopping revenue return to the bad actors pushing non-inventory."

Programmatic advertising, though lauded for its efficacy in media trading, has not always been compared favourably to more traditional media channels such as TV and outdoor, when it comes to brand safety, which have boasted more tightly controlled environments to the open exchanges. However, the study revealed that 65 per cent of marketers do not think programmatic is more risky than other channels due to poor traffic quality.

Huwawei’s global marketing director Nick Graham, who also took part in the study, said: “Risk exists everywhere media is traded. Outdoor had its crisis a couple of years ago. TV created the quality versus value discussion, and other areas of ‘accountable’ digital like paid search, create opportunity for risk, albeit via optimisation or lack thereof.”

The rise of private marketplaces (PMPs) has helped curtail the rush of fraudulent traffic, and 58 per cent of marketers polled in the survey said they believe PMPs are currently the only way to avoid quality issues.

However, a quarter of respondents said they believe that PMPs are still vulnerable to poor quality traffic, which always “finds a way through”.

The onus for who should take responsibility for improving marketplace quality seemed to fall to all members of the value chain, with 98 per cent of respondents saying the sell side, which includes publisher, supply-side platforms (SSP) and exchanges, should take responsibility for improving quality. Yet almost as many (91 per cent) said it should fall to the demand side platforms (DDP) on the buy side. Just over half (54 per cent) thought it should be law enforcement that fights the issue.

BBC Worldwide senior vice president of sales operations, Tom Bowman, agreed accountability should be divided across the chain. "From a publisher perspective; an advertiser is coming to your site because of the quality of the audience, if you degrade the quality because of the technology you use, then you will lose the advertisers [and revenue]," Bowman said.

"Publishers need to apply the same quality control when it comes to advertising as they do content and take responsibility for how they engage with their consumers through advertising. Technology companies need to stop over promising and humans need to take control," he added.

Carat's Rhind believes that the responsibility and opportunity lies with publishers: “The better the inventory quality the higher the value of advertising inventory. Whoever can prove they have the highest quality can price discriminate. Therefore the upside sits with the publisher to justify a premium.

"The role of the media buyer is to buy the right inventory at the lowest price possible, this can be a very high CPM where inventory quality is good. If someone else can buy the same inventory more cheaply, it is an unavoidable conclusion that the media buyer has done a bad job."

However, ComScore's vice president of effectiveness, Duncan Trigg, disagreed, calling it "churlish" to blame publishers. "No individual re-seller or seller should be refereeing their own game," Trigg said. Industry standards are very important but the policing of those standards needs to be independent and not tied to revenue.

He continued: "Ultimately marketplace forces will drive better standards, today fraudulent impressions are the same price as legitimate impressions. Pressure will initially come from agencies and brands but will flow downhill to the pubs and marketplaces."

The report highlighted inconsistencies between marketers’ expectations of quality today and in the future, across all aspects of the marketplace. Fraud in particular showed a sizeable discrepancy, with perceived fraud levels considerably higher than those deemed acceptable.

Marketers believe 27 per cent of traffic is fraudulent but are only willing to tolerate up to five per cent. In the sub groups publishers / media owners and brands held a zero-tolerance view.

Viewability has been a hot topic for last year among advertisers, with many having taken a tougher stance on how much they are willing to pay for ads that do not meet their standards of what is deemed in view. Unilever and Shell are among those to have laid out tougher expectations, yet the issue will remain thorny until the discrepancies between the way different vendors measure viewability are reduced.

The report highlighted major tensions between the buy and sell sides with regards to what are acceptable viewabilty levels. It revealed a 25 per cent discrepancy between marketers’ perception of viewability (45 per cent) and what they consider to be acceptable (70 per cent).

A total 36 per cent said they considered 70 per cent and over an acceptable level of viewability, though 39 per cent of the publisher/media owner group were comfortable felt 60 per cent and over was more feasible. In contrast brands revealed a more stringent view with 40 per cent selecting 80 per cent and over as an acceptable level.

The study, conducted in association with Open X, marks the first to come out of ExchangeWire's new research arm, headed up by head of insights and analysis Rebecca Muir. Her blog on the findings of the research will be published later today on The Drum.