GDPR will Trigger a Massive Transformation of Programmatic Advertising

GDPR means the end of this unrestricted use of cookies and hence the end of programmatic trading as we know it.

Programmatic advertising has become the dominant way to trade digital media because it matches ads to people on a truly vast scale. This ability to scale depends on reliable and exchangeable device identifiers. Cookies (and other device identifiers) have been used to identify and track a billion devices across the Internet to enable programmatic trading. GDPR means the end of this unrestricted use of cookies (or any device identifiers), hence the end of programmatic trading as we know it.

This coming transformation in programmatic trading comes at a time when the market itself is under fire from both advertisers and publishers alike. This is best illustrated by two big things that should be the cause of industry-wide reflection and reform. Firstly, back in January Procter & Gamble’s Chief Brand Officer Marc Pritchard made that famous speech. He made it clear that one of the world’s biggest buyers of media is fed up with the lack of transparency in programmatic trading. Describing it as “murky at best, and fraudulent at worst”. He also made the simple and damning point “There is no sustainable advantage in a complicated, non-transparent, inefficient and fraudulent media supply chain.”

Secondly, in March the Guardian, one of the world’s largest sellers of English language media, began legal proceedings against Rubicon Project “for the recovery of non-disclosed buyer fees in relation to Guardian inventory.” The two big issues for the Guardian were that “secret commissions” were allegedly obtained (by Rubicon Project) without the Guardian’s knowledge or consent; and that it did not have visibility of how much buyers were willing to pay for its media, which led to reduced ad revenue.

It seems that programmatic woes all come down to the same thing: lack of transparency. I predict that this burning and recurring issue of transparency will mean that the coming upheaval triggered by GDPR will be very rapid indeed. Lack of transparency in programmatic trading is the root cause of huge market distortions and illegal and dysfunctional behaviours:

Massive oversupply of inventory: Since publishers never get to understand the true market value of their inventory, they are incentivised and encouraged to run more placements from multiple ad platforms. Transparency with respect to real demand will solve this problem and reduce the supply of inventory to better match demand.

Cyber crime/Ad fraud: Lack of transparency in the trading environment combined with the proliferation of inventory and ad platforms means that it is much easier to commit fraud on an industrial scale (both ad placement fraud as well as adware/malware). Market transparency may not eliminate fraud but will make it much harder for large-scale fraudsters to remain undetected for long.

Uncertain ad delivery: The conflicting interests of advertisers and publishers with respect to impression counting has been exacerbated by the lack of market transparency and over-supply of inventory. This, in turn, has led to advertiser anxiety about ad viewability. Transparency will change the ecosystem to bring the ultimate buyers and sellers into better alignment and equilibrium.

What Does Lack of Transparency Really Mean?

Markets work best when information is freely shared between buyers and sellers. It is the only way for supply and demand to reach equilibrium. This is elementary economic theory.

Lack of transparency is fundamentally about gross asymmetry in how market information is distributed and controlled. To fully understand this point, lets be clear, there are two types of data involved in programmatic that are being suppressed.

Firstly audience data, which is used to target devices. Audience data comes either directly from advertisers, directly from publishers or from data brokers (indirectly from publishers). To differing degrees, this data is derived from tracking of devices via a variety of methods (e.g. cookies and device fingerprinting). The practical use of this audience data relies on the huge number of companies operating in the ecosystem being able to reliably share identify devices (as a proxy for a person) and share these identities freely. The mechanism used to do this is called cookie matching (or synching). Cookie matching enables the linking of two sets of profiles of a single device (person) in the databases of two independent companies, e.g. an ad exchange and a DSP. Without being able to match cookies, DSPs will not be able to use their 3rd party data (from data brokers) or data from advertisers/clients. For a more in-depth understanding of different types of audience data please see one of my previous blog posts here.

Secondly, there is market information or inventory pricing information, i.e. the media budget of a buyer, the price a buyer is willing to spend per impression and the price a seller is willing to accept per impression.

Lets not beat about the bush, the reason why we have information asymmetry and black box trading in programmatic is that arbitrage is the business model of the largest ad platforms (whichever way it may be calculated or justified). The most effective way to operate any arbitrage business model is to suppress market information. As in any market where there is gross information asymmetry, there is also gross inefficiency and extraordinary financial returns to players with this information. This is true of media trading as it is of any other traded commodity. Most ad platforms suppress both audience and market data. Only when true market values are hidden, is it possible to manipulate perceptions of value from both the Advertiser and Publisher perspective.

In digital media, the problem gets even worse since the largest ad platforms work for both buyers and sellers. An “End-to-End Ad Platform” is not a common industry term, but it is a useful way to describe technologies and businesses that both buy ad media from Publishers and sell ad media to Agencies or Advertisers.

Naturally, the large media agencies have realised that money is to be made with an End-to-End ad platform business model. All of them, in some way or other, are implementing End-to-End ad tech stacks in order to capture as much margin as possible. Chris Smith (formerly at Digiday) wrote an amusing article, where he described how large media agencies refer to their attempts at hiding their secret margins as “Hiding The Turd”.

The fundamental problem is the structure of incentives in the business model of End-to-End Platforms leads straight to moral hazard. It is logically impossible to structure the incentives of an End-to-End platform in such a way where all parties benefit: there will be winners and losers. End-to-End platforms are venture backed profit making businesses, which need to return an ambitious ROI for their private equity investors. End-to-End platforms profit most when they are able to simultaneously maximise media price to Advertisers and minimise media price to Publishers. This is both the reason why we have so little transparency of market price data from large ad tech companies and why so many publishers complain about eroding ad revenue.

This lack of transparency breeds a deep lack of trust between publishers and ad platforms, hence publishers tend to both work with multiple ad platforms in order to keep each one honest. Multiple ad platforms, eroding ad revenue and lack of information also encourages the proliferation of ad placements.

The constant glut of inventory is due to publishers constantly trying to make up for a system skewed against them. This excess supply of inventory and multitude of platforms per site has several damaging symptoms: it drives inventory prices down, it reduces the quality of many ads, encourages people to use ad blockers and provides the perfect environment for fraudsters. As if this was not bad enough the lack of transparency has another malicious side effect: it creates a wall of distrust between advertisers and publishers. From an advertiser perspective, this has resulted in deep anxiety about viewability of impressions as the natural conflict of interest between media owner and media buyer are unnecessarily exacerbated by ad platforms pushing for ever more unwanted supply.

GDPR: A Catalyst for Transparency

It is uncanny how the text of GDPR indirectly addresses the issue of transparency in programmatic trading. Under GDPR ad targeting relies on specific and unambiguous consent (which is qualitatively different to an opt-in for those that don’t know). If consent is required, then this means all legitimate targeting data is by definition 1stparty data. This means that 3rdparty data is effectively illegal. Even if available, no reputable brand or publisher would want to risk their reputations, let alone fines for some dodgy 3rd party data of dubious value.

The first implication and the most important is that there are only two places to gain a consent for marketing purposes: the publisher site or a brand site. Both brands and publishers will have a much easier time to become data controllers in their own right and run their own campaigns. The role of the ad platform in matching ads to people is much diminished as people are (at least in part) matching themselves to ads. Ad platforms will simply no longer have an audience data advantage over advertisers and publishers (at least the larger scale advertisers and publishers).

The second implication is that consent (especially consent for a specific brand) is binary; there is either consent or not. Hence the perceived value of any impressions is much harder to be manipulated.

The result will be that both market data and audience data will be much better accessible to both publishers and advertisers, so arbitrage by ad platforms will be much harder to achieve. The only way for an ad platform to succeed post-GDPR is to gain trust and become fully transparent. Hence GDPR will usher in a transition in market power from market intermediaries to the ultimate buyers and sellers.

For many ad platforms, especially the big end to end platforms, this is may be their Kodak moment of market disruption. Just like Kodak’s internal thinking and structure meant that it could never effectively respond to the changing market environment and totally mistook where the market was going, the same may be true of ad platforms that currently rely on secret arbitrage to survive. They must ditch the arbitrage model to gain trust, but if they do they are out of business: unless they change their product. My prediction is that “trust” is the new business model. Sorry if that sounds abstruse, but this is the business model my company, AdUnity, is working towards and this is a blog posting, not a sales pitch.

Another key consequence of GDPR will be a reduction of inventory supply as publishers will have better information to match supply to demand. This will reduce the ability of fraudsters to commit fraud and reduce advertiser anxiety about viewability. GDPR effectively recommends the use of an audit trail for targeted advertising (for complaint handling purposes). This audit trail can also be used to tackle fraud and viewability. Hence, all in all, GDPR will make serious inroads into eliminating fraud and viewability concerns of advertisers.

GDPR may be a European law, but the effect will be felt globally. This is not just because any targeted ad campaign targeting EU geos will require compliance, but because it actually solves many of the problems that have brought the industry into disrepute. Programmatic trading will be transformed within 12 months, the strange thing is that most ad tech players seem oblivious to it.

About AdUnity
ADUNITY is an ethical ad technology company based in the UK with offices in both London and Bucharest. We deliver trust, transparency and advanced technology in the programmatic ecosystem. AdUnity provides GDPR ready Consent Based Advertising solutions and transparent trading platforms for publishers (no black boxes) and does not operate an arbitrage model; just a flat revenue share. We comply with the latest programmatic standards (OpenRTB 2.4) and format standards (HTML5, VAST 2.0, VPAID 2.0, MRAID 1.0, Native Ads API 1.1).


Author: Rupert Graves

20 years commercial and product experience in digital media, mobile and advertising, with deep domain experience in programmatic advertising and media trading. Sector Experience in: Advertising, Digital Media, Mobile/Fixed Line Telecoms, Broadcast TV, Consumer Electronics, SaaS. This includes B2C and B2B marketing and product development.   View Rupert Graves's profile on LinkedIn