Programmatic advertising (PA) is a multi-faceted term. Many market players use it as a buzzword, a label for the hype that has at times raised very high expectations among lots of market players, especially advertising customers. Others frequently use PA as a synonym for automation projects that are several years overdue, especially in so-called classic media, but in which nothing is “programmatic”. At mediascale, we generally define programmatic advertising as data-driven media-buying and as a process we are only just beginning.

Therefore, the disappointment that may have occurred with one or other advertisers is not a fundamental issue for programmatic advertising. Rather, it should be an incentive to take programmatic to the next level: on the one hand, by rethinking the set-up of service providers and technology; on the other, individual expectations should be reasonably calibrated.

In recent years, it has mainly been venture capital-financed players, who wanted to be part of the media value chain, who have fuelled the programmatic hype starting from their own interests, which has led to high expectations. And of course they have claimed “their” part of the supply chain. But those who worked more intensively with the market knew that the quantity and quality of the available profile data for programmatic campaigns is limited. However, only good data can increase the efficiency of campaigns significantly enough so that the additional costs for the additional members of the value chain are reintegrated. A possible disappointment was thus an announcement or based on unrealistic expectations.

In many conversations with our customers, we have realistically presented both the possibilities as well as the limits of data-driven advertising in order to rule out exaggerated expectations of PA from the outset. In doing so, the following assumptions were made, which our customers, sometimes against their initial will, have been getting along well with so far:

• Programmatic advertising is not a new channel with completely different rules to traditional display business. Also when auctioned and backed by data, a content ad remains a content ad and will not develop the advertising impact of an instream pre-roll large format
• Meaningful, validated data is the indispensable basis for programmatic advertising. Here it is important to look carefully and carry out comprehensive auditing of the available data offers. At first glance, the data market in DMPs seems expansive. But data segments that deliver what they promise (delivering a corresponding uplift to campaigns) are by no means abundant. And they have their price.
• An impression that cannot be assigned to valuable data should not be bought programmatically. As meaningful as it is to uniformly track all advertising contacts and accumulate all campaigns and pseudonymous profile data in one system, it makes little sense to put untargeted campaign volume into systems just to have bought it “programmatically”. This results in costs and technical performance losses that are not offset by financial added value.
• The open market, open to all, originally proclaimed by many to be programmatic’s central promise of salvation, has created more problems than it solves, as it has also opened up the market to a plethora of dubious players. The efforts of the large, open sell-side platforms to push the black sheep out are commendable, but unfortunately not always successful. That is why we only buy inventories that we can thoroughly test, both technically and commercially. Furthermore, whenever possible, we buy from partners (often in private marketplaces) that we know and have established business relationships with – including any sanction options which may be necessary in the interest of the customer in an emergency.
• And we’re not forgetting the creation: What use is the most sophisticated planning on a profile basis, if only one means of advertising is available? That’s why data driven creativity is, in our view, the indispensable fourth pillar of programmatic advertising – alongside technology, media space and data.

Today, programmatic has already caused major changes in the digital media business. But we are sure that this transformation process is far from finished yet. And it will encompass more and more media types in the future: TV, out-of-home, audio, cinema and eventually also print. In five years at the latest, we will be able to plan, book and control more and more channels via programmatic. In addition, people’s media use is evolving, new, relevant platforms are being created at ever-increasing speed, and data protection requirements also require fundamental and sometimes new solutions. All these challenges keep us busy and agile. Staying at the current level of development is not a solution. Particularly as we are just scratching the surface with programmatic.

This article was first published in adzine.