Indeed, when this question arises, the fuzzier areas of these concepts become evident. In these discussions, ‘data-driven’ is frequently interchangeable with ‘programmatic buying’, ‘real-time bidding’ or ‘profile-based modulation’ – which are ultimately and can only ever be aspects of comprehensive data-driven marketing. Data-driven advertising first and foremost refers purely to what is inherent in the concepts: marketing and media are planned on the basis of data and findings. In principle, it does not conflict with good old-fashioned range-optimised media planning. Strictly speaking, data-driven methods have always been as good as any TV strategy, even based on AGF files.

The question should rather be: Are programmatic, profile-based media-planning approaches related to efficiency and meeting objectives considered as ‘simple’ reach planning?

In most cases, when referring purely to e-business, a mass medium like TV falls behind digital channels in terms of ROI, especially performance channels, but here, only the efficiency aspect is considered. With regard to effectiveness, another picture frequently emerges. In day-to-day practice, it is shown that from a certain point of view, digital performance channels and programmatic approaches are still only limited in scale and one reason for this is because lucrative profiles in the digital world are ultimately also finite. Broad-reach channels, such as TV, are indeed widely scattered, but they reach a much bigger public for a very economical cost. Demand is generated, which ultimately also pushes the digital channels. At this point, the ‘unmentionable’ should actually legitimately lead to scatter gains in a quite specific way.

Aside from this, we have only been talking about pure e-business up until now. The majority of sales still takes place over the counter and the client base reached via digital channels is in the main limited. A housewife leafing through offers in brochures that land in the letterbox on Saturday mornings is still a more common phenomenon than some might think.

If we look at the current market, it becomes clear that in the end, most providers putting out programmatic advertising or data-driven marketing on banners are limited purely to digital channels. Offline channels, like TV, are generally only considered based on ‘simple’ estimations of short-term TV uplift. Thus, only a sub-division of marketing and media is visible. Moreover, other relevant factors of influence, such as weather, seasonality, etc. are not represented.

For comprehensive data-driven marketing, aside from digital measurability, statistical approaches and algorithms are also required, in order to fully quantify the effect of marketing and media. With Plan.Net Business Intelligence as the analytical arm of the Serviceplan Group, we have hereby gradually expanded our competence in recent years and created an increasingly extensive pool of knowledge.

The results can be summarised quite easily. Large-format, broad-reach media activities are most effective in impacting brand success. Mind you, they are the most effective, but not necessarily the most efficient. Decisive for success is the successful orchestration of all channels. Broad-reach media generate general demand, even amongst consumers who do not necessarily form part of the core target group. Here, therefore, there is no significant conflict identified with programmatic buying. Additionally, consumer relevance can be enhanced with broad-reach media; for example, via dynamic content and motives. However, it is consequently essential that this increased relevance or demand is channelled and finally converted.

Incidentally, neither do we seem to be on our own in drawing these conclusions. There will certainly be a reason why, ironically, data-driven companies such as Amazon have stepped up their investment in traditional TV advertising.

Let us just look for a moment at typical target group definitions or segmentation. Most of the time, these need to aim at addressing the core user base of a product or brand – which is only right. In the end, however, it will be noted that this core target group comprises only 20-30 per cent of the actual users/buyers. Alternatively, it is of course possible to extend the target group across the board, which is hardly target-specific. Firstly, therefore, ‘reach’ cannot be substituted for anything else, especially with regard to mass-market goods. However, please don’t get me wrong, it should nevertheless be the right reach in the right target group.

Let us just assume we transfer all of this to a programmatic, profile-based approach in a completely simplified form. First, actual product search is omitted. Strictly speaking, 70-80 per cent of potential buyers would not be reached. Clever use of algorithms would of course recognise this and correspondingly extend the target group segment – in the end, the result would once again be reach-optimised planning.

There is no doubt that programmatic models are a thing of the future, even if it has not yet been ascertained how rapidly they will develop. Nevertheless, it will not mean that in future campaigns will only be modulated on individual, supposedly lucrative profiles. The more oriented the product towards the mass market, the more likely it is that communication will be based on a broad reach. Here, programmatic modulation can surely increase consumer relevance. In the course of this development, the above-mentioned target group concept will also clearly be a dynamic one.

We are now already looking at the transition between broad reach and brand campaigns and what is readily referred to as online performance becomes ever more fluid. Therefore, the exchanging of branding formats and modulation in a target group for acute, interest-based communications about product offers, and vice versa, takes place with increasing speed and momentum – ideally using devices.