Chapter 4 Owning a Data Management Platform
Tracking data helps companies to better understand consumer behavior and to customize their services. Moreover, it allows them to better target potential customers. Similar words and slides are usually used to convince advertisers to join the pack, invest in their technology stack and buy a Data Management Platform (DMP). The following paragraphs detail what advertisers should consider buying into the promise of tracking and personalization. While no two advertisers are the same and the ROI depends on advertiser specific setups and parameters, there are four pillars to take into consideration; data and technology integration, potential use cases and finally cost.
4.1 DMP data integration
User behavior on websites and platforms is mostly captured via tracking pixels (cookies). A “pixel” or “TAG” is a small piece of software that is loaded in the background of a website to collect information (undetected), about users and their behavior on the website. Providers whose pixels/TAGs are implemented on multiple websites can track user behavior across the different websites to create detailed user profiles. These profile segments are very useful from the advertiser’s perspective because they enable them to address members of their target group in a favorable environment. The key challenge with current cookie-based tracking is at least twofold. First, walled gardens such as facebook, youtube, amazon do not allow marketers to track their advertising. In general, the ability to track with DMP is very limited. This is a general trend; providers with valuable audiences will bundle data and inventory to avoid data leakage and monetize the audience on their properties. The second effect is public awareness. The Cambridge Analytica scandal increased the awareness for data privacy. Users ask for better privacy control and companies respond. Apple and Mozilla both improve the user’s possibility to opt out of any tracking. We assume that Google will follow in the next year by limiting 3rd party tracking in their browser and ad network. Finally, young, tech-informed audiences increasingly use ad and tracking blockers. To sum up, the data that DMPs will be able to collect in the future will be lacking key ad impressions, key audiences and be patchy at best.
4.2 DMP technology integration
The other key ingredient is technology integration. The insights and audience data collected in the DMP needs to be connected to systems that change the user experience. These are basically three systems; Demand Side Platforms (DSP) buying individual ad impressions for users in certain target audiences, ad and onsite personalization engines with change the message and advertising content depending on the target audience. The systems need to be connected and work seamless together. Currently this is not the case as there are no industry standards. Hence advertisers need to carefully check which technology works well with other system in the ad/martech ecosystem. Besides the pure connectivity there are two quality metrics that are important to consider; speed and match-rate. Speed describes how often the systems are synced and how fast user profiles are transferred between the systems. Currently, it can take up to 24 hours for a user profile to be transferred from the DMP to DSP. In a retargeting use case; most advertisers are better off using DSP retargeting pixel with no time delay. Finally, the metric match-rate describes how many users profiled gathered in the DMP can be used in the activation. The rates vary by country and technology combination but are usually in the range of 70-80%.
4.3 DMP media use cases
Most use cases focus on three areas: Targeting Costs, Frequency Management, Personalization.
Potential use cases for a ROI positive DMP implementation.
While the use cases can be generally be solved with the DMP; the key question is: “are there alternative solutions?” and what is the gain/lift by using the DMP versus using the alternative. In Germany, targeting cost (on display and video inventory) is less than 5% on average. Key publishers such as facebook and google provide targeting data without additional costs. One key asset DMPs claim to have exclusively is retargeting, segmentation and lookalike audiences. All these aspects can be (better) solved with the current DSP solutions. Frequency is traditionally measured in the adserver. As there is currently no link from the adserver to the buying systems (DSPs); overall frequency management is an issue if the advertiser uses multiple DSPs at the same time. Currently most advertisers in Germany follow a single DSP strategy and can use their DSP to control frequency and overexposure. Ad and Onsite Personalization: Ad personalization focuses mainly on retargeting segments and showing these segments the last (most relevant) offer. This case is due to the DSP speed advantage better served with the direct DSP integration. If the knowledge of the last viewed offer is transferred to the ad system with a delay of 24 hours, the potential advantage of segmentation and better prediction is minuscules. ## DMP cost considerations There are three cost aspects to consider: Implementation, maintenance, and license. DMP implementations are time-consuming and cost intensive. We have seen consulting companies working on an implementation for 4 months with 3 FTEs. The license cost of well-know DMP brands is around 100-200K per year. The maintenance is far from automated. Tags need to be managed, campaigns to be trafficked multiple times and clients need to build a strategy and new workflow around the technology. All in all, DMPs are not a plug and play solutions but rather a complex, cost and time-consuming project.
The sales slides of DMP vendors promise the world of complete tracking and real-time personalization. The German reality is more complex. As a rough guideline; even clients with substantial digital investments in open market video and display inventory will have a hard time to turn a DMP project into a ROI positive endeavor with media cases only.
Questions an advertiser needs to consider: - Is there a positive ROI in DMP projects? - What is the total cost of ownership? - What are the touchpoints that can reliably be measured? - Do you have user groups which have excessive ad exposure and can this be reduced? - Which of the use cases can only be handled by a DMP?