Why Identity Graphs Are Key to the Future of DMPs and DSPs
Published: December 2, 2020
Author: Feliks Malts
In my last post, I talked about the role of measurement and major media platforms (Google, Facebook) in addressing tracking and measurement in a cookieless world. In today’s post, I’d like to put the future of data-management platforms (DMPs) and demand-side platforms (DSPs) under a microscope and explain why the key to both lies in the concept of identity graphs (also referred to as ID graphs or ID resolution graphs).
Before I get into DSPs and DMPs, it’s important to know that the concept of identity graphs was introduced (less robustly) about a decade ago by Tapad, which is still the leader in the space because it has been validated for accuracy and has a large pre-established addressable base.
As shown in the image below, identity graphs are databases that store and create linkages between a wide range of user identifiers (think usernames, email addresses, phone numbers, IP addresses, loyalty card numbers, mailing addresses, etc.) – across devices and offline engagements.
How identity graphs are informing DSP advertising today – and in a cookie-less future
Just as Google, Facebook, and big unified marketing measurement (UMM) vendors (Nielsen Attribution, for example) are going to move toward a deterministic model of conversion measurement, large DSPs are following suit. Like UMM vendors, they’re building identity graphs to tie back identities of the same people on different sites by partnering with Tapad, as well as developing their own solutions in some cases.
The identity graph data comes from DSPs working with large content networks (think Hearst properties like Good Housekeeping and HGTV Magazine, which reaches a huge audience on any given day) to track the behavior of the syndicated user base of those content networks. The graphs update continuously as users authenticate and take action; identifiers like IP addresses or work email addresses allow both DMPs and DSPs to probabilistically assume user identification and send the anonymized data back to the identity graph.
In this way, DSPs get anonymized, fresh, at-scale data of user behavior that allows them to maintain robust retargeting campaigns.
Many DSPs (including Google’s DV 360) still rely on cookies to build their identity graphs, but what happens when those cookies go away? At this point, only a select few have built solutions to thrive in a post-cookie world.
The Trade Desk, a leader in the DSP space and 3Q’s preferred DSP partner, has robust integration with third-party data providers; they’re ahead of the curve because they’re partnering with Tapad to build out their own identity graphs. They’re also leading the Unified ID 2.0 initiative, building a network with other partners to unify to the same identity across a larger vendor cohort (Nielsen, for one, is joining them in this initiative). The bigger the partner ecosystem they build, the richer, more addressable, and actionable the data. Any advertiser looking for a new DSP should be looking for those positioned to perform post-cookie, and the DSPs more integrated with other partners will have a huge leg up in scaling addressable audiences.
How DMPs are working with identity graphs
Forward-looking DMPs are also building identity resolution solutions by partnering with Tapad; these solutions allow advertisers to build audience segments and target them across properties where they have partnerships in place (for instance, partnerships with Google, Facebook, and major DSPs will allow DMPs to directly sync identities across those channels and collect first-party and third-party data as the users browse). Lotame, 3Q’s preferred DMP partner, was the first DMP to proactively build a solution that lets them continue to deliver on their business model without cookies by moving third-party data dependency to identity graphs.
As with DSPs, the bigger the ecosystem of the identity graphs, the more deterministic data DMPs will be privy to, and the more value they can add to advertisers by addressing audiences using robust third-party data.
The “holy grail” of data will be…
Lotame and The Trade Desk combined is almost a perfect union; Lotame is a DMP that provides addressability at scale, and The Trade Desk is a DSP that provides a vehicle for targeting ads to that addressable base of users. (Rounding out the picture, Nielsen Attribution’s UMM solution taps into the identity graph to provide a full picture of attribution across engagements). This full solution also has a big bonus: it brings more devices and channels, including fast-growing CTV, into the addressable, measurable ecosystem.
As shown in the image below, all the dots are connected to form a full picture of the customer and the customer’s journey to purchase without sacrificing individual privacy.