What is identity resolution? A guide to building an identity graph
by Zeke Sexauer, on October 10, 2019
In this era of people-based marketing, getting identity resolution right is foundational to the success of all aspects of marketing strategy, impacting everything from insight generation and media optimization to measurement, attribution, and the delivery of better individual customer experiences.
But identity resolution isn’t easy.
Cisco’s Visual Networking Index projects that American consumers will own up to 13 devices and connections by 2022. Every time a consumer uses one of these multiple devices or channels to interact with a brand, a different identifier—such as email address, device ID, cookie, IP address, or phone number—is attributed to that individual. It’s up to the brand to understand which identifiers belong to which consumer, as well as what these various signals and behaviors mean.
Marketers must be able to find a way to connect all these disparate cross device identifiers back to the people they stand for by creating and maintaining a database of persistent customer profiles, known as an identity graph.
However, most companies’ ability to connect their first-party identifiers to individual profiles at a scale that is consistent and accurate enough to leverage for the benefits and use cases needed is limited. Thus, marketers must augment their first-party identity graphs with third-party data sources and identifiers that allow them to better activate the data.
How to build an identity graph
Typically, marketers have had two ways to augment their first-party identity graph with additional customer information: onboard to a third-party identity graph or acquire the data directly.
Use a third-party identity graph
Third-party identity graphs collect identity data from numerous identity originators and aggregate it into a proprietary identity graph. Brands can then push, or onboard, their first-party data (including offline data) from their data management platform (DMP) or data warehouse to the provider, which matches individual identifiers across touchpoints and links them to a proprietary persistent ID that follows the individual even as identifiers change.
After the linkage is made, brands then pay a fee to the vendor based on data volume, number of channels, and types of data licensed, in order to activate the graph for digital display advertising and consumer insights and analysis.
Using a third-party identity graph gives brands the ability to quickly and easily augment their customer data with additional identifiers, but at the expense of ownership of the final product. Brands don’t own the resulting customer graph, but rather “rent” it from the vendor, paying per use case and beholden to the criteria that the provider deems important, not the brand.
Brands don’t own the underlying data sets and linkages, and often have limited visibility into user-level data. Their ability to generate insights, optimize performance, and determine attribution are limited by the provider’s capabilities.
Furthermore, by renting a third-party identity graph, brands must place their trust in the matching methodology of that vendor. Third-party vendors often tout high match rates, but provide little visibility into how they determine what constitutes a match. A high match rate is of little use if the underlying data is inaccurate.
As identity becomes increasingly critical to measurement and attribution, insight generation, media optimization, and the delivery of better individual customer experiences, the costs and limitations of renting a third-party identity graph will only grow more acute.
You wouldn’t install a $10,000 speaker system in a rental car, so why would you base the core foundation of a multi-million dollar marketing strategy off of a rented identity graph?
Build an identity graph in-house
To maintain control over their identity graphs, the most progressive brands are bringing identity resolution in-house. Brands make direct deals with other companies to purchase their first-party identity data, which they can then match to their own within a customer data platform (CDP). With each new source of data, brands can fill in the holes in their customer identity profiles and create a robust ID graph custom-built for their specific needs, strategies, and goals.
Building an ID graph in-house gives brands control over the way that identity data is collected, matched, and stored. By purchasing identity data directly from the originators of that data, brands can ensure the data they’re feeding into their graphs meets their standards for quality, trustworthiness, and compliance.
Resolving identity in-house also gives brands control over how data is matched by choosing when to employ deterministic matching and when to use probabilistic techniques. When a brand controls their ID graph, they decide what tradeoffs to make for accuracy versus reach depending on the specific use cases and goals.
Furthermore, when brands own their identity graph, they can use the underlying data and linkages as they see fit. This allows them to extract maximum value from their graph beyond one campaign or application. The data can be used to activate a campaign, integrate with partners, analyze for customer insights, understand attribution, and more.
The limiting factor to this approach, however, is that it has typically been extremely time and labor intensive. For each new partnership, numerous cross-functional teams must get involved. Finance has to agree on pricing and payment, lawyers have to agree on a set of terms and conditions, and engineers have to connect systems and platforms—and that’s only after your data science team has determined that the data you want to purchase will actually be useful. Multiply this process by each supplier you partner with, and it could take numerous years and thousands of employee-hours to build an actionable identity graph.
The future of identity resolution
Brands are investing heavily in identity resolution, and will continue to do so. In fact, US spending on identity solutions is expected to increase by 30% each year between now and 2022, when total spending will reach $2.6 billion. The brands that will succeed in this environment won't be the ones using the same identity graph as their competitors. The competitive advantage will instead be given to the brands that are able to resolve identity in unique ways based on unique use cases.
What brands need is a solution that gives them the precision, transparency, and control of building an ID graph in-house, but allows them to do so with the speed, ease, and scale of using a third-party provider.
In the late 2000s, demand-side platforms (DSPs) transformed the world of digital advertising by enabling advertisers to evaluate ad space on millions of websites instantly, and then buy the inventory that met their specific criteria in real time.
Narrative is rebuilding the data industry in a similar way. Using Narrative, brands are able to evaluate identity data from numerous suppliers at once, and then purchase the precise data points they need based on their explicit criteria.
With Narrative, brands can quickly and easily build a proprietary identity graph according to their specific business needs, strategies, and goals. With full control and ownership of their identity graphs—and the data and linkages that underlie them—brands reduce risk, eliminate costly licensing fees, and gain the ability to build fully attributable, measurable, and actionable marketing strategies across the full customer experience.
Interested in learning more about how Narrative can help power your identity resolution strategy? Request a demo today.