Accuracy is a core tenet of identity resolution. In this context, it's all about continuously verifying people’s identities across all their channels and devices to ensure you know who you're talking to.
Many marketers fail to realize that it's easy to not do accuracy well or get it wrong. For example, you could think you’re messaging multiple people, when you’re actually messaging the same person with the same ad on all of their devices.
To explain this concept more, our SVP of product management, Dave Scrim, shares the two most important things to know for ensuring your programs have accurate identity.
Video transcript:
Dave Scrim: Being able to reach your target audience at scale is critical, but it's not enough. Vendors often talk about match rates or reach rates, but you rarely hear anybody talking about accuracy rates.
Accuracy is about continuously verifying someone's identity in a privacy-safe way. To get accuracy, right, you need two things, quality data and the ability to scale it.
Most identity solutions out there today use unstable or fragmented IDs, such as device IDs, cookies or email addresses. This makes it really difficult to tie a profile together and get a holistic view of the customer.
For a more accurate connection, transactions are the gold standard because when you buy something, you use your real name and address every time.
You're probably saying to yourself, I have transactions, so I can make that connection myself. The trouble is, you typically only see a couple of transactions a year, not the volume you need to really tie together all the various cookies and device IDs. In order to do that, you need a network of thousands of partners working together.
It's important that you ask your vendors not just about match rates or reach rates, but what are the accuracy rates? How do I validate them? From my perspective, you shouldn't settle for anything less than at least 95% accurate.
Accuracy is the foundation of any strong marketing program. Without it, you won't be able to target the right audience, you won't be able to get the right message out, and you won't be able to measure your programs effectively.