89 percent of marketers worldwide struggle to create a single view of their customers because, according to a survey performed by Experian, they have poor data quality, siloed departments, and the inability to link technologies. In addition, these marketers report that the top barrier to cross-channel marketing success is lack of this single customer view.
At Amperity, we often talk to CMOs, CCOs, and IT leaders who are trying to unify their customer data. Nearly all of them tell us that despite months (or years) of effort, expensive consultants, and heroic efforts by their IT teams, they still lack the single view they set out to achieve.
What’s Really Holding Everyone Back?
What’s the real problem? Is customer data unification really so hard? In our experience, customer data is often siloed in difficult-to-link technologies. But rather than poor quality data, we find that the data itself is rich and valuable. When brought together with intelligence, it forms a deep and useful customer knowledge base that can power advanced analytics, personalization, and targeting.
So while access to siloed data can be an issue, there are two more fundamental blockers that we encounter when we work with brands to unify and make their customer data actionable:
The problem of identity resolution has never been so complex as it is today, due to the explosion of customer data, channels, systems, and applications. To achieve a single customer view, data has to first be brought together from these disparate systems, which requires building, maintaining, and monitoring data pipes into and out of a centralized system.
But after data is co-located, the real challenges begin. Customer identities must be resolved across tens of sources and billions to trillions of records, even when there is no common key to link them. This requires massive computing power, machine learning intelligence, and customer data expertise.
Lack of ground truth
There is another related, but deeper, complication. Customer data is inherently uncertain. Without that common key, can you tell with assurance when two records belong to the same person? Can you tell when they don’t?
There is no way to be certain, and yet choices must be made: either leave records separate when they have a high likelihood of belonging to a single individual (producing false negatives), or unify records when you can’t be sure if they do (producing false positives).
The business impact of these choices is huge. False negatives in customer data unification means duplicated lists, multiples of the same emails and direct mailings sent to a single individual, and ads shown to customers who have already purchased the products you’re promoting. A lack of unified data also makes effective site, app, and email personalization difficult to achieve. All this rolls up to lower revenue, poor customer experiences, and loss of loyalty.
False positives, on the other hand, mean that some of the information in a customer’s profile is false. Using profiles built with potential fast positives runs the risk of sending order confirmations, shipment notifications, and information about returns to the wrong individual, with disastrous consequences.
A Better Way
We believe brands shouldn’t have to make this difficult and expensive choice. Instead, we help brands build multiple, unified customer views, connecting rich data stores to form robust customer profiles at scale. To learn more about unified customer profiles and multiple customer views, check out our Resources page.