If you’ve read any of my other posts, you know how passionately I believe that customer experience should be the organizing principle and top priority of every business. Focus on customers is, to say the least, highly correlated with success. Customer understanding—knowing your customers and anticipating their needs even before they might themselves—makes it possible to design consistently good customer experiences that build successful businesses.

One of the most crucial resources for customer understanding is data. You might be forgiven, then, for thinking that I’d be a fan of customer data platforms (CDPs). After all, they promise to pull together all of a company’s customer data from any number of sources to help analyze, predict, and respond to customer behaviors in ways that improve marketing results.

But I’m not. Here’s why:

  • We live in the real world, where customer data is imperfect, and continuously imperfect at that. The pace of change means that no customer data source can ever be perfectly up to date and consistent. And besides, although we’re getting better at prediction all the time, customers will still do things we don’t expect them to, or when we don’t expect them to. Implementing a CDP to fix that is believing in a fairy tale. Only a company in stasis is going to have a totally consistent view of customers…and a company in stasis is not likely to be a going concern for very long.
  • There’s never been one consistent source of enterprise data for any length of time. Just as customer data is continuously imperfect, so is the structure of most companies. Mergers and acquisitions, expansion into new lines of business or geographies, new strategies, experimentation with new processes or technology approaches—all of these things mean that having only one source of data and insight is, and will likely forever be, an elusive goal.
  • Marketing isn’t the only part of the business that needs a holistic view of customers. Analyzing and taking action on customer data only from a marketing perspective completely misses the bigger objective. Customer experience spans all aspects of a business. Anyone who interacts with customers needs to have access to a consistent view of those customers and insights that make each interaction better. Creating yet another data silo, especially when it purports to be a single source of customer information, is a short-sighted, ill-advised investment.
  • Different departments and teams need different views of the same customer data. What’s most useful to customer service may be very different to what’s top priority for sales teams. What we need to know about customers varies with context. Insights that drive planning, strategy, or product development, for example, have a totally different scale and cadence to what informs effective real-time marketing efforts. Each part of the organization has its own lens or filters to get to the most valuable customer insights for them, but all need to work from a holistic, shared view. And while we’ll never achieve perfection, customer data should be highly consistent no matter who’s using it.

While the promise of CDPs falls short of the mark, these are fundamental challenges that need resolution. So what does getting it right look like?

Here’s my view on the key characteristics of enterprise-grade customer data management systems:

  • Their design is driven by both a solid foundation of good data management practices and a clear set of use cases across different parts of the organization. In other words, they’re not just repositories, they’re also tools that flow seamlessly into all of the enterprise systems that require customer insight to operate effectively.
  • They can adapt to support a wide range of needs and uses, beyond those that we have already identified today.
  • They can accommodate both deterministic customer data (from systems of record) and probabilistic customer data (predictions based on observed behavior)
  • They support different cadences, from real- or near-real-time to periodic, depending on the use case.
  • They provide the ability to identify customer segments that meaningfully inform the most effective ways to interact and engage with individuals—regardless of the type of interaction (e.g., marketing, sales, service). That’s based on more than just analyzing behaviors from a marketing perspective.
  • They ensure that personally identifiable information (PII) is only used where it needs to be and where it’s permitted, whether it’s a question of regulatory requirements or customer preferences.
  • They draw from, coexist with, and feed into the tools that different teams use to get their work done.
  • They help to establish the metadata structures that will improve effective data analysis and use over time.
  • They’re straightforward to use across departments and functions.

This may all sound as much like a fairy tale as the promise of CDPs. The good news is that many companies and a whole lot of sharp minds are working on the problem. Those attempting to tackle it range from startups to established players (some of whom are already in the, ahem, CDP category) to large enterprise technology providers.

Part of the challenge is semantic. The term CDP has become muddied by undelivered promises and limited understanding of the need. Many of the more recently announced offerings in the area of customer data management have been unofficially described as “CDPs that aren’t just CDPs.” These are much more than just data management systems, however. They aim to provide intuitive tools for deriving and applying customer insights, not just consolidating customer data.

The pace with which players in this fast-evolving space are tackling the dual challenges of data management and applied use gives me hope that we’ll soon have some much more effective solutions to putting customer understanding at the heart of business operations. While I suspect we’ll never get to one single system for all customer data, there’s every reason to believe in practical solutions to this large-scale problem in the near future.

In the meantime, do some serious tire-kicking before you invest. Define the use cases that are most crucial for your organization to address immediately. Push hard to understand what a vendor's offering includes today, out of the box, and what's on the near-term roadmap.