Within an organization, a “single customer view” can mean many things. Marketers, operations managers, sales teams, and service organizations might all define and use it differently. But wherever you fit, your single customer view must rely on the same complete, up-to-date, and reliable data.
Before you can have a single customer view, you must collect and centrally integrate all extant information about each customer: their behavior, preferences, interaction histories, demographics, and more. You also need to make all that information easily accessible to all the business systems and users that need it, in a form that can easily be used.
Creating the foundation for your “golden record” takes work, but it’s eminently doable. Most of the upfront work involves getting three critical items right: records, business rules, and technology. This blog post considers each of these.
The right records are authoritative: your single customer view is only as reliable as the information it’s based on.
Often, government data sources can be key components of an authoritative customer record. For example, voter registration records typically provide reliable names, dates of birth, and address fields. Even when an address proves outdated, the fact that it was accurate when entered means it can usually be reliably updated. A voter registration list can therefore be especially valuable for matching and data integration.
Example #2: Tax ID lists. These typically include names and dates of birth (DOB), and may also include government-issued identifiers (sometimes even Social Security numbers). Example #3: postal databases, which typically present complete and accurate addresses. (It’s worth noting that postal databases aren’t usable without technology that has been CASS-certified by the USPS. This approval process ensures the technology accurately validates and standardizes address information.)
The most authoritative consumer data sources are comprehensive and reliable (and, even so, they require data matching and record testing so you can remain confident in them). Sometimes, sources are very reliable for certain types of data (e.g., birth dates), but less reliable for others. For example, third-party data sources may provide phone numbers that have since been disconnected. A database of business registrations, while useful for certain tasks, would not be authoritative for general consumer data.
Internal systems that generate controlled data can often be given greater weight than other sources. So, too, you can over-weight the value of more recently-updated data sources, especially in consumer organizations where customers have recently provided their own contact, account, or demographic details. Regardless, you can use pre-set controls in internal systems to check your data’s completeness; validate attributes such as length, data type, range, and format; and to compare internal and external data sources.
Creating a customer “golden record” requires combining the right data elements across matched records: choosing some data to survive, and discarding other data as less trustworthy.
Business rules are used to make these decisions at scale, based on characteristics such as:
- Frequency: Preferring versions of data that appear more often (e.g., a name spelling that appears consistently in multiple data sources rather than the version in a single outlier source).
- Source Priority: Preferring data from one source you already have reason to believe is more reliable.
- Completeness: Preferring data from sources that are more complete.
- First/Last Valid (Non-null): Preferring data from sources that contain full and valid names.
- Adherence to data quality rules: Preferring data sources that follow specific data quality rules they make available to you.
These tests are generally straightforward. But when rules are combined, it’s easy for them to become too complex. Wherever possible, simpler rules are better. Since everyone can understand them, everyone can help fix problems as they arise. They’re less fragile, more predictable, and easier for busy data stewards to maintain.
To keep rules simple, avoid creating additional rules for especially rare edge cases. It’s tempting to seek to eliminate all errors, including those in edge cases, but business rules become more fragile as they increase in complexity. Simple rules are easier to explain and adaptable enough to function well in the vast majority of cases.
Records and business rules are easier to manage with the right technology. Migrating to a golden record is easiest with a platform that can ingest, cleanse, match, key, and compile it using any data sources and business rules: even those you haven’t imagined yet.
To achieve this, marketers need a comprehensive customer data platform (CDP) they can manage themselves. Using a CDP, they can unify all their customer data in one persistent database organized to help them orchestrate cross-channel customer engagement in real time.
To understand the value of CDPs, consider that most companies have brought multiple point solutions online, responding to the emergence of new engagement channels such as email and social media. These “walled garden” systems typically use different customer identifiers, data formats, and ingestion rates, making it even harder to unify customer data than it already was.
“Open garden” CDPs are designed from the ground up to ingest data regardless of its velocity, structure, or volume. They use advanced algorithms to resolve customer identities on the fly, and can make that data available across the enterprise almost instantly.
What’s more, you needn’t replace the infrastructure you’ve already invested in. Rather than obsoleting those systems, CDPs can make them more valuable, because they can now deliver messaging optimized to reflect truly comprehensive and up-to-date customer knowledge.
With a CDP as your foundation, you can proactively engage at the perfect moment with contextually relevant interactions that are far likelier to anticipate the customer’s needs, wants, behaviors, preferences, and intents.
Bringing It All Together
Records: authoritative information you can test and trust. Business rules for creating the best possible golden record for every customer and prospect. CDP technology to unify all your data from every source. Put them together, and it’s possible to create a single customer view you can actually use everywhere – maximizing its value every day, with every customer.