One performance metric commonly touted in ads for sports cars is the acceleration speed from 0-60 mph. The ads, however, fail to mention that the horsepower comes with trade-offs. If you’re flooring the vehicle every time you leave the driveway, you’ll pay more in fuel and put a lot of unnecessary wear and tear on the tires and brake pads.
It’s analogous to some customer data platform (CDP) vendors who tout data ingestion speed as a definitive performance metric without telling the complete story about what’s under the hood, as it were. If one performance metric – ingestion speed – shortchanges other features, the result could compromise the intended use case, namely using customer data to provide a real-time, personalized customer experience (CX) across an omnichannel journey.
Know a CDP’s Features – Inside and Out
To avoid finding out about limitations at an inopportune time down the road, organizations should read the fine print. A proof of concept (POC) is the time to put data through its paces, looking beyond ingestion speed as a defining metric to understand the intricacies of a CDP as an engine for digital transformation.
One hidden cost of many CDPs that promise fast ingestion speeds is that customer data often remains unstructured until query, at which point a data model is applied. At first glance, quick data ingestion might seem to be an attractive feature but the downside, of course, is that by withholding structure until query the data model will be different for different users based on their security access, user access, or intended use. This will not only create inconsistencies in the data by returning different audiences, it also creates a runtime issue for the obvious reason that having to structure customer data at runtime means a longer runtime, adding to the overhead of resolving the query.
No Data Structure, No Single View
Apart from the longer runtime, returning different audiences is antithetical to everything a marketer is trying to accomplish in terms of engaging a customer with relevant, personalized experiences. The lack of audience consistency throws a wrench into this effort because you can’t be sure that the audience being extracted is truly representative of the latest data.
Many CDPs promise some version of a “single customer view.” It may be called by different names, but a core tenet is the ingestion of all customer data – every type and source. But a failure to match data until point of query makes a true single customer view impossible. When a momentary data structure collapses after a query, there is no precision. Some CDP’s may use up all their horsepower to quickly go from 0-60, but the trade-off is that performance suffers because they are then not able to match the data upon ingestion at scale and speed. To compensate, they let customer data accumulate and only take the latest data to run through the data model. Using the latest data may present ‘a’ view, but certainly not a true single customer view, the 360° unified profile that is updated in real time to provide a complete customer record.
Another consequence of failing to apply a data model until query is a lack of accountability. It creates an enormous data lineage problem, particularly worrisome in industries such as healthcare and financial services where data lineage is under constant audit. An organization unable to verify how it has used its data over time will face compliance barriers on many fronts; in addition to potential financial penalties for a failure to comply with GDPR, CCPA and other regulations, the company may also face a loss of customers whose preferences have not been met, among other trust violations.
Persistent Keys and Data Match Consistency
A second hidden cost from many CDP’s that will also create inconsistencies is with regard to data matching without persistent keys. When keys change with every match, a data match accomplishes little more than a slice-in-time view of a customer, household or audience. There is simply no connection between a match one day and one the next, or minute-to-minute, hour-to-hour or week-to-week. This makes it impossible to track key structure changes over time. Operationally, it means not knowing, for instance, whether a household composition has changed over time. It also forces a company to do a complete re-mapping of accounts to the new keyset every time they match the data.
One major drawback to this approach, aside from the time and effort having to re-map accounts with every data match, is that it handcuffs marketers from providing a next-best action for a customer that is contextually relevant to a customer’s journey. If a customer is marketed to as head of household based on a data match one week, but has since finalized a divorce and established a new residence, any engagement is likely to be irrelevant to the customer’s up-to-date profile.
To ensure consistency and accountability, the Redpoint rg1 solution immediately structures data at ingest – providing precision with matching results for a particular query. Accountability is ensured with persistent keys that augment and change over time, providing a clear picture of what is happening to a population over time, including the changing dynamic of relationships and contextual history.
Not all CDPs are the same. If your company is exploring how a CDP can power a digital transformation and engage with customers with relevant, personalized omnichannel experiences, it’s important to cut through any confusion or uncertainty about flashy features that may look good right out of the gate, but will soon leave you on the side of the road.
Related Content
What is Data Lineage and Why is it Important?
Why Data Veracity is the Foundation for a Personalized Customer Experience
No Data Left Behind: The Importance of a Closed-Loop Cycle for Intelligent Data Orchestration
Be in-the-know with all the latest customer engagement, data management, and Redpoint Global news by following us on LinkedIn, Twitter, and Facebook.
