There is an expectation from consumers today that the brands they conduct business with know who they are. Whether interacting online, in-store, via a call center, on social media, an app, or any other channel, consumers trust that a brand knows something about them – their likes, preferences, behaviors, favored devices, etc. It’s something of a litmus test for brands to demonstrate they value a customer beyond the transaction. The consequences for getting it wrong are steep. It’s likely the brand introduces friction into a customer journey through an irrelevant offer, or just by being out of step with the customer. The customer, realizing that the brand doesn’t really value the relationship, will find a brand that does.
Consider the findings from a 2021 Dynata survey commissioned by Redpoint, where 78% of consumers said that it is frustrating when a brand’s communications and marketing messages are inconsistent depending on which channel they visit. Furthermore, 80% of consumers said they are more likely to shop with brands that demonstrate a personal understanding, exhibited through relevant, personalized offers.
The research perfectly captures why data lakes and even basic customer data platforms (CDPs) are insufficient for keeping up with the customer through dynamic, omnichannel, non-linear and non-sequential customer journeys.
Before exploring the reasons why data lakes and basic CDPs fall short, it’s important to note a key distinction in that data lakes, which rose to prominence in response to the hyper-structured and summarized world of data warehouses, are a sandbox environment for data scientists, analysts and statisticians to build predictive models and work with large volumes of granular data. Whereas even the most basic CDP is very much a production system that must manage high-velocity audience queries, receive continual data updates and support real-time interactivities.
But when Hadoop and relatively inexpensive storage helped data lakes rise to prominence, a misconception grew that because data scientists were using them as creative testing environments, that they were somehow on an equal footing with a CDP in terms of an ability to develop a perfectly manicured customer profile.
While a data lake might be sufficient for basic segmentation and other analytical activities, a key limitation is that they were never intended to orchestrate highly relevant omnichannel experiences at scale, relying as they do on an outdated concept of a code-centric approach to data warehousing. When data must be processed using coded ETL’s as with Databricks, data in a data lake goes stale quickly, creating inertia that resists large amounts of daily transactional updates.
Because a personalized experience requires moving with the customer in real-time throughout a dynamic journey on both digital and physical channels, data quality processes must be embedded at the point of data ingestion. Anything less, and data will be unfit for purpose, yielding poorly executed identity resolution, minimal behavior information and a dearth of transformations (e.g., a year-over-year change in spend).
Ease Stagnation with a Single Point of Control
Limitations of a data lake for providing a differentiated, personalized CX at scale is why data-driven organizations are focusing on robust, dynamic CDPs that solve for the data “stagnation” problem by providing access to an enterprise-class dynamic data repository that continuously refreshes and links the totality of an organization’s customer data in a central hub with a single point of operational control.
Right off the bat, enterprise-grade marketing software that offers a single point of operational control eliminates the data latency and marketing program complexity familiar to organizations with a marketing technology stack comprised of standalone packages that were acquired and integrated over time.
Conversely, a dynamic CDP embraces the data challenge. Within seconds of data ingestion, an enterprise-grade platform should provide cleansing, matching, de-duplication, governance and data mastering processes – everything needed to make data fit for purpose and usable the moment of arrival. A golden customer record links together all the proxy identities for each possible customer (known and unknown) and provides a robust long-tail of transactional information that includes everything from granular behavior to KPIs to transformations summaries – essentially everything there is to know about a customer, ingested, processed and updated in milliseconds as the customer moves throughout a customer journey.
A platform with inline automated machine learning enables marketers to deliver personalized messages at scale by eliminating the need to build offline, coded models that go stale over time. Once in production, models become outdated as soon as there is new data or as soon as the business decides to optimize a different metric. Embedded automated machine learning applied to a consistently updated golden customer record ensures a hyper-personalized experience at scale that is always in the cadence of the customer.
This is a stark difference from another common personalization technique, using client-side personalization tools that build web-based customer profiles. While a profile may indeed be built off an enormous volume of web behavior, it lacks access to other data sources and therefore likely does not provide a complete picture of a customer. To compensate, marketers link building blocks of tools that provide access to other data which is neither scalable nor able to provide repeatable results. This approach may be fine for email blasts and light personalization on a brand’s website, but it is not enterprise-grade marketing software.
An enterprise-class customer data management platform creates the single point of operational control for all channels and messages, eliminating fragmented communication that is distributed on a channel-by-channel basis and limited to a local scope of data. Marketers have a single platform for controlling the utilization and orchestration of all channels, strategically engineering a holistic engagement journey for individual customers at scale.
From the customer’s perspective, consistency across channels demonstrates a deep, personal understanding of the individual customer’s desires, wants, and needs. It is as if a brand is speaking to the customer with one voice, regardless of the channel or touchpoint. As the Dynata survey shows, customers reward omnichannel relevance and a consistent experience with loyalty, which is why brands interested in differentiating on customer experience are transitioning to robust, dynamic CDPs that offer a single point of operational control.
