Almost a year ago I wrote an article, Customer Data Platforms – a Contrarian's View, that garnered negative reactions from a prominent customer data platform (CDP) practitioner.
In their 2018 Market Guide for CDPs, Gartner had highlighted the significant attention CDPs were receiving and indicated that these technologies were nearing the “Peak of Inflated Expectations,” suggesting that their popularity stemmed from the ongoing struggle to unify customer information and use it to improve marketing.
About CDPs, I wrote: “While the potential benefits are significant, these applications are by no means a silver bullet. Despite (or because of) the publicity surrounding CDPs, misconceptions and an incomplete understanding of the possible issues can quickly turn the ‘Peak of Inflated Expectations’ from a catchy analyst phrase into an expensive reality.”
Not much has changed since I wrote that article. So, today I am doubling down on some of my past impressions, and it seems that in terms of cautionary advice, I’m not alone.
Gartner cautioned in the aforementioned 2018 guide that a slide into the “Trough of Disillusionment” was imminently possible. Some of their reasons for these conclusions included:
- A rapidly changing vendor landscape rife with venture capital infusions, acquisitions and rebranding of existing capabilities to fit the perceived market momentum.
- A wide variance in product capabilities and features offered by vendors branding their products as CDPs.
- A nascent implementation landscape with many planned implementations, but few examples of documented ROI to-date.
- An implementation topography largely comprised of mid-market organizations, furthering questions about how well the CDPs can handle the complexities of large organizations.
- A potential clash around technology versus marketing-managed in terms of how to overcome integration complexities and account for processing essential business information.
- A need to reconcile overlapping capabilities between CDPs and existing tools in the martech stack.
Free webinar on data privacy and customer experience
Are data privacy and customer experiences an oxymoron? Your customers don't think so. Watch this free ondemand webinar to learn how brands can meet and even exceed their goals in the age of data privacy.
This is not to say that the demand for CDP capability is manufactured. It’s not.
Marketers have a very real need to corral customer data currently residing in disconnected silo’s both inside and outside the organization. In an HBR survey on using real-time analytics to improve customer experience, the top challenges marketers faced were legacy systems, data silos and multichannel complexities.
In a Forbes Insights study on the rise of CDPs, only 1 in 5 executives surveyed considered their companies to be leaders in customer data management and only 13% believe they fully utilize customer data. These difficulties persist despite the fact that we’ve been trying to uniquely identify customers and consolidate first-party customer information since the late 1980s.
As I highlighted last year, these systems have undergone many incarnations, but today we still see customer-oriented data warehouses and data lakes facilitating analytics and master data management applications facilitating operational activities. The on-going excitement around CDPs clearly indicates that these existing applications are not solving marketers’ problems.
Marketers have a very real need to corral customer data currently residing in disconnected silos, both inside and outside the organization.
Making informed decisions
So how do marketers avoid the “Trough of Disillusionment”?
Step 1: Cut through the confusion created by the wide array of vendors and capabilities currently in the space.
In their recent report on the myth versus reality for CDPs, Winterberry Group highlights the fact while more than 100 vendors call their technology CDP, fewer 20 really fit the criteria defined as follows:
- Ingest and integrate customer data from multiple sources.
- Offer customer profile management.
- Support real-time customer segmentation.Make customer data accessible to other systems.
This doesn’t mean that you shouldn’t look at vendors who don’t fit exactly into these criteria; but it does require a very clear definition of the problems you are trying to solve with the CDP.
If that problem centers mainly around integrating customer data into a unified and persistent source, enabling the management of that information, and making it accessible to other applications in the existing martech stack, then it’s essential that the product you select really does all these things.
Gartner lends urgency to this requirement with their discussions on the varied backgrounds of CDP vendors; operational data management vendors; marketing personalization vendors; and pure-play CDP entrants -- each of whom carry inherently different strengths and weaknesses.
Step 2: Dispel the notion that you will be able to implement or manage this technology independently of IT.
While the system may be implemented with outside technical help, e.g. the vendor or 3rd party systems integrator, the likelihood of succeeding without on-going IT involvement is very small.
Having spent most of my early career helping companies build customer profile systems of various types, I can definitively say that the robust customer identity matching solutions required to make a CDP accurate always require tech savvy. They are not solutions that businesses can drop in and forget.
Information quality degrades over time, systems generating customer information change, and the business rules regarding how and when to merge duplicate customers are never static. On-going support will be needed to ensure that the CDP stays in sync with existing customer systems. Other solutions providing customer profiles (e.g., MDM) are rarely, if ever, managed by a business unit rather than by IT.
Furthermore, the volume and complexity of applications that will feed to or receive information from the CDP make it virtually impossible that out-of-the-box APIs will work in every instance, especially in the case of the myriad of legacy applications that generate much of the data a CDP will require.
One very positive change in industry thinking around CDPs from last year to today is reflected in this area. The Winterberry report stipulates that the industry is moving away from “marketing-managed” as a criterion for a CDP and does not include it in their definition.
Marketing-managed has been dropped from the CDP Institute description. And, Gartner speculates that one possible future for this technology is that it could even be subsumed into the more technical data management space.
Step 3: Clearly delineate between the capabilities of the CDP and the capabilities in your existing martech stack.
Unsurprisingly, given the varied backgrounds of participating vendors, CDPs boast many additional capabilities to the core criteria defined by Winterberry Group. These can include audience messaging and testing, householding, next best optimization, journey analytics and discovery, and orchestration engines. Sophisticated marketing groups will already have many of these and duplicating them in a CDP will simply cause additional confusion.
Making the right decision about your CDP
So, while the CDP is a wonderful idea – designed to solve some of the very real problems that marketing encounters around unifying customer information, they also require thoughtful consideration and careful planning. Understanding exactly what problems you really need to solve and matching the solution to those problems; gauging clearly the amount of IT support you will need and securing it before purchasing; and mapping how the CDP will fit with existing applications (both integration and capability overlap) can help to ensure that you make the right decision regarding a CDP.
- 3 ways to rethink retail forecasting and demand planning The pandemic has profoundly changed consumer shopping behaviors and experiences and the increasing pressure has retailers scrambling to improve their ability to precisely predict and plan for demand. If you don’t know where to start, here are three questions to ask as you rethink your forecasting and demand planning.
- How to improve your AI marketing skillsMarketing teams can use current AI capabilities to enhance their efforts around campaign automation, dynamic pricing based on forecasting models, and by providing more relevant, real-time customer offers.
- How machine learning can provide customers with more relevant offersIn this video, industry influencer Daniel Newman offers his perspective on ways machine learning can be used to enhance customer experiences.