David Raab Q&A: Marketing, CDPs, and Digital Transformation Post-Covid
CDPs are showing rapid growth as businesses everywhere undergo rapid digital transformation in an effort to meet customers on the channels they’re using now: mobile, digital, social, and more. The CDP Institute, headed by David Raab, has long been at the center of initiatives to promote understanding this widespread push for the use of data throughout large organizations. We recently caught up with Raab in the following far-ranging interview, which covers many of the issues facing CMOs, DMOs, CTOs, and digital transformation directors. From omnichannel marketing to coping with Covid, Raab explains the centrality of this increasingly transformational martech. (This is the second in a series of interviews with Raab; the previous interview can be found here.)
Treasure Data: How has the CDP market evolved over the last couple of years (pre-pandemic)? Is there anything surprising about how it has evolved?
David Raab: On the buyer side of the market, we’ve seen expansion beyond the original buyers in media and retail to include financial services, travel, telco, B2B, and, more recently, even healthcare and education. If there’s a pattern, it is moving from industries with many small ticket purchases and a short-buying cycle to industries with fewer, larger purchases and a long-buying cycle. Since the application of CDP is generally related to more precise personalization, it makes sense that initial adoption would be in industries where the returns on personalization (i.e., better targeting) are clear and quickly measurable. Retail and media are also two industries where industry leaders (Amazon and Netflix) had widely-noted success with personalized offer selection. They are also two industries that notably lagged in assembly of customer data compared with travel, telcom, and financial services, which meant that the incremental benefit from a CDP was higher because existing capabilities were so limited. Taking all this into account, I would say that the buyer evolution has not been surprising—at least in hindsight.
On the vendor side of the market, the evolution has been more complex. Many CDP vendors started as applications for specific purposes such as predictive modeling, customer success management, tag management, customer audience creation, or web analytics. Although the applications were diverse, they all shared the need for access to unified customer profiles. In building this capability, the different types of vendors converged on CDP functionality. They then all recognized that the unified data had other uses in addition to their own application, and ultimately that the data was more valuable than any particular application that might consume it. Those firms then repositioned as CDPs. This heritage is one reason that CDPs vary so greatly in the features they provide.
In more recent years, as the CDP concept took hold, the industry attracted even more vendors who had a data assembly capability, including many whose assembly features are more limited than a true CDP (which can assemble any type of data and share it with any other system). This has been a problem but is resolving itself as other categories gain more attention and those vendors reposition into those and away from CDP. The last few years have also seen the emergence of industry-specialized CDPs, which is a normal development as markets mature. More unique to the CDP industry has been an increasingly clear split between CDPs that focus on data profile creation for use throughout the enterprise, and CDPs that offer a broad range of marketing-oriented functions including analytics, personalization, and cross-channel orchestration. This largely reflects two different sets of buyers—corporate IT for the data-oriented systems and marketing departments for the marketing-oriented systems. This development was a bit more surprising since the general trend in software is for systems to expand their functionality over time. But you could also see it as another type of specialization.
I suppose the sheer speed and scale of CDP’s growth is something of a surprise in itself. Many industry analysts were skeptical of the category and many vendors were slow to recognize how important it was to their users.
Treasure Data: How do you think the Covid-19 pandemic impacts CDP investment and martech investment in general?
David Raab: We’ve seen continued strong interest in CDPs, although some companies may have deferred signing contracts. In general, Covid-19 seems to have caused companies to try to do more with their existing martech and to pull back from new investments. But doing more often depends on better data sharing across systems, so a CDP purchase might still go forward or at least be at the top of the list at many companies that are otherwise avoiding new martech purchases.
Treasure Data: What types of companies come to you looking for CDPs these days? What type of use cases are they typically looking for?
David Raab: As I mentioned previously, interest in CDP now spans many industries. I personally have received recent inquiries from telcos, travel companies, financial services, healthcare, education, B2B, agencies, and more. The use cases do vary by industry. Most buyers are marketers who focus on more advanced use cases, such as cross-channel orchestration of customer treatments or geo-fence-triggered messaging. Those are usually done well after initial implementation, which starts with simpler cases such as customer profiling, segmentation, and journey analysis. The reason that buyers focus on the advanced cases is those are the ones that generate directly measurable revenue improvements. Simpler use cases are more foundational; they generate insights that will eventually lead to value but the connection is harder to measure.
Treasure Data: Even though they might not have a full-fledged solution yet, how do you think the entry of the cloud providers into the space has changed things?
David Raab: Entry of the cloud providers has raised the profile of CDP, perhaps even more among the big industry analyst companies (Gartner and Forrester) than among buyers. Sadly, it has also added some confusion since each of those vendors offers a slightly different definition of CDP, tailored to their own vision and architecture. Most have made CDP part of a larger platform rather than a standalone system. This makes it harder to understand where the CDP begins or ends compared with other products they sell. Many of the cloud providers argue CDP should extend beyond creation of unified profiles to include analytics and personalization, since those are capabilities those vendors provide. This gives them an argument against standalone CDPs that focus primarily on data unification. My own position is that there are many “best of breed” analytics and personalization systems available, so those need not be part of a CDP. In practice, some companies want to buy separate components and some would rather buy a single integrated product, so it’s not a question of whether either approach is correct. But I would still argue that the definition of a CDP should be limited to the core data assembly function, and this doesn’t change because many people want a product that combines features from several different classes of systems. A knife is a knife, even though you buy a multi-purpose tool that has a knife, screwdriver, hacksaw, and can opener in the same package.
Treasure Data: How has CDP awareness evolved?
David Raab: The category was first named in 2013 and awareness grew very slowly until 2016, when it took off for reasons I have never determined. After 2016, awareness of the term and the general concept grew quickly, although there was considerable confusion about the exact definition. In the earlier period, CDP was often confused with DMP (data management platforms) even though those are really very different. To a much lesser extent it was also confused with CRM. I think awareness of CDP as a term is now very wide spread, although there’s still some disagreement over the proper definition. But I’d say that most people would understand that a CDP creates unified customer profiles. Where they might disagree with my own definition is they might not understand that a CDP should take data from all sources, not just some, and share the profiles with all systems, not just the CDP’s own analytics and personalization functions.
Treasure Data: Do you think companies are realizing the value of CDPs beyond marketing? What will it take for more companies to see this value?
David Raab: This is a definite trend. It’s actually quite common to hear of companies where marketing sponsored the CDP project and other departments then realized they could use it as well. What we’re now seeing is more projects where a department outside of marketing is the initial sponsor. This department might be corporate IT, customer experience, or analytics, which see an enterprise-wide need, often as part of a digital transformation initiative. Or it could be sales, customer success, or even operations, which see specific applications that require data from systems outside of their own group. I think this value will become more obvious as companies increasingly focus on improving customer experience in using their products, which increasingly relies on data to provide convenience and personalized service. This is the most important competitive arena for many companies today.
Treasure Data: What’s the most unique use case you’ve heard of for a CDP?
David Raab: My favorite story is an airline where the CDP was purchased by marketing, but the first actual users were the operations team. They used CDP data to find which flights had the most high-value customers, so they could position spare equipment to ensure those flights were never delayed. That information had never been available before.
Treasure Data: How do you think Google’s decision to block third-party cookies will impact CDPs? What do you see as the future for DMPs?
David Raab: Companies are scrambling to replace third-party cookies as a way of identifying ad recipients. Many of the proposed solutions rely on some type of first-party data, such as hashed email addresses or device IDs. Gathering that data is the role of a CDP. DMPs were built primarily to store anonymous cookies, so they are already suffering greatly as marketers look for alternatives. Some DMPs are reconfiguring themselves to work with alternative identifiers but the real issue is that privacy rules require consent for data sharing, which the DMPs were never designed to track. In fact, DMPs generally avoid handling any personal identifiers so they don’t need to comply with privacy regulations. But it’s increasingly possible to tie nearly any identifier to an individual, reducing the ability of DMPs to claim their profiles are anonymous.
Treasure Data: How do you envision the market to change in terms of new CDP (or CDP-like) entrants, acquisitions and companies leaving the market?
David Raab: Everyone expected the big cloud providers to enter the market by purchasing existing CDPs, but they have instead generally chosen to build their own. Instead what has happened is companies like Mastercard and Dun & Bradstreet bought CDPs to extend their capabilities to serve their clients. The companies that got purchased tended to be mid-size CDPs, while the largest CDPs are still independent and the smallest CDPs are either hanging on in specialized markets or quietly fading away. New entrants now tend to be vertical specialists—often with a regional focus—or companies with delivery systems, such as email, who are backing into CDP to broaden their footprint. The net result is the number of CDPs continues to grow.
There’s also a chance that the cloud platforms themselves—Google Cloud, Amazon Web Services, Microsoft Azure, etc.—extend their capabilities to become de facto CDPs. This would require supporting services for agencies or consultancies but those could easily arise. In some ways, that’s the biggest threat to standalone CDP companies.
Treasure Data: How is the CDP market growing in Europe and Asia? How is the situation in those regions different from North America?
David Raab: The European market is growing faster than North America. Europe-based CDP vendors are much smaller on average than their North American counterparts and the European firms are more likely to offer the full set of data, analytics, and personalization functions. Many of the European vendors grew out of marketing agencies which already offered database building services. Some of the small European vendors sell in a single national market, although most sell across borders. The French market is more isolated than the others but even there we see increasing integration in both directions: non-French firms selling in France and French firms selling elsewhere.
The Asia Pacific market is quite immature. We do see a number of regional firms, often concentrated in telecom and financial services. There are some signs of growth but things are moving slowly, even in percentage terms. I do expect that to change fairly soon, as we see a great deal of interest among Asian marketers in CDP and that will likely start to translate into action. The Asian vendors are more similar to European vendors in having a broad feature set and often growing out of agencies or consultancies.
Treasure Data: Do you know of examples where CDPs are being used to re-architect new user experiences to replace the face-to-face ones that are difficult or impossible now? What are some of the best examples?
David Raab: We haven’t heard much about this. Bear in mind that most CDPs work behind the scenes to support customer-facing systems. So while CDPs no doubt are helping many firms switch from in-store retail to ecommerce, the customers are interacting directly with web systems and apps.
Treasure Data: What do you think the CDP’s role will be in the typical martech stack in 5 years?
David Raab: One of the most interesting exercises you can take is to look at any decade, see what people expected to be facing ten years later, and then consider what actually happened. It’s quite humbling. Ten years ago, Facebook and Twitter were still private companies, Instagram had not yet been founded, and Alexa was far in the future. That said, I believe we can be confident that the need for unified customer data will always be with us. And I think (with less confidence) that the box holding that data will be labeled CDP on standard architecture diagrams. But I expect that box will often be part of some larger entity, rather than a standalone CDP vendor. It’s also possible that box will be more of an interface to external data held elsewhere (or in many places) than a separate data store of a company’s own data. One of the best things about CDPs is they decouple your customer data from other systems, so they create flexibility to adapt to whatever changes happen in terms of martech: new channels, new processes, new regulations, and even new customers (machines, not people). And even if you need to swap one CDP for another, having your customer data tidily stored in a single location makes the switch much easier—like copying from a tape drive for a hard drive, if you remember tape drives—than having to locate and replace hundreds of little data pockets scattered around your organization.
About David Raab
David Raab is founder and CEO of the Customer Data Platform Institute, a vendor-neutral organization that educates marketers and technologists about customer data management. Mr. Raab has a long career as a marketing technology consultant and analyst. He coined the term Customer Data Platform in 2013. To learn more, visit the CDP Institute website.