David Raab is known to the MarTech world as the man who coined the term Customer Data Platform in 2013.
As the founder and CEO of the CDP Institute, Raab has been a leader in exploring and explaining how marketers can use CDPs to solve today’s most pressing marketing challenges. He is a marketing technology analyst and a consultant who has helped hundreds of companies to find their right systems for their needs.
Raab is one of the keynote speakers at the second edition of Vibe Marketers Fest taking place in Dusit Thani, Dubai on November 5 and 6, this year. Ahead of his much-awaited workshop, MarTech Vibe caught up with him over a quick email interaction.
In this exclusive interview, Raab speaks about the impact of AI and machine learning on marketing, digital transformation, tech acquisitions, and the future of CDP.
Data-driven marketing has been a norm over the years. Do you see AI and ML taking this approach further? How?
AI and ML can make many more individual-level decisions. Human marketers can at best create rules that specify how to handle different situations, but such rules can only handle a limited number of options. For example, rules to pick the best message for a customer would be based on a limited number of segmentation variables and assign a particular piece of content to each combination. An automated system can look at dozens of variables in combination and can match that to a near-infinite number of content options without a human needing to classify each piece. There are similar applications to refine message frequency, timing, channel selection, and other decisions. At the extreme, an automated system will simply find the best action and take it, without even explaining to humans why it made its choice. Of course, that is dangerous, so humans must have ways to monitor the automated choices and to apply constraints that avoid any serious errors.
Digital transformation is one of the primary concerns of businesses in the Middle East. What would be your advice to them, in terms of dos and don’ts?
There are some new technologies involved in Digital Transformation, but the real issues are the same as with any major organisational change. Transformers must start with business strategy to ensure they are making the right changes. They must consider organisation, process, and training issues to ensure that changes are deployed effectively. They must measure their results to see what changes are really taking place, to understand their impact, and to make improvements over time. And, of course, they must put the necessary technology in place.
What should attendees look forward to, from your presentation at the Vibe Marketers Fest this year?
I will be talking about How Not To Pick Marketing Technology. We will discuss the importance of proper technology selection practices and common mistakes to avoid. Specific topics we’ll cover include:
- How to select a marketing system that matches your business strategy
- What customers do and don’t want from technology-enabled marketing
- Building a MarTech architecture that’s both stable and flexible
- Prioritising your MarTech decisions to ensure they deliver real value
- Rewards you can expect from making the right MarTech choices
What are the things CMOs should keep in mind before investing in a CDP?
They need to understand how they will use the CDP, which means defining specific use cases they cannot execute now and expect the CDP to make possible. This will determine what features they need – which is especially important because CDPs vary widely in the scope of features they offer beyond the shared ability to build, store, and share unified customer profiles. CMOs also need to ensure that other systems and organisational resources can support their target use cases. Otherwise, they might buy the perfect CDP but still not meet their goals because the necessary data isn’t available, because their messaging systems cannot accept the CDP outputs, or because their staff doesn’t know how to use the CDP information.
What is the future of CDP? What do you see it evolving into?
We see CDP as being recognised as a required function for every organisation, similar to a CRM, order processing, or accounting system. Some companies will buy a specialised CDP that only builds the unified customer database. Others will buy a CDP with data, analytics, and campaign functions. Still others will buy a broader system such as a marketing cloud or ecommerce system that includes CDP as one of its features. We also see more use of customer data outside of marketing, in customer-facing functions such as sales, order processing and customer service, and even in operational functions such as retail store management. These all have a major impact on customer experience which is what ultimately drives business success. Any of those systems might include CDP functions: what matters is making those functions available, not which kind of system delivers them.
Why do you think tech giants and marketing cloud vendors like Adobe and Salesforce jumped onto the CDP wagon? What impact will this leave on the already crowded CDP solution provider field?
The tech giants are offering CDPs because their customers are asking for CDPs. CDP is unusual among marketing technologies because its growth is pulled by marketer demand, not pushed by vendors who invent something clever and then try to convince buyers they need it. Once the tech giants realised the demand was not going away and that their current products would not suffice, they had no choice but to develop a proper CDP capability. We think their entry will greatly expand the market but still leave room for vendors who specialise in different niches such as large enterprises, small business, specific industries, and different regions. Specialised products can compete effectively with the big tech vendors within their niches.
What are some of the myths around the CDP solution that you’d like to bust?
One is that CDP will by itself solve your data problems: the CDP can do some data cleaning and matching, but it cannot create quality data from scratch. Another is that CDPs can succeed without involvement by corporate IT departments: although the CDP greatly reduces the demands on corporate IT, the IT group still controls access to company data. Finally, let’s bust the myth that CDP is such a broad label that nearly any system can qualify as one. The CDP Institute has a rigorous definition of core CDP requirements including the ability to ingest data from all sources, to retain the original details, to store the data indefinitely, to build unified customer profiles, and to let any system access those profiles. Many companies calling themselves a CDP don’t meet those standards. They are purposely defined in simple enough terms that any marketer considering a CDP can easily determine whether a particular system is really qualified.
How can marketers push the envelope when it comes to what their CDP can do for them?
Most marketers will start with simple CDP applications such as analysing their customer base or understanding the customer journey. They then graduate to using the CDP for predictive modelling and, ultimately, to support the outbound campaign and real-time, cross-channel interactions. That final level of maturity is still quite rare, in good part because it takes so much change to existing systems and organisational structures. But pushing the envelope isn’t really a goal in itself – most marketers will get the most benefit from simple solutions that were not possible before the CDP, such as removing customers from retargeting campaigns immediately after they make a purchase, or giving telephone agents a real-time view of each customer’s recent Web site behaviours.
Tech start-up acquisitions have also been on the rise in the past couple of months. Do you see this as a consolidation or an expansion of solutions?
Acquisitions are always happening. They are usually a way to add capabilities more quickly than the acquirer could build them for itself. Some also add a significant number of new customers for the buyer or let it enter a new market segment. Occasionally the purpose is to remove a current or potential competitor. With CDP in particular, the big tech companies seem to be building their own systems rather than making acquisitions, although I wouldn’t be surprised to see some small acquisitions of firms with unusual CDP technology such as handling very large data volumes, using AI to incorporate new data sources, or identifying customers across channels. Some of the recent CDP acquisitions have been more about getting staff with rare skills, such as predictive analytics, than with adding the products of their companies.
How does a CDP help address customer data fragmentation?
Building a unified customer database is the core purpose of the CDP. It addresses fragmentation by bringing together data from multiple sources into a single system, by storing that data so identifiers that change over time can still be linked to the underlying customer, and by enabling other systems to access a consistent central database rather than each system building its own partial customer view.