EP #5: David Raab on the Role of CDPs in Delivering Personalization at Scale

EP #5: David Raab on the Role of CDPs in Delivering Personalization at Scale

About this Podcast

As part of our #PersonalizationThursdays series where we discuss the impact of personalization and predictive recommendations across key industries, we had the honour of hosting the leading Martech guru and Godfather of Customer Data Platforms (CDPs), David Raab! David answers the following about CDPs:

  • What does it mean?
  • Why do businesses require them?
  • How are they implemented?
  • How much time does it take to implement?
  • Who are the key internal stakeholders involved in its implementation?
  • How do they help delivering AI-led personalization at scale?
  • Which companies are successfully leveraging this capability?

A super-insightful episode that every marketer, product manager, and data scientist – across industries – will find valuable!

Episode Transcripts

Pradyut Hande (Host): Hi guys! and welcome to a brand new episode of the MartechnoBeat. A specially curated podcast series powered by Netcore Smartech, I am your host Pradyut Hande and today I have with me a very very special guest ‘the Godfather of the Customer Data Platforms’ the one and only Mr. David Raab.

David Raab: Thank you for having me, welcome, glad to be here.

Pradyut Hande: So before we begin, I would just like to give our listeners, not that you need any introduction but just letting them know that you are the founder of the ‘Customer Data Platform Institute which is a Vendor Neutral Educational Project that helps marketers build unified customer views that is available across the company and moreover you have been a market leader, a pioneer leader in this space and a consultant in the marketing analytics and vendor assessment space for over 32 years, sir it’s a real honor and pleasure to have you today and I am sure that we are gonna have an extremely insightful conversation.

David Raab: Great! 

Pradyut Hande: Alright, so just to lay down a little bit of context now with regards to CDPs, you know in 2018 the industry’s revenue scaled about 740 million dollars, now that just goes to show that the entire concept over CDP has gone from just merely being a buzz word to the need of the hour for businesses across industries especially with revenues in 2019 likely to exceed 1 Billion Dollars and with the fact that CDP has now been cited in Gartner’s hype cycle for Digital Marketing and Advertising 2019, it just goes to show that you know marketers and businesses at large are taking this far more seriously. So one thing that I have noticed in my conversations with marketers across the board across different geographies is that the definition or the meaning of a CDP means different things to marketers across industries so if you could just explain what a CDP means? What does it do? And why is it required by a business and if you could do that explanation to a five-year-old, how would you do that? 

David Raab: We actually did a blog post once where we do in terms of different kinds of vehicles so, CDP was actually an ice cream chuck because it has a lot of different things and different people chase after it and they are all delighted to see it as it is supposed a data warehouse which was a school bus which had the same information it was carrying and I think the data link was a dump truck so if we really had to explain it to a five-year-old we would probably use it those terms but in slightly more grown-up terms what a CDP does is – it pulls your customer data from all the different places where it sits, where its gathered which don’t usually talk to each other.

Pradyut Hande: ‘Right’.

David Raab: And puts it in one place so that then you can all use it together to get a complete picture of each customer and not just a fragmented picture that any individual system would have and then it shares that information out to any other system that wants to use it back to our ice cream truck, all the little kids chasing after the ice cream truck that needs to get particular kind of ice cream that they are interested in. So like differs from as I say something like a data warehouse which only has certain kinds of very structured data and actually other big difference is that CDPs packages you can buy or build a warehouse data typically from scratch so you know some fundamental differences that make it easier to pull together because its package software and deploys more quickly and its more flexible because it’s designed to connect to other systems both in terms ingesting data from any source and then pushing out data to any source so it gives much more flexibility than companies have with the way their data is organized in their systems.

Pradyut Hande: “Right”, that makes a lot of sense and I mean a lot of businesses that I interact with, one of the key questions that they ask is how much time does it take to implement a CDP from scratch?

David Raab: Well, again its a package software so the actual installation is a matter of days now, the implementation takes longer because you have to decide where your data is going to come from and then you have to assess the data that is in those source systems and see if it’s actually in the format your file is, how is it to used you have to do a certain amount of data modeling although it depends upon the CDPs, some will actually just load the data in its original format and then you can decide after the fact about how to extract a new structure or read a supposed structural load, so you know realistically you might get a CDP up and running in several months, it could be six months to a year but it’s not an issue the technology because you are not building anything so that would be a custom projects like a data warehouse or a data like that this is usually two or three years so it’s way faster.

Pradyut Hande: ‘Correct’, and you are also accounting for the level and degree of integrations required or the CDPs to be ingesting that sort of relevant data.

David Raab: Alright, I mean in some of CDPs, most of the CDPs have prebuilt common data sources so they are all gonna connect to your big e-commerce platforms, your big sales automation systems, and so on. So that again is just a matter of literally pushing a button, but if you have to build the custom connectors that can be a matter of easy days or weeks, not months but again depends upon the situation.

Pradyut Hande: Okay, and who would you say would be the right person or key stakeholder to be driving the implementation of a CDP at a business?

David Raab: We see a few different person that is involved, I would say that the most common buyers are sill the directive level marketing people so the head of the digital or the head of the email or the head of the advertisement sometimes somebody of that level, sometimes there is a marketing technology staff within marketing that reports to the CMO. And they might hire someone who once they sense what’s going on in the company, they quickly realize you know we lack CDP, we need one and they thrive. The other things that we are seeing now is more are corporate-level people so possibly chief data officer or chief analytics officer or the chief technology officer, the chief information officer who wants to drive not a marketing-driven product but a corporate level project because we see continued and expanded interest in custom data throughout the organization and now with regulations, we see a lot of need for the compliance people to make sure that customer data with the corporations is handled correctly. So that pulls the responsibility away from marketing backup to more of the corporate groups. 

Pradyut Hande: “Right” that makes sense given the fact that you know from a buying perspective there are greater sweet level folks involved in the decision-making process.

David Raab: ‘Yeah’ they are always going to be involved, they are usually the ones driving it, they are capable of that. 

Pradyut Hande: And setting this background we have now entered 2020, CDP has been one of the key buzzwords over the last couple of years. What are some of the Industries that you see where adoption continues to rise?

David Raab: I think it’s rising across all the Industries because it’s not really fully adopted anywhere, the initial adoptions were clustered primarily in retail and in media publishing industries and there were a couple of reasons for that but we think primarily because these industries have a large number of frequent low volume transaction customers, there is a lot of data that can very clearly benefit for better personalization and segmentation which is what CDP ultimately makes possible. Now in a couple of last years, we have seen more in some of the other industries like Finance, Healthcare and Insurance, Telecom, Travel those are all kind of rising industries right now for CDP where you know fewer larger transactions specialized data is very important customer relationships still adds to personalization we might say in the retail situation.

Pradyut Hande: “Right” that makes a lot of sense. You mentioned that adoption is seeing a rise in the financial services industry, it’s interesting you mentioned that we have got a lot of banking customers who are still using data lakes now in a situation like that how helpful a CDP would be?

David Raab: See people will often use a data lake as an input source because the data lakes essentially as you know pull the data from the data sources and put in somewhere, they put it in the data lake without particularly doing any formatting or cleaning or anything else to it so the idea of pulling the data together to create a unique view of each customer requires merging the identities across different the systems and sources of information, to look at the data like provides a place for the data scientist to pull the data in its raw form and they would do analysis and they can do that kind of data profiling that I was discussing. So CDP makes that process which data scientists are doing manually and the project becomes a more normal process as it is simply executed in the background and other systems can use that already prepared data so they are very complementary to each other.

Pradyut Hande: Okay, now that’s an interesting point you made. And that actually brings me to my next question. We spoke about true value that businesses can unlock through the adoption and implementation over a robust CDP just wanted to pick your brains on what would some of these companies or examples of some companies if you believe are leveraging CDP is optimally can be used to drive a variety of different use cases primarily the revolving around personalization at scale so which are these companies that come to mind. 

David Raab: Well, we the CDP Institute have a wire online library, and one of the things we have in the library is many case studies for various vendors so CDP Institute.org  and so we look to those statistics so as to see who is doing what. So just an example is botbox for online subscription for food they have a case study in there for the use of CDP, the customer view and then to segment Subscribers going cancel subscriptions and the period the number of experiments to find out more strategies to the messaging on the time in the promotional offers. They doubled their revenues having to have a particular segment. So there’s pork box hello fresh actually has happened another case study is a different matter but similar business doing Direct Mail up optimization where they did a number of Direct Mail Plus email campaign and they make an find the optimal approach in it at 2 X increases Advantage Rent-A-Car don’t know what that was Tiny Emily personalizing their home page based on who the visitor was where they came from where they looked at before they added 25 percent lift in cookies xanterra Cruise Lines another case that was doing automated multi-wave sure campaigns to sell cruises. Thanks. 

Pradyut Hande: “Right”.

David Raab: Industries of the companies that do this are the good ones. Of course, we’ll use the CDP for many many use cases of the whole point of the CDP is its foundational technology. Once you have those customer profiles assembled the number of uses literally infinite, so it’s not that hard to figure it out. When you start, make The Plunge in and build the thing in the first place. 

Pradyut Hande: Right, you mentioned oh how there was a customer or other company that was you know, dynamically personalizing their website and that actually brings me to my next Point essentially how CDP the fundamental thing that the CDP helps what enables a business to do is develop a 360-degree unified view of every customer and that essentially becomes the foundation for driving personalization at scale now the likes of maybe an Amazon On or Netflix that relies so heavily on their recommendation engine would employ some of this. I’m sure. 

David Raab: Yeah, absolutely one of the prime where I use cases is analytics in particular as you know does work in data science and spend most of their time on the data preparation and that is exactly the work that the CDP sizes so that they already can start with most of the day of preparation work being done and that can feed into any kind of recommendation and interpreting to algorithm or whatever a little court case that’s in the CDP for all those things you need complete data connect is exactly what the city people’s together for you so it’s a journey. All those kinds of projects are perfect.

Pradyut Hande: Another key point that has come up actually in a lot of my conversations is you know, the fact that everyone’s talking about CDP but a lot of businesses or lot of marketers think that it’s not for them or it’s for the company next door all their neighbors so to speak so what would is your take on that? What advice would you give to businesses who believe that it’s probably not for them? 

David Raab: Well, obviously I would ask them why they think such a thing, I think CDP is extremely useful for most people, there may be some practical constraints and it is certainly not for small businesses. I mean really small business. The core value is when you deploy everything down to your five-year explanation is the CDP brings down data from different sources so you only have one system that has all your data which a lot of small businesses have to do business with. Whatever their system is to do everything in particular then they do not need CDP. The new CRM system in the city because there’s nothing to unify. It’s already in one system actually are many of these cases reduce even in that situation, but by and large, those businesses are not going to do and you also need a certain level of sophistication even though many of course very simple things like customer-created houses, customer here yells is complicated but simple analysis of just pour my customers were their profiles like what kind of things they tend to buy was very simple questions in the CDP really quite easy to answer but what take it Beyond those initially use cases you want to get into complicated task that takes an at least Modestly sophisticated marker sophisticated marketing infrastructure marketing technology stack right have an email system and website and all the other systems actually do the personalization. Even if the CDP is telling you what to do, every company has those resources, but if you reach that sort of minimum level of sophistication, which I would say every midsize business today as those tools could have those tools in place. You probably would be better for us. Because very few companies provide care data and pretty much all companies can benefit from unifying the data. So maybe your company is going to suffer. So probably you want to go join your neighbor because they are going to thrive and you will suffer.

Pradyut Hande: I think that’s a very valuable point that you made and I’ve also come across a lot of marketers a lot of businesses that once they Implement CDP. That’s when they’re able to unlock their true value. That’s low. That’s when they are able to identify use cases that CDPs can solve for them. But before they get to that stage, you know, as I mentioned the general attitude is that this may not be the right time on this may not be the right fit for me. So yeah, valuable points that you made that bring me to my next question actually. We see a lot of these online startups predominantly apps and websites that are online first brands that are happy to adopt a multi-channel marketing Hub or a mobile marketing platform. So they don’t really have a standalone CDP involved in the entire equation. How do you see that trend changing or remaining constant? 

David Raab: Well, if they are only and have only one channel in my earlier point about CDPs being very dynamic and ka. I really only have one chance to integrate across multiple systems. We’re doing everything in a single system. They just are an eCommerce platform, then again either CDP is not necessarily all that great. But if using a multichannel market hub working up, for example, presuming that you are a multichannel Marketer, except we’re using that so you will have multiple channels and the value of the CDP, in that case, is that multi-channel marketing up typically well-combined data management of the CDP. So basically it does have the CDP inside of itself with the delivery system. So it actually pushes a message out to the ink does it only now do it all web personalization? So that’s one private. Is that one product great for you? That’s a great email system and a great web personalization, a great call center and great CRM, you know I’ve ever given every feature of that is absolutely everything you need right then you do not need a CDP. If you’d rather use some other email platform or some other web personalization or ingest some other recommendation engine. It does a little better than what’s built-in now. Many of those multi-channel marketing Hubs don’t really have the setup to share that data with other systems and I think that’s just as English has CDP from a system that assembles stated just for its own uses. A lot of people, a lot of systems actually just ingest external data but then they aren’t really sure how to expose it. So if you’re in a situation where you’re not convinced that one product is going to do everything for you then you might want to use some other product for someplace in the processing. That’s where CDPs are going to add value. So we expect to see more and more because the reality is very few companies are really confined to one system that does everything they need as well as to be done. 

Pradyut Hande: So, I mean, it’s you know, the entire strategy is about optimizing your market stack, and rarely would you find a full-stack solution that does everything to the best of its ability or to the best of your expectations? So it’s a question of optimizing the entire stack. 

David Raab: ” Yep!”

Pradyut Hande: All right. So on that note, I’d like to know the end by asking my favorite question, you know, we spoke about how critical CDPs are to businesses that have achieved a certain degree of scale and are chasing personalization at scale. We spoke about how you know, who would be the key stakeholders driving the adoption and implementation of a CDP and the timeline involved in its Setting up and so we spoke about multiple aspects of a customer data platform and how valuable it is to businesses across Industries. I’d just like to know your thoughts on where the future lies when it comes to CDPs. How do you see this entire space involving? What changes do you foresee your thoughts on? 

David Raab: Well you know it’s a very dynamic space and we see a lot of changes that are continually happening and what we see more than anything else is broader diversity among CDP, classes of customer databases shared among all the systems, that is defined but there are also a set of functions that if I persist in customer database year, but other systems are our official definition but is also certain functions, sharing that data, right? So what we see more is we see some systems that are two CDP systems that that’s all they do is they build the database right? And then we see other systems that offer the CDP plus analytical capabilities and loss of orchestration capabilities, even in some cases. Its delivery capabilities are more all along than the multi-channel marketing up bottle that we were just discussing except that being a CDP actually has the openness to share its data with an external system so it doesn’t lock you in. I can point to literally dozens of vendors in each of those categories and now we also see more vertical specialization. So CP specifically for travel specifically for healthcare specifically for retail, what makes them specialize in those is that they have connectors into the operational system. So it’s a ticketing system for travel or internal medical record systems for health care and so on so there’s a lot of growth in vertical-specific CDPs.

Pradyut Hande: “Right”. 

David Raab: So the more amazing thing that we see is the capability which is that the really high-profile now being embedded in more and more systems that are doing other things in addition. So it’s unfortunately very confusing for the CDP buyers because as you all are CDPs which one do I pick? 

Pradyut Hande: ‘Correct’

David Raab: The reality is, of course, once you know what your needs are? Once you know what industry you’re in? What do you know? What is your scale? Once you know the scope of functionality you want, you just want the data better. Do I want the analytical bits and the campaigning bits as well? Then he quickly kind of Narrows down to a more manageable number of factors, but that initial shock of looking oh my God! they all call ourselves saying.

Pradyut Hande: “Yeah” can be rather overwhelming. 

David Raab:  It is definitely overwhelming. This is definitely where I got at the Institute with a lot of work to go to classify vendors and to make it simpler to do that so you can stay tuned for some announcements, but that probably is that changes so that the notion of the CDP nobody argues anymore. Yes. I need someplace to have all my customer data unified and available at profiles and we did interact with netpal five years ago, people didn’t even realize that was an option. Oh, Kelly knows it’s an option now. They said my God Kelly. Is it a notch by hundreds of options? How do I choose what I want? That’s the second-order problem. It’s a better problem to have not knowing you have a solution available at all. So we’ll take that one next.

Pradyut Hande: Now that makes a lot of sense and you certainly shared a lot of very very valuable insights that I’m sure our listeners are going to benefit from. On behalf of Netcore Smartech, I’d like to thank you once again for your time David. It’s been a pleasure hosting you.

David Raab: It’s been a pleasure to chat!



Unlock unmatched customer experiences,
get started now
Let us show you what's possible with Netcore.