Lemnisk hosted a Customer Data Platform (CDP) Virtual Summit for the North American region on June 29th, 2022. The CDP Summit’s aim was to make enterprise marketers understand how they could create exceptional customer experiences using CDP-led hyper-personalization and increase their digital engagement and conversions. It comprised insightful fireside chats and a panel discussion with leading industry leaders. This article focuses on one of the fireside chats titled: “Notes from the CDP Battlefront” with David Raab, Founder, CDP Institute.
The key takeaways from the discussion are as follows:
Notes from the CDP Battlefront
1. CDPs have very quickly evolved into one of the most strategic tools for CMOs. Do you think business decisions are getting clearer and quicker now?
Well, I think that people do have a better understanding of what the CDP would do for them. I don’t know if organizations are moving more quickly. I think that just because of the way the world is, people are being fairly cautious in their tech investments. Not too long ago, there was a very interesting study where they asked people if there was one tool that they could keep, and which one would that be. And actually, the CDP was at top of the list of all the tools that they had.
So firstly, there’s an understanding that we’re going to be more digital across every industry imaginable. Even the ones that we thought were physical. And secondly, if you’re gonna be more digital, you need to get your customer data organized. And the only way to really do that, or at least the best way to do that is with the CDP. So the recognition of the need is I think, stronger than ever.
And then the understanding that it really is the only tool that’s designed to do that is something that people are not understanding 100%. So definitely there’s an easier sales cycle now to the CDP book. It’s still a big acquisition, an important decision for organizations, and they do tend to take their time. One of the things that we are also seeing is that the CDP now also has a very strong, strategic understanding amongst the other C-suite members such as the chief product officer, chief analytics officer, head of business intelligence, and so on.
2. Do you think a large workforce is also one of the reasons that is contributing to quicker decision-making in terms of adopting CDP?
Well, there definitely has been a shift in who buys the CDP originally. It was almost always a market. Today we see more of an enterprise view of the CDP because companies have recognized that customer data has applications beyond marketing, operations, sales, and support in all these other areas. So, what that does is it changes who the buyer is.
Because now the buyer is not a department head. The buyer is the head of IT or the head of data or the head of analytics. So they are different people that are solving for a broader range of use cases because they need to service more than one function or one department. And that makes the decision a little more complicated for them.
But it also makes it more pressing because what happens is those people are hearing requests from their constituents in all these different departments. And they see that there are multiple problems that the CDP is gonna solve and multiple problems that won’t get solved until they buy a CDP or put a CDP in place. So they are more eager to move ahead with these projects.
3. In the recent past, we saw slightly bigger discussions around CDP versus DMP versus CRM and so on. But we are slowly seeing that get cleared up a little. There are some organizations (not as digitally mature) for whom it still needs to be addressed. But others are mostly kind of getting it. Talk us through how that’s been over time.
That’s been the message that the CDP Institute has really been delivering for six years now. The whole point of the CDP is that it really is built to do this one thing. And all those other tools were built to do something else.
Your marketing automation is more or less sending emails. Your CRM systems are really helping your sales people and your service agents. Your data management platforms are really about audiences for advertising. They all have a very specific purpose and they’re all optimized in their technical design to meet that purpose.
The CDP has a broad purpose of taking all your customer data and making it accessible and shareable. Once they understand what the purpose of the CDP is and it’s not the same as the purpose of a CRM or a marketing automation platform or a DMP. They can grasp that if it’s built for something different, it probably needs to be a specialized tool. So I think that message has really been understood pretty well.
4. Some organizations adopted a CDP pretty early on. Do you think that the focus for them has now shifted to questions like:
Did we make the right decision as to what kind of CDP should we have?
Should we have a more data-focused CDP?
Do we need a more campaign-focused CDP?
Do we just need a CDP that just takes in data and then plug & play with a lot of different innovative, downstream tools?
How are these conversations currently playing out as you’re seeing them in some of the more digitally mature organizations?
Of course, it depends on the organization. People will be talking about whatever pain points they happen to be feeling. If they’re very comfortable with their campaign tools or their personalization tools or their email engine, they’re not going to be looking to the CDP to solve that problem as it’s not a problem.
If on the other hand they have a great CDP, but they find they have issues because they can’t activate the data. Or because one of those other tools that are in place aren’t working, then they may say, “I really wish my CDP could send email”. So they’ll go out and look for a CDP that sends emails. But we definitely do see people replacing their first CDP with the second CDP.
We also see people not replacing but supplementing. They’ll have a CDP that’s doing a pretty good job at building profiles during the data side of things. But it’s not particularly equipped with a good campaign engine or a good real-time interaction manager. And they don’t want to get rid of their CDP but want to add this other functionality.
And sometimes that functionality might be provided by another CDP. So it is not at all uncommon to see companies with multiple CDPs in place because they’re good at different things. And they’ve decided to use both of them for what that particular system is good at.
5. What are some of the top mistakes organizations could have done a little better when they were selecting or implementing a CDP? What are some of the best practices?
Best practices and worst practices are kind of complementary to each other. The number one mistake that you make is you don’t understand your requirements. Then you go out and buy the wrong system. After you buy it, you find out you wanted it to do one particular thing but it doesn’t.
Now, one of the interesting things about the CDP market is that there are a lot of CDPs. So chances are there’s one out there that really does what you need. But you have to look carefully. And it can be very confusing because they sound alike, in the initial level of description that you get from the vendors. Hence, you have to look a little deeper to understand what the scope is of this system versus that system.
But of course, you also have to understand what the scope is that you need. So where do bad requirements come from? They come from people not doing their research and really not engaging in the organization. The thing about it is that you have to talk to the people who are going to use the CDP. To find out what their problems are and what problems need to be solved. Then you have a second level of which of these problems the CDP versus some other system solve. But first, you have to get a list of problems, which really comes down to gaps and use cases.
And then the other mistake that they make is they’ll get that initial set of requirements to find, buy the CDP and get very excited. But the larger organization will be disengaged from it. All the people who are supposed to use it were not engaged through the process. They also don’t get trained. The next step should be to make sure these people really do get trained to use the tool. And many organizations will skip training for one reason or another.
It’s rare for a CDP to fail on a technical level. What happens is an organization bought the wrong CDP cause they didn’t understand their requirements or they didn’t train their people to use it properly. Or there are other organizational issues around cooperation or alignment, etc. Therefore, it really is important to pay attention to those organizational issues, those training issues, those people issues, as well as the technical issues.
6. What are some of the most prominent expectation versus reality dissonances that you’ve seen amongst not just senior stakeholders but across the organization?
Well, certainly data quality is the number one issue that you run into. But it depends on who you ask really. And yeah, there’s an expectation that the CDP will be able to do things. It can’t do so often. Recognizing that identifying anonymous website visitors is a class example. You can’t identify anonymous website visitors, they’re anonymous.
There are certain things that you can do in some cases. But it’s limited. And some people just think the CDP will solve that problem. It’s magically going to identify all their anonymous visitors. Then, there are other issues related specifically to identity resolution where they think the CDP is going to be able to match up identifiers that are not connected. You can only do what your data lets you do. And yes, there are some techniques that can do a little bit of that, but it’s very limited. That’s probably the biggest expectation.
Beyond that, there are expectations regarding how easy it will be to do things. Data is complicated. It takes a while to get the data in there and get the data models built and get things unified. And then you have to train people again, to understand what’s in there to understand the limits of the data. It’s not going to be perfect. It’ll never be perfect. You have to keep continuing to put work into it, into data governance, and into data quality.
Now, on the other hand, we haven’t seen too many people complaining about privacy. CDPs are probably a bigger part of the solution to the privacy problem. Then many people realize that to deploy a CDP, you have to identify all your data sources. You have to connect them. You have to do the identity resolution. You have to do a lot of the same things that you do actually to comply with most privacy regimes, which requires pulling all that data together. And then if someone wants to, you can make a subject access request to share and manage with them.
Your uses of the data are legitimately authorized based on consent or whatever the proper legal authorization is for that use. And the CDP can really solve a lot of those problems. And we’re just beginning to see people recognize how much the CDP can help in that particular area. So maybe at some point again, people will say, I thought the CDP was gonna solve all my privacy problems. But so far they haven’t said that. They’re like delighted that it solves any of their privacy problems.
7. Have you seen any implementation-related best practices that you might want to call out?
The accepted and correct wisdom is you do an incremental deployment, one step at a time. There’s of course, the crawl, walk, run approach. But I think that one best practice that’s not focused on enough is focusing on one department at a time because it’s very tempting.
It’s much easier to train one department and get them used to using the CDP and understand all the potentials of the CDP and to do 3 or 4 or 5 things in a row within that department. Then do 5 separate things in 5 separate departments and try and train each department.
8. Which department do you think is the second likeliest to adopt a CDP other than marketing?
Well, you know, often marketing has multiple teams within it, right? So they’re the advertising people and the email people, and so on. Usually, the CDP by its nature deals with known customers. The first adopters are usually going to be the people who are working on customer retention or on customer development.
We’re now beginning to see more and more interest in the CDP as an acquisition tool. The CDP is all about first-party data. But using your known customers to build lookalike models, to go out and find other people similar to them using the CDP is a way to do retargeting better, even anonymous retargeting.
The CDP has advantages in connecting different systems. To make their retargeting work a little more efficiently. We see more of the CDP, even within marketing being used more broadly. So I would say that it’s not just the department level, it’s the sub-department level within marketing.
Once you get outside of marketing, you’ll see sales groups, operations teams, and customer support teams as probably the most common second users of the CDP because they are very interested in seeing the data that’s not captured in the CRM. So often the CRMs only have some subset of the customer data. And there’s other interesting information – what they’ve done on their website, e-commerce system, or other operational systems that are not connected to the CRM, that can make customer support interactions work more smoothly. So the CDP is often applied to that.
Now my favorite CDP story is that of an airline. They built a CDP and the operations group used it first to understand which routes had the most frequent flyers on them and had the most flight delays. This was done to decide where to put the spare equipment and to minimize the impact of flight delays on their highest-value customers. So it’s nothing to do with marketing. It was totally operational, but that was the group that had the most benefit immediately from the CDP.
9. Do you have a framework to sort of calculate the value of a CDP?
We do have a framework. If you must do an ROI analysis, you need to look at the impact on revenue and the impact on costs. And of course, most of the benefit of the CDP comes from revenue. At the end of the day, you’re going to make more money from your customers. You’re going to sell more. You’re going to have strong relationships. There are soft metrics like loyalty and engagement that also boil down ultimately to revenue.
The problem with that is that it’s hard to attribute any of those gains just to the CDP. The CDP lets you do better marketing campaigns, but you have to do those marketing campaigns as well. Hence, what value came from the CDP? Did it come from the marketing campaign? The answer is you couldn’t do that marketing campaign without the CDP. But the marketing campaign certainly has some value of its own.
So there’s some kind of allocation that you have to do to see how much of that value should be credited to the CDP. That being said, it’s pretty easy to come up with numbers to justify the CDP along those lines. The other problem with that which is the problem with all marketing technology is that you don’t really know what the results are going to be of these new campaigns until you run them.
The other way to justify the CDP is to look at the cost savings and the operational cost savings. So my people are going to need less IT support. They are going to spend less time manipulating data. They’re going to have these kinds of very concrete productivity benefits for the staff itself. And that’s a much easier calculation to make. It’s still obviously gonna be an estimate, but you can come up with fairly reasonable estimates.
10. How do you quantify new capabilities added to the CDP?
What you do is look at what the programs or capabilities are that the CDP will add. There’s a certain kind of marketing program like retargeting, which is a trivial example. And I want to retarget three people who visited my website via email. While those two systems probably aren’t connected. I probably have a way to do that today. The CDP will let that happen in minutes. I didn’t have that capability before and now I do have that capability.
So that’s a program that I couldn’t run before without the CDP. What’s the value of that program going to be? I can’t estimate it until I run it. And that’s the problem with anything that’s brand new. But I can certainly come up with examples of these things that I couldn’t do before that I’ll be able to do now with the CDP. And now let me put some value against those and that’s how you quantify it.
11. How is the looming recession going to impact the kind of decisions around CDP?
We hear different signals from the market. Some people tell us that they’ve been seeing a slow down in CDP purchases. Others tell us they haven’t seen that. We don’t really know. People are certainly worried about the economy in general, the global economy, and that will no doubt have an impact on how freely they can spend on marketing technology.
But again, to go back to the earlier example that I cited, the CDP is often seen as a very critical, strategic long-term investment. We don’t expect that there will be too much of a slowdown in CDP purchases. People are more likely to actually try to consolidate their stack. And they may therefore begin to look for CDPs that have a broader scope of functionality so they can get rid of a couple of other tools that just do one little thing.
The CDP does that and a whole bunch of other things, and that also saves integration costs, which are substantial, especially in smaller organizations. So I think overall the CDP demand will hold up fairly well, even though it may shift a bit into the kinds of CDPs that people buy.
By Bijoy K.B | Senior Marketing Manager at Lemnisk