Eureka! How to understand a crucial CDP concept

Gorgeous black couple
Summary: Some people have a hard time understanding the relationship between online and offline data in a CDP. Here’s a new way to think about it.

A customer data platform is a system that collects, harmonizes, and combines customer data from across your enterprise to create something like a single customer record. The CDP is connected to your website, so it makes a profile for every web visitor.

Some people mistakenly think that if you have a web profile for a customer, and then you load additional information about that customer into the CDP, that web profile will be updated with the new information.

Well … not so fast. There has to an identifier that’s common to both records to allow you to merge that information. For example, if a record from your fulfillment bureau has an email address, and the web profile has the same email address, then yes, you can merge the records.

Here’s the part some people have a hard time getting. Even if Joe Smith has a web profile, and the very same Joe Smith has a profile from some backend system, the CDP can’t merge those two records without a common identifier.

People struggle with that concept, so I spent some time thinking about a better way to explain it. While I was thinking, I remembered a conversation with my brother about two ways to frame the same issue. I can’t remember the exact example my brother used, but here’s one to give you a flavor. I believe this is from Daniel Kahneman.

Imagine you are presented with four cards. Each card has a number on one side and a letter on the other. You are given a rule: If a card has a vowel on one side, then it has an even number on the other side. Your task is to determine which cards you need to flip to check if the rule is being followed.

The cards you can see show A, 3, 6, and M. Which cards would you turn over to check the rule?

Now, imagine a scenario involving people at a party. You are told that if a person is drinking alcohol, then they must be over 18 years old. You have four cards representing those four people. One side shows their age, and one side shows if they’re drinking.

The cards you see who one is Drinking beer, one is 16 years old, one is drinking Coke, and one is 22 years old. Which cards do you flip?

For a lot of people, it’s easier to solve the question when it’s expressed the second way, because it’s relying on things we know, and are familiar.

My challenge last night was to find a way to explain the CDP issue I mentioned above in a way that’s easier for people to understand. Here’s what I came up with, with a little help from ChatGPT.

You’re managing a party, and you’ve prepared the guest list. You don’t know any of these people, and you don’t have their photos, but you have lots of other information about them. You know their names, addresses, email addresses, and phone numbers.

The doorman ensures that everyone who comes to the party is on the guest list, but when they enter the hall, you have no idea who they are. You have a very nice database that explains lots of details about all of them, but you can’t match the person to the database entry until they give you their name.

That’s what going on with your offline data and your web visitors.

Leave a Reply

Your email address will not be published. Required fields are marked *