What is a Customer Data Platform?

The customer data platform (CDP) is one of the fastest-growing categories of marketing technology today. In order to understand why, you have to look at some of the underlying challenges in marketing related to data and personalization.

What does a customer data platform do?

We live in an era where the customer is in control. Amazon can predict what products we will buy next. Netflix recommends the shows we like with great accuracy. Customers want personalized experiences and fast service, and expect companies to have an intimate understanding of their preferences. It’s no longer a marketing advantage to deliver on this — it’s table stakes.

Customers want the interactions they have on a company’s website to translate to their mobile app experiences and even in-store visits. And they want everything updated in real time, so their current needs are being met on every channel. The problem is that, for most companies, those environments operate from different data sets — even though the customer is the same. 

As customers move from channel to channel, they expect their experiences to be consistent and “in the moment.” Most customer journeys involve three or more different channels (email, web, and mobile app, for example), and customers tend to move seamlessly and quickly between these channels. Many companies, however, don’t have these data environments connected in real time.

The result is a disconnected experience for consumers and the lack of a single source of customer data for companies.

That’s where a CDP comes in. It combines all of that data in real time for companies, allowing them to offer hyper-personalized experiences that feel like magic for customers.

How does a customer data platform work?

The first thing CDPs do is connect all of a company’s customer data in a single place. That means not only stitching together a single customer ID from many different customer interactions, but also tying together databases that traditionally don’t share data, like marketing clouds, service software, and ecommerce engines. We call that customer resolution.

The next thing CDPs have to do is reconcile our identities for our known customers (like email and mobile numbers) with what we know about customers before they share their data (anonymous cookies and mobile device IDs, as an example). This way, we can start to associate a journey that started with an email campaign and continued onto the website with the same customer. We think of this as a cross-device identity.

Once the CDP has created unified profiles of customers, the system has to make that data available in real time so companies can deliver personalized experiences. That means connecting customer data to many different types of systems, such as email-send engines, demand-side platforms, and content management systems. 

In a nutshell, CDPs are concerned with these primary tasks: data collection, data unification, data activation, and data insights.

As the amount of consumer data grows, however, companies must be aware of and respect data privacy concerns. Many consumers are willing to have some of their data used to deliver magical, personalized experiences, but they expect that companies will protect this data and use it ethically. These are the building blocks of a trusting relationship with customers, and companies need to use consumer data in the right ways.

What are customer data platforms used for?

Here are a few examples.

The right ads at the right time 

Sometimes the best use of data in marketing isn’t used to better target consumers — but to not target them at all. We’ve all had the experience of being targeted online by ads for things we’ve already bought. 

The reason companies have a hard time stopping ads for the sneakers (or cars) we have already purchased is disconnected data. A unified profile that connects marketing and purchase data enables marketers to be smarter with their budgets by suppressing consumers that have already made a purchase and redirecting those dollars toward prospective customers, or recommending other products to existing customers.


Say a consumer comes to your website, browses a particular product — say, a new red convertible — and leaves. Wouldn’t it be great if you could tie everything you’ve learned about that customer to a personalized offer via email or push notification? For example: “Test drive that Honda today and get $500 off MSRP!” You can only do that by connecting that consumer’s identity to your marketing engine.

CDPs solve this problem. Customers who see content tailored to their interests (test drive this red convertible today!) are much more likely to engage with a brand.


What drives better marketing? The answer has always been better customer insights, but most analytics systems operate in silos.

Email engagement data is separate from website analytics data, which is separate from display ad data. Stitching that customer data together, and tying all of those interactions with the same consumer, can be a herculean effort when you are using legacy systems.

What if an outdoor retailer had a customer’s marketing interactions (email and advertising engagement) tied together with their ecommerce data (purchase history) and website interaction data (products viewed multiple times) – and made that information available to a service rep in the call center?

A bit of data science (customers who purchased this tent online, and opened email promotions for these hiking boots, and spend between $250 and $1,000 annually, usually buy this backpack) can reveal the right product recommendations to the call center agent, who can make a personalized, relevant offer on the spot. This type of personalization can turn a $15/hour call center rep into a $100,000 a year salesperson. That’s the power of a CDP.

We originally published this article in 2019 and updated it in September 2022. Chris O’Hara, Salesforce’s former vice president of product marketing, contributed reporting and writing to this piece.

Source link

Interesting Posts

Leave a Reply

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