Why would I need a Customer Data Platform?

The term Customer Data Platform or CDP for short is quite a hyped topic in the field of marketing applications. Many software vendors are currently jumping on this trend. But what exactly is behind the buzzword CDP and what can a Customer Data Platform do for me?

CDP unified customer profile

What is a Customer Data Platform?

Let’s start simple with the Gartner definition:

A CDP is a marketing system that unifies a company’s customer data from marketing and other channels to enable customer modeling and optimize the timing and targeting of messages and offers — Lizzy Foo Kune, James Meyers — Gartner Research März 2018.

Let me put it in my own words:
A customer data platform brings together customer data from all channels to better understand customers and to use this knowledge directly for further marketing activities.

I think this makes sense immediately, but where does the need for a CDP come from? What is currently missing in my system landscape?

What problem does a Customer Data Platform solve?

Let’s think back a few decades: In the past, people used to buy their goods in local shops and had a personal relationship with the sales staff. Salespeople knew exactly what their customers’ needs were and could respond to them.

Today, an increasing part of the interaction with brands and companies takes place through digital channels. This means that customers interact more and more with software systems rather than people. This has many advantages, such as shopping independent of location and time.

At the same time, however, companies are finding it increasingly difficult to respond to the personal needs of customers, especially when customers communicate with the brand through many different channels. Behind each channel there is one or more software systems that process and store customer data.

disconnected systems

The challenge now is that each of these systems contains different information about each customer, e.g.

• Website — Cookies
• Web Tracking — Browsing behaviour
• E-mail marketing — newsletter opt-in, opening rates, click rates, etc.
• eCommerce — online purchases, returns, etc.
• Call centre — service and complaints
• CRM — customer master data
• and much more.

What is missing, however, is a central overview of all data from all systems. Many companies are now trying to merge the data in a large data warehouse or in a data lake. There are two main challenges:

• Matching the contacts from the different systems is not trivial. It is difficult to recognise that different profiles belong to one and the same customer because the customers leave different data about themselves. Once it is an e-mail address, once a telephone number, then maybe just the name…
• Access to this data and its use is often difficult because the IT department has its hands on the central database and only technically experienced people can handle it. If the marketing department wants to use the data for analysis or campaigns, it often takes a long time until someone sits down and makes this information available.

From the point of view of the end customer, the user experience suffers from these data silos. You can see this, for example, in the fact that you keep seeing ads for products that you have already bought, or that you keep receiving offers by e-mail that are not the least bit interesting.

Even a customer service employee has a hard time because he or she does not have all the information at hand that is relevant at the moment of the call. What has the customer already looked at on the website? Has he already added a product to the shopping basket?

How does a Customer Data Platform work?

Actually it is relatively simple. The CDP is now placed in the middle between all existing systems and exchanges data in both directions. First, all available customer data is taken from the source systems, then it is cleansed and merged to create a clean database. On this basis, it is now possible to carry out analyses and run machine learning models over it to make predictions, keyword predictive marketing.

CDP architecture

Everything is aimed at better understanding the customers and ultimately creating target groups of customers that can then be used immediately for marketing campaigns. If I really understand my customer, then I can make him an offer that is relevant to him in the right channel at the right time. The whole thing is then a data first approach, so to speak. I first need a complete and clean database so that I can make valid and intelligent decisions.

working with a CDP

How do I use a CDP?

The complete process of using a CDP is as follows:

1. Data Ingestion: Everything starts with the integration of the source systems. Here it is important to decide which data is required and that it is mapped cleanly.
2. Identity Resolution: The next step is to standardise the data so that they can be compared with each other as well as possible. The profiles of a customer from the different systems are now merged into one uniform profile. There are many buzzwords for this, such as the 360° customer profile, the unified customer profile or the golden customer profile. This is where all information about the customer, his transactions, his browsing and clicking behaviour, etc. flow together.
3. Analytics & Machine Learning: Now that we have a clean database, we can analyse our customers and their behaviour. Machine learning algorithms help us to recognise behaviour patterns, form clusters and gain insights on the basis of which further marketing activities can be carried out. Basically it is about optimizing marketing efforts so that you don’t waste money by targeting the wrong people with irrelevant offers in the wrong channel. With the machine learning you can make predictions like whether someone has a high likelyhood to buy today or to engage in a certain channel. You can also make product recommendations or optimize the sendout time of messages.
4. Segmentation: Now we are moving step by step towards campaigns. We form target groups of customers with common characteristics, such as all female customers from Bavaria who have bought something within the last year, are interested in mountain sports articles, respond well to the e-mail channel and have a high probability of buying something.
5. Action: I now send this customer segment as a campaign to an execution platform, i.e. to a system that executes the campaign or sends it out. This could be an email marketing system that sends out a newsletter, an advertising system that displays ads, or a re-targeting campaign in social media. The campaigns can also be automated, e.g. recurring every week.

Once everything has been set up, marketers only work with parts of the platform in their daily operations, e.g. with the campaigns or the analysis functions. But other employees from other departments can also be granted access to the platform. For example, call centre agents from customer service can send live queries to the system to view customer data and provide better advice to customers.

What distinguishes a CDP from other tools on the market?

A CDP is not a DMP! A Data Management Platform works primarily with anonymous / third party data that is intended for advertising. A Customer Data Platform works with known first party customer data.

A CDP is not a CRM! A Customer Relationship Management System stores the master data and communication history of the customers and can therefore serve as a data source for a Customer Data Platform. The CDP is rather a marketing tool with which the customer data can be used for marketing purposes.

A CDP is not a DWH or Data Lake! A Data Warehouse or Data Lake is a collection of large databases managed by the IT department. Access to the data is usually reserved for IT staff. In a CDP, marketers have direct access to the data and can use it immediately for campaigns.

The CDP market

Although this topic is still relatively new, there are already a lot of software manufacturers on the market that offer a customer data platform. Some of them are still very new and small, but even the very big Marketing Cloud providers are in the process of positioning their own or acquired solutions.

The maturity level of the solutions, however, differs considerably in some cases. Especially the impact of the Identity Resolution is often underestimated. There are also enormous differences in the scalability of the solutions (the amount of data that can be processed with high performance).

A good source for more information about CDPs is the CDP Institute. There you can also find comparisons of the different platforms.

As I work for Acquia, I can of course recommend our own platform, the Acquia Customer Data Platform. Our CDP is already successfully used by many well-known customers such as Hugo Boss, Clarks or Lululemon. Not only can our customers use it to optimise their marketing activities, they are also generally experiencing enormous improvements in their operational processes through the use of the CDP.

Finally, I would like to make a book recommendation. The founder of our CDP, Omer Artun, has written a book together with Dominique Levin called “Predictive Marketing” in which he explains how to use the software successfully for your business.

I hope you enjoyed my little insight into the CDP world. If you need more information, feel free to contact me at any time.