The CDP promises to provide the key to comprehensive data-driven marketing, a very attractive marketing concept where all your customer data is combined for marketing (and other) uses. Without the data and the management of the data, the ‘marketing brain’ that allows for smarter campaigns simply can’t function. So bringing together the data is very important.
There are some distinct characteristics of a CDP that explain why major brands are looking at them instead of relying purely on other types of data systems like CRM (customer relationship management) or a DMP (data management platform).
So what is a customer data platform and what is the difference with other systems?
What is a customer data platform?
The aim of the CDP is to bring together all customer data and stitch the data together into unified customer profiles. So a marketer can easily work with it.
You could say we have already heard the story of tools that can function as a hub or central place for customer data. There certainly has never been a shortage of technologies that claim to provide a ‘single customer view‘ or a ‘360 degree customer profile’.
But in reality there hasn’t been one platform type out there that has the potential to bring together all the data and at the same time properly make it useful for the marketer.
David Raab, a respected martech analyst, was the first to define the CDP category in 2013, and the definition reads as follows:
A customer data platform is a marketer-managed system that creates a persistent, unified customer database that is accessible to other systems.
Visualisation of a CDP (source: Emailvendorselection.com)
How is a customer data platform different from what we know?
What are the big differences between a CDP and systems like CRM or DMPs?
1) Different than CRM
CRM systems are built to engage with customers, this is on the basis of historical and general customer data to create a persistent customer profile. They aren’t built to ingest huge volumes of data from other sources.
A CDP is able to connect all types and sources of customer data, whether internal or external, structured or unstructured, batch or streaming. This allows you to form a much more comprehensive view and to better understand your customers, and act on it even in real-time.
The data sources will include different categories:
Transactional and order data: Ecommerce, administration and sales systems that generate data on purchases, order- and renewal dates, customer and product value, abandoned baskets, returns, and more. This is without exception very valuable data for marketing and sales efforts.
Behavioural, web and mobile data: Products and categories browsed, clicks, store, interaction data, number of pages visited, and more. SDK anyone? Data that event driven campaigns need, your analytics love and predictive models crave. This is the data your mother told you to pay attention to as it speaks volumes on current and expected behaviour and preferences.
Profile data: Without customers, you wouldn’t have a business. Knowing who they are and what they want, will lead to more effective marketing. This category starts with contact data and opt-in, then you can enrich with psychographic data points – details about lifestyle, context, preferences and personality.
Product data This is not customer data, but unmistakably necessary to include and execute on personalised engagement with the customer. Think about stock levels and pricing as just two crucial data points.
The management of products and their data has increased in scale as assortments become bigger and product life-cycles shorter. Many products are only available one season, for instance a bike will be there for a few months then is replaced with a new edition. Typically this data enters the customer data universe through product data management systems or ecommerce platforms. Bigger companies will have standardized exchange formats with their vendors, to increase efficiency.
CRM & offline data sources: Here we have the CRM! Typical CRM data is profile card type. Phone, post, email, permission and suppression data, firmographics. But this type also includes information from web forms, surveys and noted down in CRM by agents.
The above examples are just the tip of the iceberg, but that is no problem as the CDP is built to handle any type and category of data (currently known). The ability to flexibly add and change new types of sources makes it more future proof in the function of a central repository.
The only constant is change and sources do and will change in the digital marketing environment. Just think of the upcoming data avalanche from Internet of Things applications.
2) Different than custom integrations
A CDP is built with the marketing function in mind. Could you build a CDP yourself? In theory you could, and many custom IT projects try to achieve what a CDP does. Obviously there are quite some big investments, time and risks involved in a custom project like that.
CDPs are not just databases. They standardize and package “hidden” features. These features contain pre-built marketing database and packaged tools to make database creation and operation easier. You can think of data handling features to standardise, stitch and cleanse data.
Some CDPs have additional functionalities built in, like BI, analytics, reporting or tracking.
3) Different than delivery platforms
In the CDP world, the systems that interact at channel touchpoints are called delivery platforms or engagement platforms. These are for instance your email marketing or marketing automation software, website, account based marketing system, social media management platform, apps.
These delivery systems interact with the CDP to send out messages, but also collect engagement and campaign data to feed back into the CDP. They aren’t part of the CDP, but some CDPs have tight integrations with the delivery platforms and make is possible to plan campaigns and even send out messages directly from within a CDP.
4) Different than DMPs
It is easy to confuse a DMP with a CDP. DMPs were designed to serve advertisements and enable retargeting using cookies. They focus more on anonymous segments and categories than single customers. In a DMP, much of the info is anonymous and typically expires after 90 days (the cookie lifetime).
The CDP creates a persistent customer profile. That means it stores the data and keeps the history. Then by combining it with all the data about your customers comes out with a single record.
DMPs were designed to target (and especially retarget) anonymous users for advertising, whereas CDPs create a database of your identified customers to use for more than only advertising.
5) Different than data warehouses
Traditional data warehouses are built and run by IT teams with big technology know-how, but little marketing feel. The data warehouse aims to bring data together, but the marketer has to rely heavily on the IT department, which often make going from ideation to execution the slow lane to frustration. The reality is that most valuable customer data demands to be quickly accessible and ‘actionable’ customer data.
Some technical involvement is still required, vendors commonly can handle this in collaboration with your IT department. Marketers get greater control over the marketing database and additional features.
What makes a customer data platform different from other systems?
A CDP is a packaged system that creates a persistent, unified customer database. It can ingest all types of data sources and gives open access to other systems while being marketer-centric.
Comparison chart of the functions of CDP versus data warehouse, DMP and CRM.
Moments are the building blocks of relationships
Creating a comprehensive customer view is admirable, and sounds like a great marketing-philosophy, but… and there is a but…. it stays a philosophy until there is an actual use case. That is why we should start by looking at customer moments.
Moments are the building blocks of relationships, and can be enabled by a CDP. They are what customers will remember and, taken as a whole, will define how your brand is perceived. That’s why companies should be able to make the most out of these moments across multiple (maybe even all) channels.
In their book “The Power of Moments: Why Certain Experiences Have Extraordinary Impact”, Chip and Dan Heath explain what psychologists now call the Peak-End rule:
When people assess an experience, they tend to forget or ignore its length. Instead, they seem to rate the experience based on two key moments: (1) the best or worst moment, known as the peak and (2) the ending. [..] What’s indisputable is that when we assess our experiences, we don’t average our minute-by minute sensations.
Marketers therefore should be architecting extraordinary and memorable peak moments.
How Deutsche Bahn uses a CDP to create memorable moments
A few weeks back, I spoke with Dr. Markus Wuebben, founder of the CDP platform CrossEngage. He showed an example from one of their clients that was presented at the OMR Masterclass event.
The German railway company Deutsche Bahn has started working with a CDP and crafting moments.
From research, Deutsche Bahn know how strenuous tightly packed travel can be, and that their customers are more likely to upgrade to the first class in this situation – if they were only given the chance. Using the CDP they can offer travellers a discounted upgrade to the first class when second-class is crowded.
The upgrade is offered through the Deutsche Bahn app via push notification. It doesn’t matter if the user purchased his ticket via the app, on the website or a different channel, because these are all connected to the CDP.
One thing they need to know for this campaign is the travel destination of the traveller. Because the CDP combines the personal data with contextual data (how busy the train is on the trip), to make this happen. It is a special moment a customer will definitely remember, as well as direct profit generation for DB.
You might wonder how many customers are identifiable to give them this upgrade. Turns out that over three quarter (77%) of customers are logged in when booking, so can be targeted via the CDP. With 1.7 Million bookings and 2.7 Million logins a month that is nothing less than a very substantial group.
Customer Data Platforms have the potential to create a single customer view. But the real value of that is making it actionable and gaining the insights to do so. Think of the questions that can be answered by CDPs:
- What was the product this customer bought before their current purchase?
- Which segments / target groups does this customer belong to?
- Is this customer likely to churn?
- What has he shown interest in lately?
- What is their (purchase) intent, and timing?
- What is the value and predicted future value of this customer?
- Where do they prefer to interact and create moments?
- What are their preferences and where are they in the customer journey?
Obviously this is just a sample, and it might seem that it is all about analytics. But it is about easily deciding on the right segments, user journeys, messages, channels, and timings (while validating hypotheses, doing A/B tests etc.) to improve CLV and prevent churn.
Once you combine the underlying data and insights, they can lead to even more remarkable, valuable moments.
Online training courses
To learn more about this topic, see Econsultancy’s Online Fast Track Data and Analytics course, or explore the rest of Econsultancy’s online training offering.
Who are the main CDP vendors? The most recent list I can find is for 2016 and it doesn’t mention CrossEngage described above.
http://www.crossengage.io is mentioned in this article because they shared a real CDP use case of Deutshe Bahn with me. That is what I would recommend before looking into CDPs, always have your use case(s) in mind.
There is quite some confusion on the subject of what a CDP is and is not. I imagine the vendors all like to be called a CDP (because that is what marketers are looking for) but are some are really more of a DMP or CRM solution. Which I think is perfectly fine, because if a company knows what they plan to do, they select the type of solution that is fit for it.
Some others just not fit with All of the requirements such as “marketer managed” or “open to other source or delivery platforms”.
See the chart I made for the differences between the different systems.
For your question, David Raab keeps an overview of CDPs here:
If you end up looking for a marketing automation or email marketing software (mainly delivery platforms). You can have a look at the list I keep on automation and ESPs.
Hope that answers your question, keep ‘m coming happy to answer all I can 😀
Thanks both for your answers.
Great overview on CDPs, especially the very important differences between that and CRMs/DMPs. Really useful to consider, as much more data is available and collectible–which for marketers usually means much more headaches trying to fit it all into one smooth, flexible system. Poorly implemented management systems (or just poor marketing) can lead to bad moment-creation, and I’m excited to see how CDPs alleviate that going forward. Thanks for the great post Jordie!
Nice description of how they differ. From your perspective, what would you describe as the key elements a CDP needs?
Hi Garry, thanks for asking. Companies will be seeing, what is the ideal set of features for them. And that depends on their use-cases, what they will be doing with it.
The pre-packaged features, those are interesting.
Putting together the different datasources on one side and combining them.
On the other side, you probably want some of the Descisioning (segmentation, next best product / Action) or even campaign management / Automation. See that the role a CDP plays is also depending on the current rest of your marketing tools / MarTech stack.
Garry, what do you see that your clients are using most in the beginning and are asking for the most?
Cheers for that Jordie. Makes a lot of sense. In terms of what we have been seeing, it tends to break into two areas. One is those that want there campaign management built into the CDP, it smooths work they are doing and enhances automation, particularly speed to get live. On the flip side its about utilising the data and more and more people are asking how to access the data more and for it to drive new routes like AI and predictive modelling
Hi Jordie, terrific article – thanks for that.
Artificial Intelligence can also be used as part of that data layer to match customer preferences with product attributes. In practical terms, an integrated customer-product data platform will:
– Analyse the retailer’s product catalog, deconstructing each product into a series of attributes
– Learn the preferences of each consumer for each attribute, based on signals generated by their behaviour
– Calculate a value that represents the relevance of each product for each customer
– Serve the most relevant products for each individual consumer
Customer and product data have a yin-yang relationship: the ability to match customer preferences with product attributes drives relevance for shoppers and boosts revenue for brands.
Hi Peter, definitely agree that in many cases the product information and attributes needs to be unlocked and be available if you want to match the right person with the right products. automation becomes more important as catalog grows, products are selected because of varying attributes and sales-cycles become shorter.
Many companies have been struggling with this, or not even considering it. CDPs will have varying levels of built in Decisioning, some smarter, some require that to be outside of the CDP somewhere else in the technology stack.
I would say – CDP is a new creative way to understand customers and leads other than cold calling or email marketing because with CDP we can track the real behavior of customer and lead on our website.
The case study with Deutsche Bahn is really interesting. With an emerging approach to bringing all the data we have on customers in one place CDP will be more popular over time.
Thanks for sharing this Jordie.
I loved the way you summarized the article.
I agree that a CDP has an upper hand when compared to a CRM system and that’s what we advocate our clients at http://www.wavex.co.uk/