“The Different Flavours of a CDP”
by Holly Ripley, on 24 November, 2020
Apoorv Durga delivers another knock-out sesh at #MarTechFest Dial Up Series 2.
Sitting at the heart of your marketing stack, a CDP holds the capability to connect your data, allowing rich customer data to flow from multiple platforms to a single destination where that data can be collated into a single customer view. Now if that’s not what every modern-day marketer dreams of, we don’t know what is…
While implementing a CDP may make your life easier by destroying dreaded data silos, selecting the perfect CDP for your business is no walk in the park.
The CDP space is exploding, with new vendors emerging almost every week, along with a handful of vendors already out there, waiting to be discovered. We’re seeing a lot of noise surrounding this space (and with good reason) but this noise can turn into a bit of a headache when trying to select the right CDP.
The marketplace is jam packed with products that claim to have CDP-like qualities and often describing themselves as the “largest and fastest growing”, “the martech leader”, “the first to bring AI to CDP’s” and my personal favourite, “the world’s largest provider of…” Some solutions will call themselves a CDP vendor even if all they have are basic integration capabilities. Our good friend, Apoorv tells us that step one is to call out the bull$h!t (he may have worded this a little more delicately) in order to filter out the products that are not relevant to you.
What Can a CDP Do?
But before we get carried away talking about how to find the shiny new tech and colourful logos that can be relevant to you, let’s take a step back. What can a CDP actually do? Apoorv answers this question wonderfully in 5 simple steps of the customer data life cycle:
1. Data Ingestion
A CDP brings in data from multiple different sources. From offline to online and the collection of first, second and third party data. A CDP supports different types of data sources and holds data ingestion capabilities.
2. Data / profile management
CDPs can perform various data management tasks such as data cleansing and normalising. When the data is normalised a CDP can stich a unified profile of your users creating a 360 degree or singular customer view (SCV).
Once you have created profiles of your customer a CDP can segment these users based on different attributes. They slice and dice the profiles to create audiences. Manual rules can be created for these audiences or the system can recommend segments based on machine learning.
These segments form a basis for analytics. There are several ways these tools can support analytics; Some tools provide basic dashboards and reporting while others can provide much more sophisticated analytical capabilities e.g. predictive analytics, predicting next best actions and so on.
Next is the outputting and setting all the data and segments to work. CDPs mostly rely on integrations from other platforms and APIs for activation. They usually do this by sending audiences and activations to external applications that set personalisation agents.
The above is a high-level overview of the data life cycle a CDP can provide. Apoorv states that not all CDPs should offer all 5 capabilities or can do all 5 well but it’s important to know which capabilities your business will benefit the most from when selecting your CDP.
Some key ways to categorise CDPs are as follow:
- Based on data life cycle (as discussed above)
- Suites Vs best-of-breed
- Industry vertical / domain
- Support for business use cases
-Personalisation and recommendations
-Loyalty and rewards
- Deployment approaches
Apoorv tells us that not all products will excel in all categories, he says that you’ll likely find most vendors specialise in one or two of these areas.
So, you need to ask yourself, “what do you want your CDP to do?” For example, if you’re a large enterprise and have several other products, perhaps for analytics or personalisation you’ll have several choices and considerations for building your stack and how your CDP is going to fit in there.
Let’s take the category ‘Suites Vs best-of-breed’ and dive in a little deeper…
Suites Vs Best-of-Breed
One of the main considerations you’re likely to face is whether to use a CDP from a suite Vs pureplay. In other words, do you want a specialist CDP provider, or do you want to use CDP capabilities from a marketing cloud that provides other key capabilities?
Vendors such as IBM, Adobe, Salesforce or SAP all have some sort of CDP capabilities. Suite vendors can provide the solutions across multiple different categories. It can be tempting to licence one of these big suite providers as your solution across multiple martech requirements, and often customers start believing that a single suite will solve all their problems. When in fact, research shows otherwise.
Apoorv explains that many suite vendors have cobbled together their suites based on several acquisitions and as a result, their offering is not a single user experience (UX) and architecture but often a badly cobbled together package of multiple products and tools.
As a stack owner you need to decide whether your customer data should be apart of a larger platform, like IBM or SAP or should you be more conscious of separation of concerns and invest in an independent stand alone, pureplay CDP?
Of course, both approaches have pros and cons. Most small vendors offer CDPs targeted at marketers which means CDP offering is purpose built and works out of the box for specific marketing needs. In that sense, CDPs often represent a highly purpose-built solution.
If you’ve already invested in big software vendors and already have licenses to these marketing platforms it might make sense to look at their CDP offering. On the other hand, if you value a cleaner separation in your architecture you should consider an independent CDP.
Apoorv gives us two key takeaways to keep in mind when considering suites Vs best-of-breed CDPs:
- Consider CDP capabilities of a suite’s CDP platform based on its own merit and not because they are part of the suite.
- Don’t assume that all the components of the suite or cloud are well integrated - because they’re often not. So, be sure to plan for additional resources and efforts to be able to integrate them well.
Apoorv tells us that the first and most important stage of your CDP selection process should be to produce a filtering criteria based on a selection of categories. These categories don’t have to be the same as Apoorv listed above, your categories will vary depending on your needs.
By producing a filtering criteria, this will help you filter out products that don’t match your needs early on, while also give you a good understanding and overview of the technology landscape. (Our CDP Directory may also help with this, more info here).
CDPs are one of hottest topics in martech right now, and while we’d have loved to type until the end of time about the exploding space on this little page, we’ll do you one better and offer you a variety of interactive ways to get your learning on…
- Sign up to our MarTech Alliance Membership to watch Apoorv’s full sesh on ‘The Different Flavours of a CDP’ (You can also catch all other #MarTechFest Dial Up Series 2 sessions to binge on demand to your heart’s content)
- Check out our CDP Buyer Guide eBook and download a copy to keep
- Register your interest for The Fundamentals of Customer Data Platforms Digital Learning Course (coming soon)
- Keep your eyes peeled for web sessions exploring all things CDP