If you have or are considering a composable customer data platform (CDP), you’re likely familiar with all the reasons why it’s quickly becoming the preferred solution for powering a personalized customer experience (CX). You decide which components to use to build a better customer understanding. You control how your APIs are set up. You maintain flexibility, easily changing how your CDP operates to meet new trends or use cases. And, if you run your composable CDP on a modern data cloud or on-premises, you have a “data-in-place” environment where your data remains in your control, within your own security perimeter.
If this is your experience – congratulations! You’ve hit a composable CDP home run. But if you’re instead finding that it now takes you more steps to accomplish the same amount of work, or that you don’t have quite an accurate view of your customer as you expected, then perhaps you’re over-composed.
Over-Composed? Spot the Red Flags
A CDP that is over-composed is composed at too low a level. What this means is that there are so many different components that the essence of a CDP gets lost. You find it harder – not easier – to power a personalized CX based on deep customer insight because it seems you’re just moving data from one place to another.
How can you tell if you’re over-composed? Technical terms aside, one giveaway is if it becomes harder to accomplish a business function such as identity resolution, building an identity graph or a golden record, data transformations, data formatting, etc. Here are some other telltale signs your CDP is over-composed:
- You have to navigate between multiple programs to accomplish routine tasks.
- You are forced to learn SQL,or need to rely on IT to get what you. need.
- You are given data that is not ready for prime time (it’s messy, there is no single source of truth for the customer).
- You have to make assumptions about the quality and the trustworthiness of your customer data (Is it accurate, deduplicated and complete? What about householding?)
- You have multiple tools with overlapping functionality
- The components you’ve tasked with completing a business function (e.g., identity resolution) have various standards for the quality of data, and – for some – their standards do not meet your requirements.
In short, the purpose of composability is to bring various components that you need together in a sensible way to make it easy to complete a business process. Anything that detracts from that goal is a sign that you’re over-composed.
Composability Done Right
What, then, is the right way to compose your CDP to avoid the problems of over-composability?
To understand the answer to that question, it helps to understand why over-composing is a problem. The reason is that some CDPs that tout “composability” limit the definition to having a system that sits on a modern data cloud that essentially moves data from one place to another. It might do one function well (e.g., identity resolution, reverse ETL) but does not even pretend to have complete CDP functionality.
There are two points to make here. First is that while most composable CDPs do in fact sit on top of a data cloud, a CDP can still maintain composability in a private cloud, on-premise or even in a SaaS environment. Second is that a CDP does have certain requirements to fulfill to be called a “CDP.” These include the ability to ingest data from every source, capture details of the ingested data, create and maintain persistent keys and create unified profiles among them. A CDP that outsources all but one core function forfeits the right to call itself a CDP.
In other words, the hard things such as data quality, identity resolution and creating unified customer profiles that persist over time don’t disappear just because your system sits on a data cloud.
On their own, various components may not only fail to meet your requirements, but an over-composed system becomes an operational nightmare. You have one interface to pull in your data, another for identity resolution, a third to create a segment — likely using SQL — a fourth for your reverse ETL and to activate your segment, etc. Not a single application encompasses the whole business function of bringing in your customer data, creating a trusted unified profile and activating it to your end channels in a controlled way.
If that describes the day-to-day responsibilities of an operational marketer working with your composable CDP, then your “composition” is really just a collection of individual applications that a marketer has to learn how to use and coordinate in order to complete a business function. That sounds like a process made more complicated and less efficient, not more streamlined and effective.
The purpose of a CDP is to complete all those tasks in one place; log in, pull your data in, check to see if your data is ready, build a segment, push the segment out, look at your results, run an experiment. One interface. One source of truth for your customer data. One CDP that maintains an impeccable standard for data quality across the complete business function.
To discover how the complete, composable Redpoint CDP helps companies get their data right, make it easy to use and evolve with your use cases and technology stack, click here.
