One unfortunate, and somewhat common, misconception about composable CDPs is that they cannot support real-time interactions. Be assured that this is not the case. While SOME composable CDPs can’t support real-time interactions, a good composable CDP CAN.
But where did the notion that real-time and composability can’t coexist come from? One observation is that some CDP vendors – coincidentally those that do not support real-time interactions – have claimed that real time is categorically outside the scope of what it means to be “composable.”
The truth is that while real-time interactions might be out of scope for some CDP vendors, that’s not the case for every composable CDP.
Composability Means Choice, Not Limitations
The fact that composability means choice holds true for real time, just as it does for other components that together assemble a CDP that is purpose-built to accomplish your unique business and customer experience (CX) use cases.
Whatever your requirements are for a composable CDP – real-time data ingestion, advanced identity resolution, data quality, dynamic segmentation and real-time interactions among them – true composability allows you to choose the capabilities and features you need to power personalized customer experiences.
Some vendors narrowly frame a composable CDP as software that connects to a data cloud that allows you to build segments and activate data to your end channels. Boiling that down, what they’re saying is that a composable CDP requires a reverse ETL tool or APIs that sit directly on the data cloud to provide needed access to individual customer records. Full stop.
If that’s your definition, then by default you would view anything outside that framework – such as the ability to support real-time interaction – as outside the scope of a composable CDP.
From their point of view, real time is only possible when a digital experience platform (DXP) or similar system grabs information about a customer in real time to decide on a next-best action. The underlying – and incorrect – assumption is that a composable CDP cannot make this happen because it does not support a real time cache.
Their Limits Don’t Have to be Yours
A complete, composable CDP can support real-time updates to a unified customer profile as new data comes in, without requiring integration with external systems. This is real time at every layer, from data through activation. It is different than the more limited approach of making a “real time decision” on data that is updated in a non-real-time cadence. Redpoint’s composable CDP supports real time at every layer, because for many businesses, those moments matter.
A CDP can be composable and support real time updates whether the database is maintained by the CDP or a data cloud. A composable CDP can exist on-premises, in a private cloud, a data cloud or even installed as SaaS and maintain composability regardless of where the data resides or of who controls the database.
Artificial constraints aside, composability simply means a CDP customer has the choice of which best-of-breed systems the CDP can connect to for the purposes of powering a personalized customer experience – an experience that may include real time, depending on your needs and goals.
When you’re defining and building the framework of a composable CDP, it’s imperative to make sure you’re putting something together that accomplishes the key tasks and goals you want to achieve. There’s a little bit of a “Goldilocks” approach here – once you outline what you need the system to do, you’ll start to identify the components that can do those tasks – but you’ll need to balance being “under-composed” and not accomplishing those use cases with making sure you don’t end up “over-composed,” with a tangled web of disparate pieces that end up being too complicated, and too costly, for your organization.
Strike the Right Balance of Composability
A complete, composable CDP should have everything a customer needs to strike the right balance – real-time data ingestion, data quality, advanced identity resolution, dynamic segmentation, activation and, yes, real-time decisioning. It should also be flexible enough to run on-premises (in a private cloud), in your cloud data warehouse, or in the CDP’s cloud data warehouse – whichever makes the most sense for your business. After all, composability means you have the ability to customize your CDP setup to meet your needs.
If you’re interested in a composable CDP, be wary of the narrative that there are, or should be, artificial constraints about what a composable CDP entails. Many vendors specialize in only one or two areas of CDP functionality (e.g., reverse ETL, identity resolution, etc.) and if it is a capability they don’t support (e.g., real time) then according to them it’s not part of a composable CDP.
