Let’s start with a simple truth: you can’t do anything with data if you can’t move it around. Data has to come in from where you collect it and go out to where you need it.
Of course, just being able to move data smoothly doesn’t get you too far either. Turning customer data into better customer experiences and better business results means addressing the full spectrum of customer data challenges, from identity resolution to assembling and accessing the elusive customer 360 in real-time to uncovering and activating the customer intelligence that matters. For customer data to be truly transformational to a business, a Customer Data Platform (CDP) needs to be able to do all of that — but none of it can happen if there isn’t a smooth flow of data between different systems.
At a Glance:
Three ways to move data: Pre-built Integrations, File Exchanges, and Full Database/Table Exchanges
You need all three for maximum flexibility to address all potential use cases
Most standalone CDPs don’t have all three and typically lean on pre-built integrations, significantly limiting the ability to work with other systems
Marketing Cloud CDPs are designed to connect to their own systems, cutting off many options and forcing brands to uproot systems that deliver strong value
Enterprise brands today often have data in 100+ different systems — point of sale, email, loyalty membership, website browsing data, and social media channels, just to name a few — all of which needs to be brought into the CDP. Then after all the intelligence, analytics, segmentation, and predictions happen in the CDP, curated audience lists need to go out into other systems for campaign delivery, customer service, and dozens of other use cases.
Data comes in, wild and untamed. Data gets organized and analyzed. Data goes out to change the world.
But like we said: you can’t do any of this if you can’t move the data around. The processes outlined above depend on numerous exchanges of data between systems and subsystems that perform the different pieces of the puzzle — from the point of collection of consumer interactions through ingestion, transformation, storage, querying, processing, and outputting. Thanks to the sprawling marketing tech stacks most brands have today (according to one study, large companies use an average of 163 SaaS tools with a 4-year growth rate of 68%), this becomes a complicated endeavor. Not only is it complicated, it can be really frustrating — a recent IDC survey found that application integration and connection is the top challenge and least satisfying aspect of all things related to marketing tech:
The problem lies in the fact that most of these systems were not designed with data exchange as a top priority. This has led CDPs to offer pre-built integrations to the most popular and commonly-used systems, so that data flow is not a blocker.
The Power of Pre-Built Integrations
At the simplest level, moving data between systems depends on three keys:
1. Creating a connection (e.g., giving the source access to the destination system)
2. Having matching abilities between source and destination to send and receive the data (e.g., acceptable formats, scale, etc.)
3. Enabling the receiving system to make sense of the source data (e.g., mapping the column “last_name” in the source to the column “surname” in the destination).
Setting these up in-house usually requires brands to make a significant investment of engineering resources. A CDP with a pre-built integration between two systems theoretically means that these requirements are all taken care of, so users can move data between systems smoothly and cost-effectively (although in practice these integrations still need to be turned on for each account, requiring some support, despite CDPs claiming that it is automatic — more on this below). Once the integration is activated, business users — marketing, customer support, analytics, e-commerce managers, and so on — are free to roam. They can dream up and execute new campaigns, adapt to changing customer journeys and preferences, experiment with offers on the fly, and more. But when these integrations do not exist, business users rely on IT to make each new connection. IT typically has a very long backlog, meaning that only the absolute highest priority new items are picked up while the majority of ideas for great customer experience languish in the queue.
Beyond Pre-built Integrations
So what happens when you need to connect to a system for which your CDP has no pre-built integration? It’s bound to happen sooner or later — marketing technology tools are an ever-expanding universe (the famous Martech 5000 currently has over 8000 logos… maybe time for a rebrand). The proliferation of tools enables brands to take a test-and-learn approach and find what works best for them, experimenting with new types of campaigns, personalized interactions, and channels. That may involve moving data somewhere that has no existing connector, so it’s essential that their data foundation have the flexibility to connect to any valuable new technology.
There are two ways to foster this kind of flexibility: file exchange capabilities and first-class support for popular file storage and database systems.
With file exchange, authorized users can directly download segments, lists, or tables from their CDP in industry-standard formats like CSV or Apache Parquet, and then manually upload that data to any system. This approach gives users the flexibility to explore and use data where they need it, when they need it.
The other approach is to send segments, tables, full databases, or files to a secure storage location or database where multiple systems can extract and use that data. Examples include databases like AWS Redshift, Snowflake, and Apache Parquet; and object storage systems like Amazon S3, Azure Blob Storage, and Azure Data Lake Storage. This approach is efficient, flexible, and durable, supporting multiple use cases in tandem and never breaking when the destination systems you want to send to change their requirements or schemas.
Common Gaps in CDP Approaches
What we outlined above is an ideal state for data circulation: an extensive list of pre-built integrations with use case specific systems (such as an email automation tool, or a customer service system), plus supplemental solutions like file exchange and more general purpose integrations with popular storage systems for maximum flexibility as your needs evolve. Unfortunately, the majority of the customer data solutions on the market won’t get you there. Customer data management comes in two flavors, the standalone CDP and as a component of large marketing clouds — and both have their pitfalls. Here are some considerations when evaluating customer data solutions:
Standalone CDPs often only provide pre-built integrations, without the file drop or shared database option. Most vendors in this category don’t have enough integrations, offering 100 or fewer. A couple (we know of two) that made this a core part of their value proposition have a lot — 200 or more — but these may not be as seamless as advertised, requiring some engineering involvement from your IT team. Given these limitations, it’s important to probe if a standalone CDP has other methods of circulating data between systems beyond pre-built integrations, and also to understand whether the pre-built integrations are truly ready as-is or require additional work.
Marketing cloud CDPs take a different approach, funneling users to connect to data systems within their own ecosystem, which doesn't help if your existing tools are not in that ecosystem. Rather than building on what’s already working, you could end up with no choice but to replace tools you currently rely on in order to use the marketing cloud CDP. And that marketing cloud CDP likely doesn’t even have all the functionality that you need in terms of identity resolution, 360 degree views of the customer, and advanced analytics — as Tony Byrne, Founder and Principal Analyst at The Real Story Group, wrote in a Nov 2020 evaluation of the CDP space, “Major suite vendor CDPs remain comparatively immature.”
The Takeaway: Look for Full Spectrum Capabilities & Full Data Flexibility
Let’s go back to the two basic premises: for creating the best customer experiences that build long term loyalty, a CDP needs to be able to cover the full spectrum of customer data challenges, and it needs to be able to take data from anywhere and send it to anywhere. That being the case, when buying a CDP you should focus on the comprehensiveness of functionality and maximizing flexibility of data flow. Neither one is enough on its own, but together the results can transform your business, and the way your customers experience it.
At Amperity we have followed these principles and launched nearly 200 pre-built integrations, combining these with file exchange and full database/table exchange. This helps enterprise brands build exactly the tech stack they want, to support any use case they need. For more information, check out our list of pre-built integrations here.