The rise of data needs little introduction – the numbers speak for themselves:
- Data from Statista shows that investment in Big Data is 10x higher in 2020 than in 2011.
- A 2021 study by New Vantage Partners found that 99% of surveyed Fortune 1000 companies were investing in both AI and Big Data.
- Research by IBM suggests that AI investment will exceed some $16 trillion by 2030.
- The Big Data analytics market is set to reach some $103 billion by 2023
As such, the technology that supports everything related to data is also developing at an exponential rate.
The CDP is one of the latest developments in business data processing, handling and instrumentation and are angled specifically at customer data.
Software development moves fast in this day and age, but still, every few years a new data product enters the market and disrupts the status quo.
We’ve already seen this with CRMs, email marketing platforms, productivity and collaboration software and all manner of data collection and warehousing platforms.
Now, the Customer Data Platform (CDP) enters the fray as yet another 3-letter acronym with great ramifications for businesses worldwide.
Table of Contents
What Are Customer Data Platforms?
Segment and mParticle probably created the original CDP category, and to be fair to them, despite CDPs being essentially a commercialised solution for managing and analysing customer data, the idea was still unique. Since then, other companies have pretty much renamed their original offerings to align themselves with what a CDP is thought to be.
There are other acronymised services popping up too, like Intercom’s ‘conversational relationship platform’ (CRP). It’s a pretty confusing environment in the business software engineering industry right now, likely because of the money being ploughed into it by major investors.
Let’s firstly address what a customer data platform really does.
What Does a Customer Data Platform Do?
In simple terms, a customer data platform is focussed entirely on the collection, organisation, management, analysis (sometimes) and implementation of customer data.
A CDP can’t do all of these things in its own right, but it can support each of these processes.
CDPs are simple at heart – they’re intermediary services that centralise customer data and allow it to be sent elsewhere in a clean, ready-to-use-and-implement state.
A big part of what makes the CDP unique is its ability to apply identity resolution and segmentation to customers, either using deterministic analysis or probabilistic analysis. This allows large quantities of disparate customer data with lots of different attributes to be combined into a single robust and unified aggregate of customer data.
In a post-apocalyptic society, the vast majority of the world is poverty-stricken. Governments order the redistribution of material wealth; everything from food and water to clothing, property, vehicles, medicine and commodities must be redistributed.
The problem is, without an intermediary agency – like a charity – there is no easy way to distribute and allocate all of this stuff. Plus, people have varying needs.
An intermediary is able to collect goods, classify them, and distribute them. This is the customer data platform. In fact, the customer data platform analogies well with many intermediary services that aim to centralise and distribute material from different sources to different locations.
This is all a customer data platform really is. The data goes in, the data goes out. It’s stored securely using grade-A cloud storage technology and the provider also offers a suite of in-house tools that allow businesses to clean the data, analyse it and use it.
The In:Out Flow of CDPs
So, a CDP is an intermediary service. It allows businesses to gather customer data and deliver it to other locations. It’s a sort of interface.
Customer data platforms connect to a huge range of data sources, the broadest of which are websites and apps, customer service and support systems, sales channels, advertising channels, surveys and a myriad of third-party tools used for sales and marketing.
The destination of this data could be any of a number of third-party services. Data can be sent for analysis and reporting, or used to train ML models, or analysed for business intelligence (BI) purposes. Customer data could be used to develop AI software, or implemented in order management systems (OMSs), CRMs or pretty much anything else that can receive it.
The source and destination of data that uses the CDP as an ‘interface’ are both fairly open-ended. The CDP is a junction, a pit-stop.
Below is a diagram that perfectly describes the in:out design of the CDP.
CDPs Provide Control
Since CDPs are purpose-built for customer data, this is naturally where their strengths truly excel. Namely, a CDP allows customer data to be segmented and organised with abject ease. For example, marketing teams can effortlessly segment users based on groups of attributes and pipe this data elsewhere.
An example would be creating an advertising audience within the CDP using existing customer data, visualising this audience and their various attributes within the CDP and sending it to say, Pinterest, Facebook or Instagram.
CDPs are often described as providing a ‘single source of truth’. What this means is that they act as centralised databases with their own data governance and data democratisation controls. They’re excellent for businesses looking to undergo some level of data democratisation, allowing employees from across the business or organisation to use the CDP as a self-service customer database. Since CDPs are quite intuitive and simple, they can also be used and understood by those with lower technical data or IT knowledge.
Want to improve your data skills?
See the best data engineering & data science books
When Do Businesses Need a CDP?
CDPs are exceptionally useful when a business is collecting data from multiple channels. Disparate data engineering slows businesses down and a CDP offers a pretty seamless way to manage vast volumes of customer data collected from multiple channels.
Consolidating data in one location is very powerful in itself. CDPs ensure that multiple teams from advertising to marketing and customer service can access that ‘single customer view’. CDPs offer a streamlined and holistic approach to customer data engineering.
Choosing Customer Data Platforms
CDPs are now not in short supply. As soon as demand surfaced for these low-code, centralised customer data platforms, many software publishers jumped on the bandwagon and developed their own CDPs. Lots of not-quite-CDPs also rebranded similar products as CDPs.
Resultantly, not all CDPs are created the same.
First-Party Data Sources
CDPs offer a vast selection of client-side and server-side SDKs for data collection and tracking from a range of sources:
- HTTP, API, Python, Node, etc
But, when it comes to connected platforms such as Roku, Alexa, Xbox, etc, not all CDPs provide native first-party support. Whilst this is a pretty niche problem for most, it’s still wise to first consider whether your primary data sources are indeed supported by the CDP. ETL pipelines and CDIs (customer data infrastructure) is also used to ingest and sync customer data from all available touchpoints.
Another point of differentiation between CDPs is their third-party integrations. These should be reliable and well-documented. The main contenders, Segment and mParticle, offer tons of integrations with analytics tools, marketing automation services, product analysis tools, business intelligence (BI) and other data processing services.
It’s also worth considering whether the depth of each integration is sufficient for the business’s uses. One example is Mixpanel, a product analytics platform. Most CDPs support Mixpanel as a data destination, but what about as a data source? Since most CDPs don’t support Mixpanel as a data source, you’d have to push data using the HTTP API.
CDPs are designed to be useful for any and all teams that require the access and use of customer data. That means they need to be clean and easy to use, accessible and intuitive enough to allow IT to take a hands-off approach to allow departments to access them of their own accord.
CDPs are adept at taking customer data from any and all customer channels and unifying these. Customers will, therefore, be identifiable across all connected sources.
So, with unified, centralised customer data – the ‘single customer view’ or ‘single source of truth’ – the following should all be readily accessible:
- Event data taken from any and all connected channels (e.g. websites, apps, IoT devices like fitness trackers, brick-and-mortar stores)
- Interaction data from social media and search
- User traits; demographics, personas, preferences, etc
All of this data should be readily searchable and accessible without too much technical hassle. Of course, the CDP will need to be properly configured first.
With this data, segmentation based on queries is simple. Businesses can create audiences and send these to other tools and platforms.
The use cases of customer data platforms are diverse:
- Unify customer profiles and gain a 360 view of customers, as we’ve already covered
- Segment customers using easy-to-use queries and drag-and-drop menus
- Use data to personalise experiences across different channels
- Combine online and offline data for true omnichannel data implementation
- Predictive analytics; predict and prevent churn, predict high-value customers and leads
- Retarget adverts to optimised audience segments
- Customer journey optimisation; optimise omnichannel marketing strategies
- Build recommendations engines and loyalty programs
- Improve activation and onboarding
What Isn’t a CDP?
Let’s take a look at what a CDP doesn’t do, as they’re fairly similar to other products in the customer data industry.
CDPs are not CRMs
Customer relationship management software is not the same as CDPs. CDPs can be connected to CRMs, though, but CRMs are more focused on the specific relationships customers have with a business and aren’t focused on creating a single source of truth or 360-degree view of customers for implementation in a variety of other platforms.
CDPs are not DMPs
Data management platforms (DMPs) are somewhat similar in architecture to CDPs, but are more focussed on NPII like IP addresses and browser information. They’re not optimised for use with customer data, though, and won’t create the same unified customer view as a CDP. DMPs also lack the same integrations as CDPs.
CDPs are Not All-in-One Tools
CDPs do have some native built-in tools and developers will likely add more to CDPs to reduce reliance on integrations. But right now, CDPs do not provide all-in-one functionality for analysing or implementing customer data.
CDPs and Data Democratisation
CDPs are excellent for creating simple and unified self-service data tools. Read up about data democratisation – the process of businesses disseminating their data products to a wider internal audience – here.
Security and compliance are probably the elephants in the room when it comes to CDPs, data access and data democratisation. Most CDPs come with advanced permissions and data governance controls to stratify data for security purposes.
Summary: What Are Customer Data Platforms
Customer data platforms are ‘hot software’ and are marketed as a must-have for any business that wishes to use customer data efficiently and effectively. Is that a fair summary? Probably, yes – customer data platforms are genuinely very handy and useful. But, if a business’s existing customer data architecture solves its main problems then it’d be fair to remain sceptical about making the switch.
CDPs excel when businesses are collecting large amounts of customer data from multiple channels. In the omnichannel retail era, businesses might induct data from so many sources, thus leaving their data architecture somewhat disparate and strewn.
CDPs solve these issues by centralising complex customer data and creating a ‘single view’ or ‘single source of truth’, as it’s often described. Customer data becomes easily accessible and readily usable in virtually any integrated service. This saves both time and money – the two main ingredients of a successful business.
Are Customer Data Platforms Worth it?
Customer data platforms (CDPs) are a new wave of data products designed for businesses that use and manage vast quantities of customer data. Customer data is complex and may contain different datasets depending on the channels(s) it’s inducted from. Resultingly, customer data can become messy and hard to combine for a 360-degree view of customers. CDPs are a customer-data-optimised tool for storing, managing, segmenting and using customer data collected from multiple sources or channels.
Are CDPs Any Good?
Customer data platforms do not seem like a flash in the pan – they’re not a fad. CDPs provide genuine value to businesses aiming to achieve a single view of their customer data. They come with a vast array of integrations making them easy to integrate with other data platforms. CDPs are built and tuned specifically for customer data, so that’s unsurprisingly where they excel.
What are CDPs?
Customer data platforms – CDPs – are data products designed to centralise customer data in one well-integrated database for viewing, managing and using customer data. CDPs are built and optimised specifically for customer data and are ideal for businesses that collect and use customer data from multiple channels.