November 3, 2021 | Written by: Kuntal Hansaria
Share this post:
Nowadays, many consumer facing companies are building / buying a Customer Data Platform (CDP). There are literally 100s of vendors in the market currently and more continue to emerge. Some studies project it to be a USD 15 Billion market by 2026, with a CAGR of 35%!
What is CDP?
At a high level, CDP includes ability to ingest data at scale (real-time & batch), creating a 360-degree customer view (using deterministic / probabilistic techniques to identify and join customers across different data sources), space to build data science models for segmentation, next best action, etc. and data connectors to activate these segments across various marketing platforms and digital marketing solutions. Vendors differentiate themselves citing number of data sources they readily connect to using APIs, real time ingestion, auto machine learning / predictive modeling features, number of connectors for marketing campaign activation, etc.
The CDP trend has accelerated recently given recent changes in market dynamics.
- Post COVID digital transformation – Digital channel’s importance in marketing mix has increased tremendously. This has led to increased need for managing prospect and customer interactions across owned and paid digital channels.
2. Death of third-party cookies – Given data privacy concerns, support for third- party cookies has been decreasing across key browsers. The old ways of cookie- based targeting on ad platforms are no longer reliable. Earlier, one could select an audience (list of cookies) on any ad platform and target them. Now, this option is less reliable as key browsers no longer support these cookies. Companies need to rely on their own cookies (first-party data). Simply put, CDPs enable tracking, storing & marketing customers using first-party cookies.
3. Hunger for intelligent interactions – Companies want to leverage power of AI to interact intelligently with their customers. They aspire to store each interaction, join them across channels and use them for intelligent interventions.
While the above forces are creating a demand for CDPs, it remains to be seen whether this is just another wave or are organizations really positioned to leverage the power of this tool. CDPs by themselves are just a platform. You need to nurture them with the right data, deploy data science resources to build models and then link these attributes to your marketing efforts. Unless this pipeline and effort is well synchronized, CDP investments may not yield results. Some CDPs claim to have in-built, pre-built models that could do away with the need for data scientists. Well, this is probably as good as giving a self-driving car to someone who does not know how to drive. One can make some progress, but the chances of accident remain high!
One needs to take a hard look at one’s requirements before embarking on a CDP journey. Most platforms offer much more than one can consume immediately. The lure of innumerable data sources to connect to and activate may be unfounded unless your current marketing efforts are as extensive currently.
Many organizations have basic requirements:
- Customer base on CDP <10 million
- Customer data sources <10 (mostly internal, with batch ingestion)
- Deterministic 360 view of the customer (using phone number or email as key to join across sources)
- Handful of predictive models for specific use cases
- Activation of customer segments / models using existing marketing stack options
- Overall, understand customer’s needs and talk to them when & where it matters
For basic requirements, many out-of-box CDP platforms could be an overkill. This could be achieved even with a custom-built stack using serverless services on cloud platforms; in fact, this IS a growing trend currently.
If your needs are bigger: Tracking interactions across many more & increasing real-time / near real-time sources, augmentating with external data, array of models specific to segments, products, geographies, etc. and activation on various walled and open digital advertising platforms; then you could then be better served by premium CDP stacks out there.
Data residency is also a big talking point for CDPs as most data on them is personally identifiable. For some, you can implement them on a dedicated instance within your private / public cloud helping keep the data within your cloud; however, the support for this in India is currently limited. If your business has a hard requirement of storing data within India, then most out-of-box CDP options may not work for you currently.
CDPs are here to stay and thrive. They have existed in some form even before. However, like in case of any tech product, companies using them can thrive only if they think through their requirements and usage imperatives well enough!
Data and analytics consulting services
Accelerate cognitive-driven marketing transformation with customer data platforms
Try Virtual Garage