To CDP or not to CDP...that is the question

To compete effectively in today’s fast moving world, most organisations understand they need to provide a relevant and joined up customer experience across channels, requiring collection and management of real-time customer data. However, what’s become clear to us in client conversations over the past 6-12 months is that a new tech stack ‘hero’ appears to be waiting in the wings to solve all the problems of the increasingly challenged Marketing department in 2020: the Customer Data Platform (CDP).

Now with a dedicated Institute, CDPs are experiencing a rapid rise in popularity and demand as marketers continue to seek a way to make sense of the volume and variety of customer data at their fingertips and create a usable single view of the customer that powers and orchestrates multi-channel personalised marketing at scale.   

Sound familiar? Weren’t we singing the praise of the DMP just 2 years ago as the champion and saviour to all our data woes and challenges? Is this the emperor in newer clothes? What has changed and why is the CDP any different? In this article I outline where the CDP can play a valuable role and reiterate the need for clients to be very clear exactly HOW and WHERE they would use such a tool, and how they will pilot it, before they jump with both feet into  a very expensive commitment.

CDP or DMP…or both?

Ultimately, both the CDP and DMP are tools to help a marketer collect, analyse, manage and use data to deliver more relevant and targeted marketing to audiences. A CDP is defined as:

‘A marketer-managed system that brings data together to create a persistent, unified customer database, accessible to multiple operational systems.’

It can connect all types and sources of customer data, whether internal or external, structured or unstructured, to create a comprehensive customer context that marketers can act on in real-time.

The focus of the Data Management Platform (DMP) is to aggregate behavioural digital data and categorise it into taxonomies which can be used to build addressable and buyable audience segments, primarily to enable the buying (and selling) of programmatic media.

CDPs are distinct from Data Management Platforms (DMPs) because where DMPs focus on anonymous data (primarily cookies and device IDs) to create audiences made up of multiple customers, CDPs can store individual level known personally identifiable data (i.e. data the customer actively and knowingly gives to brands such as email and contact number) meaning they can connect with more applications to support person level and audience level activation across multiple channels.

Despite significant effort by the large MarTech vendors for DMPs to take on the role of holistic data integrators providing identity resolution across channels, because DMPs focus on anonymous cookies only (mostly third party), meaningful identity resolution and multi-channel activation could not be achieved. Given this, we found many DMP users (especially those with a very limited owned data asset) experiencing disappointing campaign results, largely because they a) had very little practical input in defining the audiences and b) the third party data used to define the audience was of dubious origin and often poor quality; meaning the end messages delivered to customers were not at all relevant. The consequence for brands was rather than improving targeting, they found their digital marketing contained ever increasing levels of poor targeting, irrelevant messaging and disappointing results. Never was the concept of ‘garbage in, garbage out’ more appropriate.

The Rise of First Party Data – Quality over Quantity

At the same time as these DMP problems surfacing, GDPR compliance requirements led to a marked increase in focus on first party data, as clients shifted to transparent and open data collection, favouring quality over quantity of data and cutting back the reliance on third-party data.  Indeed, our own 2018 ‘Data-driven Marketing’ research report, delivered in partnership with the World Federation of Advertisers (WFA), found that 67% of clients saw first party data as business critical vs just 11% saying the same for third party.

Martin Sorrell, Founder & Chairman of S4 Capital and former WPP CEO, also recently announced that “the battle in Marketing today is for first party data”, and it is no secret his next major acquisition will be one that meets his desire to “have a significant first party data asset”.

This inexorable rise in the focus on first party known data is driving the boom in CDPs.

It’s also interesting to note that at the last WFA Digital Governance Exchange client forum in October (’19), Jerry Daykin (EMEA Senior Media Director at GSK) and Gerry D’Angelo (Global Media Director at P&G) led a scenario planning workshop called ‘Preparing for the data-limited future’ discussing the critical questions advertisers need to be asking themselves to prepare for a future with limited data.

The Unified Data Platform

So, given the above, what is needed is a unified data management set up within the organisational tech stack that provides a reliable single view of the customer (i.e. CDP functionality) that can be analysed and modelled to provide an understanding of people and audiences for subsequent orchestration and activation across touchpoints via tools such as the DMP and DSP for media and the CRM for direct comms.    

Indeed, it’s hardly surprising to see Salesforce launch their new ‘Customer 360 Truth’ solution at the recent Dreamforce 2019 conference, which contains a CDP element (‘Customer 360 Data Manager’) and a DMP/media element ('Customer 360 Audiences') aimed at giving clients the best of both worlds.

What to Do Next?

We would counsel extreme caution at this point and a check on your reality.

The aforementioned Salesforce product is what we call the ‘Ferrari solution’ i.e. a top-end and complex solution that is likely only needed by the most advanced and highly data savvy (and data rich) organisations.

Every business is different and designing a fit for purpose answer to your needs requires a clear definition of HOW and WHERE a data management solution will be used to deliver your business and marketing objectives. As the CDP marketplace explodes (vendors alone have grown from 23 in 2016 to over 100 in 2019, in a market that is forecast to be worth over $1bn in 2019) it is no wonder Forrester released a paper in October 2018 titled ‘For B2C Marketers, CDP’s overpromise and underdeliver’.

In our recent experience, clients are still jumping in too quickly and without undertaking the due diligence needed to define the use cases and value to be delivered by the tech stack. This is costing them £millions in failed or stalling initiatives because they either over-invest and end up with solutions that are over-specified and unfit for purpose or buy solutions that sit in IT with no usage by Marketing thanks to lack of understanding and know-how.

We have consulted with multiple enterprise clients across industries, and we have learned that the following 5 step process will set-up you up for success:

Step 1: Identify use cases and desired consumer experiences

Informed by clearly defined business and marketing objectives, first define and prioritise the end customer experiences you are trying to deliver across the path to purchase. Our approach is never to start with the capabilities of the technology, instead we help clients consider how they need to transform each customer touchpoint across the entire end-to-end customer journey. Only once this is clear do we look at the role data and technology will play in delivering and enabling this future state.

This approach helps clients define a clear set of business requirements and a functional specification that will ensure the end solution addresses the use cases and not whatever the latest CDP or Unified Data Platform can offer. It also helps prioritise activities as we move into Step 2.

Step 2: Design the Optimal Operating Model (do this before vendor selection and go-live!)

Experience has shown that its crucial to consider the legacy internal organisational structure and existing ways of working and operating model in which the tech will operate. The vision of unified data and multi-channel real-time automated orchestration may not fit with your existing processes, structures and people skills and so changes may be required internally to these areas before go-live. Flicking the switch on new IT is not transformative when it fails to be adopted by the business!

It will be critical to re-engineer key processes, train people (internal and agency partners) and change critical behaviours, set governance checks and communicate appropriately before go-live. Also consider whether any global or regional teams can develop any common standards or guidelines to support local markets.

As Forrester stated in their aforementioned report “For many B2C Marketers, CDP’s are a band-aid solution on a much larger challenge” – make sure you address the internal changes needed as a priority.

Step 3 - Select the Right Vendor – Fit for Purpose, not Ferrari

With so many vendors out there, both big and small, don’t immediately jump to the enterprise level vendors. Consider the smaller operators first, especially if your needs are basic and then grow as maturity and capability grows over time. We have recently helped a multi-billion dollar CPG business select a very small CDP vendor because it was the best fit for their purpose and use cases. Start small, learn and then scale fast as needed.

The same goes if you are tied into an Enterprise mar-tech vendor (i.e. Adobe, Salesforce) – consider how to test and learn before biting off (and paying for) more than you can chew.

Step 4: Set up a Proof of Concept and Pilot Plan

Based on your priority use cases, your internal operating model and your chosen vendor, set up a pilot program that will let you test the chosen vendor/s (maybe even test multiple vendors before locking in on one) and learn which is best suited to your specific requirements and ways of working. Start small and then scale fast. Successful transformation is more successful if it delivers value in the short terms as you build towards the longer term vision. Agile planning is key, where we plan for delivery of a ‘minimal viable product’, leading to early commercial impact.

Step 5: Test, Measure & Learn

Implementation is not the end point. Keep testing and learning and be super critical on measuring the value of CDP driven activity with clear learning objectives and KPI’s to track impact and success.

This is a complex topic and one that needs careful consideration to ensure the tech stack becomes an enabler of your data-driven Marketing objectives and not a barrier. Don’t hesitate to get in touch if you would like to discuss your plans or your existing challenges.

TCF are specialists in helping develop world class marketing organisations and build the operational plans to realise their business objectives. For more information on how TCF can help your organisation, please contact nick.broomfield@thecustomerframework.com.