These days, Customer Data Platforms, or in brief CDPs, are one of the most discussed topics in the CX industry. Last time I looked, the CDP Institute counted more than 150 CDPs of different flavors that target different business challenges.
This raises a lot of questions, chiefly: What is a CDP? Which business challenges do companies solve with a CDP and how do they approach the solution, i.e. how do they proceed implementing one?
Let’s start at the beginning. According to the CDP Institute, a CDP is “packaged software that creates a persistent, unified customer database that is accessible to other systems”.
Figure 1: A simple model of CDP functionalities
Functionalities offered by CDPs reach from simple data gathering and unification to the activation and even active usage of information derived from customer data. This already shows the different business challenges that can be addressed with a CDP. It also shows, that the business challenges can be addressed in a way that provides increasing value to the business.
As an aside: A good CDP always bases on a solid foundation of customer identities and the consents that customers have given.
The second important question is how a CDP can or should be implemented. What is a viable approach?
To learn more about these questions, we spoke with Mario Kurmann, Senior Product Manager CRM at Migros Fachmarkt AG in Switzerland. Mario was responsible for the initial CDP implementation and now for its continuous improvement. Migros is one of Switzerland’s biggest retailers. Fachmarkt, as part of Migros, basically operates five independent brands in the areas of sports, electronics, home improvement, furniture with online shops and brick and mortar presences. All these brands, and the Migros group overall, have separate data silos. One more silo is the Migros loyalty program, Cumulus.
All these data silos were aggregated in different buckets, e.g. data warehouses, then serving different purposes. The basic question that arose from this was: Ho can we work with this treasure trove of data that also includes transactional, surf patterns and behavioral data.
Consequently, the first main objective of the implementation was to bust the silos and to harmonize customer data across these different brands and the loyalty program. The second goal was to then be able to activate the data and to use it operationally, always looking at it with a CRM perspective.
Long term objectives of the implementation are, of course, increased loyalty, more revenue and more precise customer engagement. However, according to Mario, the first implementation had another strategic goal altogether: Learning the capabilities of a CDP by exploring! After all, a CDP is rather a technical than a business component.
As trivial as it sounds, maintains Mario, Migros started with two simple use cases: re-engaging with customers in case of an abandoned shopping cart and personally welcoming new in-store customers to the brands. This with the background of Migros not having a clear customer database with clear and unique customer IDs. Cumulus, for example, works on household level and not with individual customers. How to deal with this is one of the questions that still needs some thought.
Migros looked at the CDP implementation as a business project, albeit the tool is of rather technical nature. As said above, the team started the implementation with simple use cases that target at improving the customer experience at important touch points via defined use cases.
Talking of the team: The team was deliberately kept slim, too. In alignment with Fachmarkt AG and Migros itself, this team was empowered to approach the topic using ecommerce data, also together with an implementation partner that already implemented the online shops.
The team maintained close ties with the business units, also educating them what the CDP can help with and where its limitations lie.
Communication, communication, communication
It is very important to have and maintain an alignment with the various stakeholders. The degrees of freedom and the timelines that the project team has as well as any limitations need to be clear to the project team as well as to the management.
Data governance is an important topic, as a CDP is privacy relevant. The availability and usability of the used data needs to be clarified with the legal departments.
It is similarly important to stay in constant conversations with the customers, managing expectations and educating them about capabilities of the CDP and the scope of the implementation while learning about the requirements to fulfil.
Actionable data is not equal to engagement
One of the big early learnings is that a CDP is not an engagement tool. It is essentially a database which helps activating the data. The actual engagement needs an engagement tool. It proved helpful to work with a separate engagement tool that didn’t conflict with other tools that were already in production use for the two use cases. The team chose e-mail as the communication channel. In doing this, the team staid independent of the daily operations, which sped up the project considerably.
Don’t boil the ocean
Starting with small, well defined use cases helped as well. These use cases worked with a small number of upstream and downstream systems. This kept the complexity under control and led to having consistent data coming into the system. This helped in delivering results fast. It also helped in gaining more understanding of the CDP as such.
Know your infrastructure
The ability to work with internal as well as with external partners was another point that increased the ability to deliver and the delivery speed. This way, there is no need to re-engineer the infrastructure and, even more importantly, a trustful relationship is already established.
One little downside
The one thing that was not established early is the harmonization of data, as the ingested data was already quite clean, because of the selection of upstream systems. This is something that will be resolved in later project stages.
In consequence, there was a win for all participants. The use of commercial use cases kept the project effort low while quickly showing results. The project showed good opening rates of the sent mails resulting in increased revenue via an increased conversion rate. This raised the appetite for more.
Based upon the established success and value, the next steps will now tackle more sophisticated use cases including campaign automation and helping customer service.
This text is based upon the CXChangersTalk with Mario Kurman, If you are interested in the original, you can watch it here