How to prepare for your Customer Data Platform implementation

As companies across industries focus on building strong relationships with their customers, the ability to deliver personalized experiences and protect data privacy is becoming increasingly important. 

Isolated data sets–when every tool operates with its own data plan, identity architecture, and compliance workflows–are one of the biggest blockers to using customer data productively and securely.

Many companies are eliminating data silos by implementing Customer Data Platforms (CDPs) at the infrastructure layer of their tech stack. An infrastructure CDP is a single customer data system that centralizes data collection, quality management, and identity resolution, and makes it easy to connect that universal data set out to any tool in the stack via a vast library of server-side integrations. This means that whether you’re building a dashboard in your analytics tool or handling a customer support ticket, you’re able to access the exact same data and customer identifiers.

Implementing a Customer Data Platform requires a fundamental shift in the way you think about and operate customer data infrastructure across the organization. As you may have supposed, it’s by no means a trivial task.

To help, we’ve put together these three steps you can take to prepare for your Customer Data Platform implementation and maximize your return on investment.

Customer Data Platforms provide the infrastructure that enables you to realize your customer data strategy, from standard use cases to your wildest data dreams. To ensure your CDP is set up to meet the right goals, it’s important to clearly define your data strategy.

Set your customer data goals

Customer data can be leveraged in an infinite number of ways, from powering marketing campaigns, to improving customer support, to building customer intelligence into the product experience. 

To make the most of your data, it’s important to pull cross-functional stakeholders together and evaluate all the ways the data will be leveraged across various departments. Next, you’ll want to establish the specific use cases and KPIs that you plan to execute or optimize through the adoption of a CDP so that you can measure impact. You may also want to look out for use cases you are already executing that could be made more efficient with a CDP.

Prioritize and plan phases of implementation

A Customer Data Platform can help you accomplish the most advanced data use cases, such as machine learning recommendations and building real-time customer data intelligence directly into your product experience. 

This doesn’t mean, however, that you need to begin using your CDP for these use cases overnight. The most successful CDP implementations we’ve seen have followed a crawl, walk, run approach that accounts for the design, implementation, training, and activation needed for each use case. 

To follow this framework, clearly define implementation stages based on your teams’ needs, capacity, skillset and more. For the best results, define how your teams will measure success at each stage.

Highlight potential dependencies or blockers

Occasionally, there is additional work required beyond the implementation of a CDP to fully operationalize your customer data architecture. Internal systems may require updates before they can be integrated with your CDP. Major customer-facing features may need to be shipped before you implement CDP vendor SDKs. It’s important to identify these potential dependencies or blockers so that you can plan accordingly.

To maximize the value of your Customer Data Platform, it needs to be integrated seamlessly with your existing infrastructure and processes. Identifying where data comes from, and what information you’re tracking is important to success.  Determine what your universal data layer will need to include in order to service all consumers, and identify gaps in your current data set.  Plan your implementation so that the infrastructure you set up is comprehensive, robust and continues to serve you over time.

Identify all sources of customer data

First, identify the various sources of customer data and each source’s integration capabilities. These may be apps, websites, internal databases, or other tools in your stack, and each may have different capabilities in terms of how customer data can be collected and integrated. For example, data can often be collected directly from apps and websites via client-side SDKs, whereas backend databases will require server-side integration. Surfacing these channel-level requirements will help you define the scope of your CDP implementation.

Determine your identity resolution architecture

Building unified customer profiles requires the union of various identifiers across your data sources. To understand your ideal approach to identity resolution,  you should catalog what identifiers are available from each data source. Once the full set of identifiers is compiled, you should define which ones can be used to accurately identify registered users versus anonymous users, as well as which identifiers should take priority when multiple are available.

It’s also important to establish the situations in which you do and don’t want to merge known and anonymous profiles. For example, if a customer signs up and is linked to a known profile, will you unify their pre-sign up browsing activity, which is linked to an anonymous profile, to the new known profile, or will you keep them separate? Your profile merge strategy will depend on the nuances of your customer journey and data governance policy.

Define what your universal data layer will look like

One of the biggest delays to time-to-data-value is bad data quality. It’s highly recommended that you create a comprehensive data plan that defines the data points you’ll be collecting and naming conventions for each data point, across platforms and properties. 

As you create the data plan, take time to align with both data engineers and business users on what data to expect and in what format. Ideally, you’ll create and maintain a single dictionary or catalog for your customer data so that all teams can understand what the data represents. 

Best in class CDPs will enable you to upload your plan into your CDP and enforce it as data is being collected, and also to download your data plan in the formats that meet the needs of teams across the organization.

Identify customer data activation systems

Being able to get data out of your CDP is as important as getting data into it. For that reason, it’s critical to identify the downstream data activations systems that your team would like to feed high-quality customer data to through your CDP. These could be tools that you’re already using today, or tools that you’d like to start using once you have your CDP in place. Make sure to tie this step back to the use cases that you hope to execute with your CDP.

Once you’ve identified the tools that will make-up your customer data technology stack, make sure the CDPs you’re evaluating can support integrations with these systems, both internal and external to your organization.

Account for privacy regulations and considerations

In order to stay compliant with data privacy rules and regulations, you should review any regulations that apply to your company and account for those in the design of your customer data architecture. A Customer Data Platform can help you securely integrate your data from point to point, but it’s essential that your data collection and federation reflect both regulatory requirements, e.g. CCPA and GDPR, as well as your company values.

Ensuring your CDP project is both successful in the long run requires support from business leadership, the implementing teams and your CDP end users.  

Assemble a cross-functional customer data team

Successful deployment and adoption of a CDP requires coordination and input across many teams within the organization. A successful CDP implementation will not only serve to reduce marketing’s reliance on engineering teams for access and control customer data, it will also serve to break down information silos across teams.

It’s important to ensure that there is buy-in from stakeholders across marketing, product, analytics/BI, and engineering to achieve the best results in the design, development, and rollout of the CDP solution. The most successful brands have engagement from stakeholders across all teams that interact with customer data, collaborating on decisions around data planning, data governance, and implementation roadmaps.

Identify the business owner(s) of your Customer Data Platform

Depending on the scope of the project, we recommend you assign a person or team to promote the project, collaborate with the teams involved, and drive internal adoption.  

Implementing a CDP represents a fundamental shift in the way data is owned and accessed across the organization, since they allow for customer data to be centrally managed through a single platform. For this reason, it’s important to identify and assign a central business owner for the CDP. This person will not only drive the initial implementation of the platform, but also the rollout of new business processes and workflows across the various internal teams.

With a CDP in place, business owners of individual data activation platforms (such as email or analytics tools) will no longer individually own the collection of data at the source. In this case, there may be a need for a central owner of the data collection across all sources, to which the data consumers subscribe and provide requirements.

The CDP owner(s) may also be responsible for creating a new customer data workflow for the business’ internal teams. This can entail creating processes to manage new data sources, data requests, and data consumers as they develop.

Identify resources for key CDP roles

Implementation and adoption both require time and personnel.  For a timely setup, you should identify people for each of the following roles:

Enable your people!

Once implementation is complete, the biggest blocker we often see is lack of training on for with the actual CDP end users. It’s important to prepare CDP users for tasks such as audience building, data forwarding, and profile lookups so that they can make the most of your CDP.

Certain CDP vendors will provide customized onboarding and training sessions that help you train team members on CDP use cases.