Construction workers observe the work on top of a tall building.
PHOTO: Shutterstock

The promise of customer data platforms, or CDPs, is an appealing one for marketers: unify customer data into an accessible and regularly updated golden customer record.

That’s how researchers at Omdia painted the CDP picture in their “Ovum Market Radar: Customer Data Platforms ” report published in November. (Ovum is now Omdia.) 

Beneath the appeal lies a significant amount of work. CDP implementation is an effort that requires extensive data-quality checks, clearly identified case studies and consultation with technology stakeholders beyond the marketing function.

“The biggest mistake companies make with CDP implementations is not bringing IT and marketing teams together to make sure the implementation delivers the promised benefits of the investment,” said Raj Kini, director of professional services of Arm Treasure Data, which offers CDP software solutions. “Just like a pilot and co-pilot running through an extensive checklist prior to take-off, companies should assign executive leads from IT and marketing to work together on their CDP implementation.”

Identify Use Cases Before Selecting a Vendor

Organizations need to get a handle on CDP implementation use cases before selecting a vendor, according to Venu Gooty, director of strategy services at HGS Digital. “Many companies make the mistake of evaluating CDP vendors just on technical capabilities without first identifying what the marketing and business needs and use cases are,” Gooty said. “The use cases could vary significantly and different CDP vendors have strengths that meet certain use cases better than others.”

Gooty cited some example of use cases:

  • Email segmentation and campaigns
  • Web personalization
  • Programmatic ads optimization
  • Attribution modeling
  • Targeted social campaigns
  • Loyalty based personalization

Marketing must make the organization’s goals clear, Kini said. Identify which use cases matter most, why they matter and what ROI they are expecting. Examples include better user acquisition, less churn, stronger conversion rates, adherence to data security and compliance standards. Use cases must have measurable KPIs so progress can be tracked and real-time adjustments made, he added.

Related Article: Developing CDP Use Cases: A Guide for Marketers

Create a Customer 180 View

Many organizations talk about their desire to get a 360 degree view of the customer but meeting that goal takes multiple steps and the unification of many data sources, according to Gooty. He suggests first building a few integrations to create an intermediate view (customer 180 view) that can provide an incremental value to the business that’s not available via other systems right now.

“As you add more data sources across phases,” Gooty said, “ROI can continue to increase as the scope of use cases grow.”

Check the State of Your Data

David Raab, founder of the CDP Institute, said the no. 1 CDP implementation tip “by far” is to check the state of your data. “Unexpected data access or data quality issues can easily slow down an implementation or force a change in direction from an intended use case,” Raab said. “The corollary is: be prepared to spend more time and budget on data wrangling than you expect.”

Marketers usually underestimate the amount of effort involved in data-hygiene activities including data cleansing and unification, according to Gooty. “Since the different source systems may have different data-quality challenges, combining that data together into CDP would require a well thought-through cleansing and unification process,” he said. “Perform that assessment first if you need to create a financial business case.”

Related Article: Lessons Learned From CDP Implementations

Unify Data Sources

During the process of checking the state of your data, organizations should identify current data usage and the most important data sources. IT must optimize data integration, standardization and access. Kini said many organizations overlook this key step and fail to bring a subject matter expert to the table. 

“Not having a clear picture of all the important data sources and relevant data elements will often lead to lack of clear insights into a user’s behavior and subsequently leads to lack of results and adoption of the CDP," Kini said. "The age old saying ‘garbage in, garbage out’ holds true with the CDP as well, and it is important to take a step back to identify the data sources and then unify them with the right approach.”

Gain Maximum Value from Existing Systems

Understanding the different capabilities that CDP, data lakes and CRM would bring to the table and how they interact with one another is crucial to architecting the right customer data technology stack, according to Gooty. One definition of data lakes, according to CDP vendor mParticle, is "collections of data, usually in the same form as the original source systems."

CRM and data lakes already provide unified customer views that are internal facing. Designing your CDP to complement the value existing systems already provide can improve corporate-wide adoption, Gooty said.

Related Article: Beyond CDP or Bust: The Looming CDP Reality

Respecting Privacy and the Process

Data must always be used appropriately. In addition to adhering to government regulations and security and privacy standards, procedures during a CDP implementation should reflect the company’s values and desired customer relationships, Kini said.

“Organizations also need to identify who within the organization has access to what parts of the data,” Kini said. “Giving everyone access to PII (personally identifiable information) can lead to careless decisions such as leaving unencrypted data on a non-secure server, data breaches, etc.”

Assign a Dedicated CDP Manager

When starting the process, it's also important to assign one business owner to oversee onboarding and training programs as well as the ongoing success of the program, according to Kini. The success of the CDP depends on not just the implementation but the subsequent onboarding and adoption by the marketing teams.

“Change management sometimes can be a problem and identifying these things up front in the process is important for the success of the CDP,” Kini said. “Having a clearly identified owner of the CDP can help with the onboarding of the CDP and ensuring adoption. This CDP owner can be part of the marketing organization or IT or sit in a program management group, but should be in a position to make decisions and change course when needed.”