Care factors for a healthy DMP program – Part II – Data

In continuation of the Care factors for a healthy DMP program Part I – Overview article, I would like to explain the data related care factors in this article.

Data comprehension

This involves understanding of your data schema (structure), data definitions (field descriptions), data quality (reliability), data relationships (like hierarchy), data relevance (for personalisation) and data velocity (update frequency). I can not stress it enough but if you do not have someone who really understands your data as a core member of your DMP program, your audience will always lack the trust factor and it will be very difficult to scale the DMP program after initial install.

Audience structure

Spend enough time figuring out your audience strategy early in the DMP journey. It will save you lot of rework and frustration later on. Do not restrict yourself to top 5 or 10 personas, but think ahead and get detailed with defining your audience strategy. Audience building in a DMP should not be an ad-hoc approach managed by one or two resources, it should be a planned undertaking approved by all relevant teams who plan to leverage DMP in future. Many organisations who tend to mismanage audience taxonomy early on, may soon end up with a sloppy DMP which no one internally trusts or wants to use. You can read my article which talks about a data to audience approach.

Data privacy

If there was an award for most searched and least comprehended term this year in martech industry, GDPR will be a top contender. Data security and customer privacy are more important than ever for your DMP program. For example, Audience Manager provides you with features like data export controls (Adobe patented), role-based access control (RBAC), time to live (TTL), IP obfuscation and custom exports to comply to these regulations. But features are only useful if they are used. Some clients still use PII like email for Id stitching, and may not be aware of features that can help them respect customer’s data privacy. This also causes a lot of delay with internal security approvals who have no clue why and how customer data is being requested to be sent to a big data cloud storage called DMP. So, having a good understanding of DMP data flow architecture and features that support data privacy, you will hopefully have a cleaner, quicker and compliant DMP instance.