Bold prediction #2: Data Products will have united the world(s)

In every company there are well established data product in the ownership (yes, real ownership and data mesh) of the data producer and the data consumers pretty without explanations can use the data, of course because it is well described. Operational and analytical worlds are united by data products.

Data Producers can publish & update the data products on consumer requests within 1 day (of course backward compatible) and as these products are used in the whole organization as standard, no need for maintenance of hundreds of interfaces, data loads and data duplications. 

For Data consumers the cognitive load of data transformation & understanding the source is minimized (yes I saw cases where data consumer are directly querying on the mainframe e.g. for payment data) and they get the data needed served on a golden plate for their business processes in highest possible data quality & consistency.

Gartner is forced to revise their hype cycle for Data Management and declare Data Mesh is NOT “obsolete before plateau”.

Reality check 2025

Obviously, this will not be the case and everything above in some similar flavor was tried over the last 10 years.

BUT the reality is that I am currently in a handful projects in several countries where we design data products to have clearer ownership, better data quality and a consumer aligned representation of the data. In one case even as a prerequisite for mainframe transformation to cloud.

In another banking case the original data producers in future are responsible for the complete transformation into centrally governed data products. From there on the data consumer is enabled to consume in a self-service mode. The advantage compared to current DWH bottlenecks in banking data architectures? (more information will follow in a few weeks about that project)

• No delution of ownership by a central DWH/transformation team between data producer & consumer
• faster time2data for changes consumer requests by decentralized implementation (instead of release monsters in the DWH environment)
• central data team shifts is focus to provide and operate the central platform/marketplace, necessary governances and new technology capabilities

Currently we have weekly meetings about data mesh, data products and state-of-the art data governance concepts.

Ad. Hype cycle: one very interesting question from participant of the executive event: “now we have 18months of genAI hype, why don’t we see productive applications and cases yet?"

At least data products/data mesh are addressing challenges companies currently have instead of having an answer and searching the problems/cases 😉