The German startup HelloFresh is in a transition towards a data mesh architecture, coming from a very centralized analytics-heavy perspective. While they are still very much in transition, I like a couple of things about their approach and would like to highlight them.
First of all, HelloFresh identified its flywheel
and its data as a strategic asset. That’s very much in line with my thinking, see e.g. my post on Good Data vs. Bad Data
“Good Data, derives the data strategy from the company strategy, feeding into the datacisions cycle. Bad Data has lots of “initiatives” flying around the company, without a coherent data strategy.”
They have a clear vision of getting all data, operational and analytical, into one “bucket”, one domain, while they want to keep conformity by normalizing dimensions across different domains.
Above all, I really like what they started with, the basics, that is data ownership and data literacy. They realized, without ownership & knowledge there is no way forward into a data-focused company architecture.
Anyway, I enjoyed the talk so check it out if you’re also in transition or just want to get a feeling of what kind of journey you’re embarking on because I guess you can tell from this short summary — it’s a long and painful one, but well worth the effort.