Data Products Are Not a Technology Problem

Every capability you need already exists

Schema registries, quality monitoring, semantic layers, access policy engines, discovery catalogs — every technical building block needed to run a data product operating model is mature, well-understood, and available off the shelf or as part of a platform like Contivra. And yet most enterprises that attempt data mesh do not get past a handful of half-adopted products.

That gap is the tell. If the technology were the bottleneck, mature tooling would have solved it by now. The bottleneck is almost always organisational: unclear ownership, misaligned incentives, and no shared operating model for what 'done' looks like for a data product.

The pattern behind stalled initiatives

The common failure sequence looks the same across industries: a platform team stands up excellent infrastructure, domain teams are asked to publish data products on top of it, a few products get built as one-off efforts, and then momentum stalls because there was never a clear answer to who is accountable when a product's quality degrades six months after its original champion moved to a different project.

What actually unblocks it

The initiatives that break through this stall point share a common move: they stop treating data products as a technology rollout and start treating them as an organisational design problem, with the same rigor applied to ownership, incentives and operating cadence as to the platform architecture itself. Concretely, that means naming accountable owners before building, funding maintenance as ongoing work rather than a one-time project, and defining shared quality standards before the first product ships rather than after the first complaint.

The technology's real job

None of this means the platform doesn't matter — it does, because good tooling is what makes the organisational model sustainable rather than a constant manual burden. But the platform's job is to make the right organisational behavior the path of least resistance, not to substitute for the organisational decisions that have to be made regardless of which tools sit underneath them.