What Capital Markets Teams Get Wrong About Data Platforms

Capital markets teams are, rightly, proud of their engineering. The data platforms they build are often genuinely advanced. And yet a familiar pattern recurs: the platform is impressive, the investment is real, and the business still relies on spreadsheets and side channels. The problem is rarely technical sophistication. It is fit.

A platform built primarily to satisfy its builders optimises for the wrong things — elegance, generality, technical purity — while under-serving the traders, risk managers and analysts who are supposed to depend on it. If using the platform is harder than the workaround, people will keep the workaround, whatever the platform can theoretically do.

Adoption is the only real metric

The platforms that succeed start from the decisions the business needs to make and work backwards to the data and interfaces required. They treat adoption, not architecture, as the measure of success. They accept some technical compromise in exchange for being genuinely usable — a trade many engineering cultures resist.

None of this diminishes the engineering; it directs it. The most valuable data platform is not the most advanced one. It is the one the business actually uses to make better decisions, day after day. That is a product problem as much as a platform one — which is why data product thinking matters here more than most places.