The single biggest predictor of whether a platform investment pays off isn't the technology choice — it's whether anyone treats the platform's engineers as having users at all. A platform team that ships based on what it thinks is elegant, rather than what removes real friction for other teams, builds something nobody adopts.
That's why we start platform engagements the way we'd start a product: interviewing the engineers who'll actually use it, before writing a line of infrastructure code.
The justification to leadership should follow the same logic: don't pitch a platform on its technical merits, pitch it on the specific, measurable friction it removes — deployment time, provisioning delay, duplicated effort across teams. Then measure adoption the way a product team measures activation, and expand scope only once the thin first version has proven it's actually being used.
Platforms that compound value do it because they were built to be used, not just built to be correct.