Nobody gets excited about building a platform. They get excited about the thing the platform makes possible.
That's the core tension in platform work. You're asking a team to invest in infrastructure that won't show its value until someone builds on top of it. In an environment where roadmaps are quarterly and success is measured by features shipped, that's a hard argument to make.
The way I've found to make it: stop calling it a platform. Talk instead about the specific thing that becomes possible once it exists, and how long it would take without it.
"We're building a configurable data layer" lands differently than "right now, surfacing a new insight to field agents takes two months. With this, it takes two days." Same work. Completely different conversation.
Platform thinking is the discipline of asking what you'll want to do six months from now that you can't do today, then building so that it's a two-day job instead of a two-month one. Teams that do this build faster over time. Teams that skip it end up rebuilding the same foundations under different names. The payoff isn't immediate, but it compounds.