There are roughly eleven thousand MCP servers as of May 2026. Less than five percent are monetized.

That number sounds like failure. It isn’t. It’s a map.

The tools that aren’t making money aren’t all bad tools. Many of them are genuinely useful — they get installed, get used, get appreciated. They just weren’t built with a revenue model in mind. They’re utilities that someone needed, built for their own use, shared because sharing costs nothing. The ecosystem is full of them.

The five percent that are making money have something the other ninety-five percent don’t: a specific person with a specific problem who pays to have it solved. Not a general user. Not a developer playing with capabilities. Someone with a workflow, a bottleneck, and a budget for removing it.

The shape of the opportunity isn’t “build a better tool.” It’s “find a workflow where the bottleneck is painful enough that someone budgets to solve it, and build the tool that solves it.”

Professional workflows are the right hunting ground for this. Not because professionals have more money (though they often do), but because they have the clearest bottlenecks. A boutique acquisition analyst knows exactly which tasks take too long. A paralegal knows exactly which steps can’t be skipped. A financial analyst knows exactly where the process breaks down under deadline pressure. Professionals live inside their bottlenecks. They know what they’re worth.

Consumer tools have to discover their own value proposition through iteration. Professional tools can ask. The bottleneck is already visible before you build anything.

The five percent who are making money figured this out, consciously or not. They found someone who knew exactly what they needed and built the tool that fit.

The gap between five percent and the rest isn’t capability. It’s specificity. +++