I have 22 plugins on the JetBrains Marketplace. Here is what the business side taught me that I wish someone had written down.
We tried free-only early on. Downloads were fine, feedback was zero. The moment we added a trial-to-paid flow, we started getting bug reports, feature requests, and actual emails from users. Paying users care about the thing working. Free users quietly leave.
Now every plugin starts with a 30-day trial, then monthly or annual. JetBrains handles billing, licensing, and trial management through the Marketplace. You build the plugin, they handle the commerce. For a small team, that split is exactly right. The price depends on the audience. Our ABAP plugin targets SAP developers at large companies, an enterprise audience, so the price reflects it. Our Arduino plugin targets hobbyists, so it costs less.
Discovery
People find plugins mostly through marketplace search, and it is keyword-based. If someone types “Julia IDE support” they might find our Julia plugin. If they search “Julia language” they probably will not. Optimizing your listing is its own skill.
Outside the marketplace it is harder. Blog posts help. Answering Stack Overflow questions helps. But your market size is capped by the language community. Our OpenSCAD plugin serves a small niche. The users love it, but the economics are what they are.
Review queue and support
You submit an update, it enters a queue, and you wait. Sometimes two days, sometimes two weeks. No status page, no estimated timeline. If you have a critical bug fix, you submit it and hope. I have learned to batch updates and avoid submitting right before holidays. That is about as much control as you get.
Support is GitHub Issues, mostly. Some users email. A few write marketplace reviews when something breaks, which is suboptimal but understandable. I try to respond within a day, and for 22 plugins that means a lot of context-switching. The support load scales with the user base, not the plugin count.
Should you build one?
If you have a problem that an IDE plugin would solve, yes. The IntelliJ Platform is better than the documentation suggests, and starting with a problem you actually have makes you your own first user, which is the best quality filter there is.
But do not plan a plugin business. Build one plugin and see if people use it. If the infrastructure turns out to be reusable, and it will because the platform is designed that way, build the next one. I did not plan 22. They accumulated.