The default move when a team needs a new internal tool is to subscribe to one. Six months later the team is paying hundreds of dollars a month for tools it barely uses, integrated poorly, and still cannot produce the one report leadership actually asks for.
The economics of this decision changed. The cost of building a small, focused internal tool has fallen dramatically, for a defined workflow with a small surface area, building is now frequently cheaper than a year of subscriptions.
The advantages compound. You own the data. You can change the workflow in an afternoon. You can integrate it with your own systems in ways a vendor will never expose. And you stop budgeting for tools you forgot you authorized.
When should you still buy? When the network effect is the product, when compliance is the real cost, or when the surface area is honestly larger than it looks, payments and error monitoring are the classic examples. Everything else deserves at least a build-versus-buy estimate.
The discipline that makes this work: keep a tools backlog beside the product backlog. When a subscription irritates the team twice in a month, it becomes a candidate. Price the replacement before the renewal date, not after.