Internal product. Does it need a product strategy?

Internal products have users, a backlog and a team, and almost never a strategy. The reasoning goes: nobody buys it, so there is nothing to position. That was survivable when building an internal tool took a quarter. It is not survivable when anyone with a coding agent can build one in a week, because that is how a company ends up with six half-tools that each do part of the same job.

We did this to ourselves

This month we built an internal tool for our own marketing work: readings, tasks, a lead pipeline, a queue of drafts waiting for approval. It started as one screen. Within days it had grown a tab every time someone had a request, and the only thing that stopped it becoming a mess was a rule written on day one: it reads the numbers, tracks the tasks and holds the pipeline, and nothing else. That sentence was its strategy. Every feature request since has been checked against it, and most were refused.

What an internal strategy has to say

• Who the user is, and who the customer is. On an internal product they are different people. The user is the team that works in it; the customer is whoever pays for the team's time. Both have to be named.
• The one job. The thing the tool does that would otherwise be done by hand, and the cost of the hand version. That number is the tool's reason to exist.
• What it will not do. The neighbouring jobs that already have an owner. Two lists of the same thing drift the moment one changes, so an internal tool that duplicates another system is worse than none.
• When it is done. Internal products never launch, so they never finish. Write down what finished looks like.

Why coding agents make this urgent

Because the cost of building an internal tool has dropped below the cost of deciding whether to. Teams now build first and find out later that someone else built the same thing. A one-page strategy, checked before the first prompt, is the cheapest form of not doing that. And once the page exists, the same loop as any product applies: what the tool can do today, the gap to what it should do, the next change written as a story with acceptance criteria before the agent starts.

A test

Ask the owner of any internal tool what it is for, in one sentence, and what it will not do. If the answer takes a paragraph, the tool has no strategy and is about to grow.