Return On Intent: deciding what to build before building it

A north star is the intent, written down in plain words. It answers three questions:

• What is this really for?
• Whose life or work does it change?
• What would a true return on that intent look like?

A good north star names a change in someone's day, not a feature. "Build a puzzle app" is a request. "Open the chat and the board is ready: no install, no account, no noise" is a north star. It tells you what to build, and it tells you what to refuse.

We work in five movements. Each one is measured against the north star named in the first.

1. 01Name the north star. Before goals, features, or roadmaps, we define the intent together and what a Return On Intent would look like. We write the answers down, and everything after is measured against them.
2. 02Enter the real context. Understanding is earned in the room. We work inside your rhythm until your constraints and unwritten rules become ours too, because solutions built outside a context rarely survive contact with it.
3. 03Reveal the need beneath the request. The brief is a symptom. We keep asking until the real need surfaces, then name it together, so everyone is building the same thing for the same reason.
4. 04Shape it with the community. Early, rough, and in real hands. The people who will live with the tool are the ones who shape it. Their corrections are the design.
5. 05Refine with ihsan. Ihsan is excellence as a form of care: doing the work beautifully because how it is made matters. We iterate past working toward obvious, until the tool feels like it always existed and quietly returns on the intent it began with.

The full approach is on our Approach page.

Shafra is a picture-logic puzzle played entirely inside Telegram: quiet, tactile, and endlessly replayable. What it becomes next is steered by the people who play it.

That steering needed a tool of its own, so we built one. TG-Review lets an app's real users propose and vote on what gets built next. It began life steering Shafra's roadmap, and is being shaped into a tool any community can hold.

Together they show the method at small scale. The intent sets the filter. The community supplies the evidence. A suggestion earns its place by serving the intent, however popular it is.

Not with a dashboard first. With a sentence, and a habit.

• Write the intent as one sentence that names no feature.
• Name two or three signs you would see in people's days if it were working. Choose signs you can observe, not numbers that only look good in a report.
• Check each decision against it. Does this serve the intent, or only the request?
• Revisit it with the people it is for, early and often. Intent drifts quietly when nobody says it out loud.
• Stop work that does not serve it, even when it is half built. That is often where the return is.

1. 01What is this really for?
2. 02Whose day changes if it works?
3. 03What would they notice first?
4. 04What are we willing not to build?
5. 05Who should shape it before it is finished?

If you can answer these in plain words, you have a north star. If you can't yet, that is the right place to begin.