Choose the next feature by naming the user problem first, setting a time budget, and defining the smallest useful result. If you cannot explain who is stuck, what they cannot do, and how you will check the fix, keep researching before committing to a build.
Preparing a product launch? Launch your startup on Maidensail when you are ready to share what it does.
Turn the request into a problem
"Add a dashboard" is a proposed solution. "The owner cannot tell which invoices need chasing" is a problem you can investigate.
Ask the person who requested the feature to walk through the last time they needed it. What were they doing? Where did they stop? What workaround did they use? Record the actual situation before deciding what to build.
Basecamp gives a useful example in its Shape Up chapter on setting boundaries: a request for complex permissions turned out to concern someone archiving a file without knowing it would disappear for everyone. The team identified a warning on the archive action as a smaller response to that problem. The lesson to borrow is the question, not the exact solution: what went wrong behind the request?
Decide how much time the problem deserves
In Shape Up, Basecamp calls this an "appetite": a time budget that constrains the solution, rather than an estimate of a design you have already chosen. Its principle is "fixed time, variable scope."
For a small team, try choosing a bounded amount of time you can afford before drawing the full feature. You do not need to copy Basecamp's team sizes or cycle lengths. The useful distinction is between "this might take a week" and "we are willing to spend a week solving this problem."
Then ask what a useful version could leave out. Cut optional scope, not data protection, accessibility, or checks needed to make the result safe to use. If no useful version fits, reconsider the budget or postpone the feature.
Compare a short list, not every idea
Pick a few candidate problems and write the same notes for each:
- User: who needs this, and are they part of the audience you serve?
- Evidence: a recent example, support conversation, or observed failure. Label guesses as guesses.
- Consequence: what happens if you leave the problem unsolved for now?
- Smallest useful result: what would the user be able to finish after the change?
- Time budget and exclusions: what are you willing to spend, and what will this version deliberately skip?
Use those notes to make a judgment. A numerical score is optional; do not let a precise-looking total hide uncertain evidence. A request from one person may still deserve attention if it exposes a serious failure. A popular request may need more research if you do not understand the underlying job.
Write a finish line before starting
For a hypothetical invoicing tool, a bounded feature note might read:
Help an owner find overdue invoices and open the right customer record. Include an overdue filter and a link to the existing record. Exclude automated reminders and a new reporting dashboard. Check whether the owner can complete the task without exporting a spreadsheet.
That is a proposed test, not a promised improvement. After release, watch someone use it and compare the result with the original problem. The guide to your first metrics offers related prompts for looking at useful actions rather than attention alone.
The takeaway
Choose a problem you can explain, fund it with a deliberate time budget, and ship a bounded answer you can check. Keep the other ideas available without treating every request as a promise.
Ready to share the product you have built? Launch your startup on Maidensail.
Source note: the appetite and fixed-time, variable-scope concepts above come from Basecamp's Shape Up. The comparison checklist and invoicing example are suggested exercises, not research findings or claims about Maidensail users.

