guesstimate-prof

Release 1: MoSCoW prioritisation

Date: 27 September 2026 Method: Story map slicing to check each story against the end-to-end journey, then MoSCoW to label the result. Test applied to each story: could the journey (nudge, choose, solve, feedback, progress) still work end to end without it?

Release 1 scope

1. You pick a difficulty and get one question. You can skip it.
2. You answer, by voice if the voice spike passes, otherwise by typing.
3. If you’re stuck, you tap for a hint (or turn hints off).
4. You submit and see the model structure and whether your number was in the right ballpark.
5. The app remembers your results.

Prioritisation

Story	MoSCoW	Reason
Choose difficulty	Must	Builds skill from easy upward; low effort once difficulty rules exist
Skip to another question	Must	Stops a disliked question becoming a reason to close the app
Guided prompts by archetype	Must	Makes the app a trainer rather than a quiz
Choose guidance level	Must	Hints on tap, or none; practising without help mirrors the interview
Compare with model structure	Must	Structure, not the final number, is the real measure of skill
Ballpark check	Must	Catches a broken structure; Release 1 uses verifiable questions only
Save my history	Must	Without it, the app forgets every attempt; scope line 5 depends on it
[Spike] Define guesstimate archetype	Must	Unblocks difficulty rules, prompts and progress tracking
[Spike] Voice capture feasibility	Must	Resolves the largest technical risk cheaply before committing
Notification at a scheduled time	Should	App works without it; wanted for daily accountability
Voice exchange (scripted)	Should	Conditional on the voice spike; typing is the fallback
Track my progress by archetype	Should	Useful only once enough answers exist; blocked by the archetype spike
Reference figures	Could	Not needed for the MVP; recalling figures is part of the skill
Release 2 stories (AI)	Won’t (R1)	Depend on the AI, planned for Release 2

Design decisions made during prioritisation

• One question at a time, with its difficulty shown; skip offers another at the same difficulty.
• Difficulty is chosen by the user and can be changed at any time.
• Guidance is on or off. On: start alone and tap for hints. Off: no hints.
• Hints are the archetype prompts, revealed one per tap, in order; hint use is recorded.
• Reference figures follow guidance: shown when on, hidden when off.
• Feedback leads with structure; the number is judged only as a ballpark check.
• No automatic step-up suggestions; the user controls guidance.
• Release 1 uses questions with verifiable answers only. Each question stores its benchmark as an optional field so unverifiable questions can be added later without a rebuild.
• Progress tracking is split: a plain history (Must) and an archetype breakdown (Should).

Open

• How skill is measured (decided in the archetype spike)
• The archetype list (archetype spike)
• Whether voice is feasible (voice spike)

Parked for after Release 1

• Skip tracking (planned for the second half of Release 1 if time allows)
• Questions without published answers