ShipOrStop
Decide whether this idea, direction, or live project deserves your next 12 weeks.
Start with what you already have: a rough idea, search terms, a niche, a feature request, competing directions, or launch evidence. ShipOrStop turns those signals into a scorecard, Build Plan, milestone tracking, and a founder-confirmed review.
Start with what you have
Six user states, one 12-week decision record.
Pick the row that matches the evidence you already have. Public tools help shape weak inputs; sign-in unlocks live provider data, competitor context, project persistence, and the Build Plan handoff.
Good fit
Solo founders, indie hackers, one-person companies, and early product teams comparing one idea, niche, feature bet, or launched project against the next 12-week build cycle.
Not a fit
Keyword discovery, article generation, team research ops, task management, or automatic stop decisions.
I only have a rough idea
Use the public worksheet when the idea is still a sentence. Add one description and a few demand phrases before spending provider credits.
Result: weak spots, the next evidence to collect, and whether the idea deserves saved Phase 0.
Research rough ideaI have an idea and search terms
Use saved Phase 0 when the idea has known keywords and needs live search data, competitor context, and a saved recommendation.
Result: AI-assisted classification, 13-question confidence upgrade, four-light recommendation, and Build Plan handoff.
Run saved Phase 0I only have keywords
Use the keyword worksheet when a query list exists but the user, job, or reachable wedge is still unclear.
Result: intent, visible competitors, and the smaller gap to test before Phase 0.
Analyze keywordsI have launch or GSC evidence
Use the sample review to see how launch evidence, milestone status, and original assumptions are compared before saved tracking.
Result: know which GSC, milestone, and assumption fields the final review will need.
Preview launch reviewI am choosing between ideas
Use Direction Finder when several ideas compete for the same build cycle and need comparable inputs.
Result: clearer user, channel, next test, and the one direction worth taking into Phase 0.
Compare directionsI am checking a niche, market, or feature bet
Start with a niche, audience, problem area, roadmap question, or feature request plus one representative query.
Result: decide whether to narrow the market, improve keywords, or run a saved evaluation.
Check market wedgeWhat the product actually gives you
From a first read to the final review, each phase has a visible output.
ShipOrStop is not a content generator or task board. It is a product demand validation record: score the bet, turn accepted evidence into a Build Plan, track the launch, then explain the next move with the limits still visible.
Text-only demand read
Input: a rough idea, feature request, niche, or direction with a few phrases people might search.
Output: weak spots, the next evidence to collect, and whether to move into saved Phase 0.
Demand scorecard
Input: known search terms, product context, and user answers for the full confidence upgrade.
Output: AI classification, market visibility, score limits, and a founder-confirmed four-light recommendation.
12-week Build Plan
Input: the accepted Phase 0 result, assumptions, weak dimensions, and stop conditions.
Output: reviewable Markdown with scope, milestone plan, cold-start tasks, assumptions, and stop rules.
Milestone and GSC tracking
Input: launch date, accepted Build Plan, current milestone, and GSC property after OAuth.
Output: milestone checklist status, exposure trend, weekly focus, and data limits against the original bet.
Founder-confirmed final review
Input: Phase 0 assumptions, Phase 2 history, GSC trend, and any available business evidence.
Output: AI-written explanation, visible limits, and a founder-confirmed ship, adjust, observe, or stop choice.
Real project example
EPUB Fixer is the owned-project example, not a made-up case study.
This section uses the public EPUB Fixer site plus ShipOrStop's current text-only score tool. It is internal dogfooding, not a customer result and not traffic, upload, download, or revenue data.
View public source pageLayer 1 - public facts checked
Public product
EPUB Fixer is a scanner for EPUB files rejected by KDP, Kindle Previewer, or EPUBCheck.
Visible entry paths
The public page lets visitors start from an EPUB file, error text, or a sample report.
Shown output
The sample report shows upload blockers, repairable items, manual review, and a second EPUBCheck pass.
Clear boundary
It handles clear EPUB package wiring and stops before conversion, formatting, layout, or manuscript changes.
Layer 2 - ShipOrStop text-only output
Product Demand Score
77/100
Top keyword to check next: epub validation checker.
Demand exists
Strong
88/100
Competition is reachable
Strong
88/100
Monetization path
Weak
49/100
Inputs and handoff
Product idea: scanner for EPUB files rejected by KDP, Kindle Previewer, or EPUBCheck.
Keywords: epub repair tool, kindle epub error fix, epub validation checker.
Source type: public page facts plus the current text-only score tool.
Phase 0 next check
Run live volume and reachable competitor checks before treating the 77/100 text-only score as a build signal.
Phase 1 Build Plan shape
Keep upload, error-guide discovery, sample report, repair gate, and unsupported cases inside the first 12-week plan.
Phase 2 / 3 evidence needed
Track GSC exposure, launch milestones, repair boundaries, and user-confirmed review criteria before deciding to ship, adjust, observe, or stop.
No connected GSC, upload, download, conversion, or revenue metrics are shown here because those are not public facts and were not provided for this pass.
Method and access
Know what is estimated, what is live, and what waits for your confirmation.
This is the missing layer before a user trusts the workflow: how the score is read, when sign-in matters, where launched projects enter, and what happens when GSC data is unavailable.
Scoring method
What the score reads
Public tools read idea specificity, search phrase intent, and whether a paid path is named.
Saved Phase 0 adds provider-backed keyword metrics, competitor context, and founder answers.
AI explains weak dimensions and next checks; the final recommendation stays editable.
Access boundary
What works before and after sign-in
Without sign-in: homepage, demo, sample report, and lightweight public worksheets.
After sign-in: provider calls, project persistence, full confidence upgrade, and Build Plan handoff.
The MVP does not run Stripe checkout; paid intent and manual confirmation stay separate.
Live project path
Where a launched project goes
Use the sample review first when you only want to see the fields ShipOrStop checks.
Create a saved project when the original assumptions, launch date, and milestones need to persist.
Connect GSC inside the saved workflow before Phase 2 and Phase 3 compare launch evidence.
Data trust
What happens to GSC data
GSC is an evidence channel, not the product identity or a full analytics warehouse.
OAuth data is handled server-side and is not exposed in frontend state, logs, or user-facing errors.
Missing metrics stay missing; AI does not fill empty GSC, upload, revenue, or conversion values.
AI explains the signal. You decide the move.
ShipOrStop will not decide for you.
AI explains the evidence, drafts next steps, and keeps the limits visible.
It uses the search terms you bring; it will not discover new keyword markets for you.
It will not write SEO pages or landing-page content.
It will not mark a project as dead on its own.
It will not fill missing GSC data with guesses.
Do I need Google Search Console before starting?
No. You can start with a rough product idea and a few search terms people might use. GSC is used later to compare launch evidence with the assumptions you accepted.
Is this just an SEO tool?
No. Search demand is the first evidence channel, but the output is a project decision path: scorecard, Build Plan, milestone tracking, and final review.
Can I use it for a feature request or an existing project?
Yes, if you can describe the user problem and bring search terms or launch evidence. Public tools shape the input; the saved workflow keeps the decision record.
Does AI decide whether I should stop?
No. AI explains the evidence, limits, and suggested path. You confirm whether to ship, adjust, observe, or stop.
Have one direction and a few known keywords?