Know which idea deserves your next 30 days.
Explore evidence-backed startup assessments, compare the economics, and see what would make each idea worth continuing, narrowing, pivoting, or stopping.
Free to explore · no sign-up · plain language, with every and explained
You reach $1k MRR in 3 mo. Month 24: $7.1k.
Ideas that plug into tools small businesses already pay for
- YouTube
- Confluence
- Atlassian Forge
- Shopify
- Veeam
- Synology
- QuickBooks
- Xero
- Razorpay
- Chrome
- Airtable
- Google Sheets
- Jira
- Square
- Acronis
- Klaviyo
- Jobber
- Housecall Pro
- Google Business Profile
- Microsoft Teams
- Amazon KDP
Start with your question
Wherever you are, start here.
No jargon, and every tool is free. Pick the question that sounds most like you.
- I have an idea. Is it worth my time?See how ideas like it were judged, what they need to work, and get yours reviewed by a person.Get my idea assessed →
- I want to build something, but what?Answer five quick questions about your time, goal and strengths. Get your three best-fit ideas, with odds and a plan.Find my idea in 5 questions →
- I've been building for months. Keep going?Use dated checkpoints to decide: continue, narrow, pivot or stop, before you lose another quarter.See how checkpoints work →
- Will the numbers ever add up?Enter your price and how many customers you can win. See how long it takes to reach your goal.Check my numbers →
Ideas are easy. Decisions are expensive. A good-sounding idea can still cost you months. Idea Atlas shows what it requires before you commit.
What you can do
One system, from first idea to a clear decision.
The idea catalog
Not every idea is equally worth building.
Tell us your hours and the monthly income you want. The ranking uses each assessment's own odds and economics, and shows its method.
Ranked by each assessment's judgment odds of reaching $1k MRR by month 24 at ~15 hrs/week. Months use each plan's own price, churn and pace. How this works
- 01
Compliance & audit evidence
StatusLens
Change-control evidence for audited Jira teams (SOC 2, ISO 27001, SOX): approvals, exceptions and an exportable evidence pack, built as a Forge app
As described:StopNarrowed:NarrowOdds of $1k
≈ 55 in 100
Plan pace
3 mo
- 02
Field operations & asset records
TradeLedger
System of record for small US fire-extinguisher inspection and service contractors: every unit at every site, due dates, routes, reports and city submission packs, and quotes for failed units
As described:StopNarrowed:NarrowOdds of $1k
≈ 55 in 100
Plan pace
3 mo
- 03
Accounting, billing & payments
RecoverPay
Failed-payment recovery for Indian subscription businesses on Razorpay and UPI Autopay: WhatsApp and SMS chasers with one-tap UPI links, and a monthly recovery report measured against a holdout group
As described:StopNarrowed:NarrowOdds of $1k
≈ 50 in 100
Plan pace
4 mo
- 04
Scheduling, bookings & reminders
Slotwise
Lesson booking and student-hour records for independent US teen driver-ed schools with 2–15 instructors, starting in Ohio
As described:StopNarrowed:ContinueOdds of $1k
≈ 50 in 100
Plan pace
3 mo
- 05
Compliance & audit evidence
AccessCheck
Honest monitoring, evidence files and accessibility-statement reporting for the European Accessibility Act, sold white-label to EU web agencies in one market (not an overlay, never claims compliance)
As described:StopNarrowed:NarrowOdds of $1k
≈ 45 in 100
Plan pace
4 mo
- 06
Consumer, creator & learning
AstroCompanion
Vedic-first AI astrologer that finds the user's real birth time from their past and remembers their life, for Indian and Indian-origin users
As described:StopNarrowed:ContinueOdds of $1k
≈ 45 in 100
Plan pace
2 mo
Why it's different
Not a score. A decision you can inspect.
A generic idea validator
- One AI-generated score
- Encouraging summary
- Large market claims
- Generic revenue potential
- Final verdict
- Assumes every founder is equal
- Hides uncertainty
Idea Atlas
- Inspectable reasoning
- Continue, narrow, pivot, and stop conditions
- Visible sources and unknowns
- Price, customer, churn, acquisition, and hour scenarios
- Dated checkpoints
- Founder-fit conditions
- Separates fact, evidence, estimate, judgment, and unknown
How the evidence is built
Human research. Visible sources.
Every assessment follows the same seven steps. Sources are read and checked by a person; nothing here is an automated scan.
- Company, pricing & industry sites1,029 cited
- Government & regulators78 cited
- App marketplaces41 cited
- Indie Hackers21 cited
- Capterra17 cited
Human-reviewed assessment
Verdict
Continue · Narrow · Stop
Economics
Customers needed, months to goal
Founder fit
What only you can supply
Decision gates
Dated checkpoints and kill criteria
Sources actually cited (Company, pricing & industry sites: 1029, Government & regulators: 78, App marketplaces: 41, Indie Hackers: 21, Capterra: 17) feed a human-reviewed assessment, which produces a verdict, economics, founder-fit conditions and decision gates.
Top source categories by real citation count. Sample research process · human-reviewed · nothing is an automated scan.
- 38
- structured assessments
- 495
- research pages
- 1,225
- cited sources
- 37/38
- “Stop” as first described
- 01Define the decisionWho must care, what they pay today, and what the assessment has to decide.
- 02Map existing alternativesIncumbents, cheaper tools, spreadsheets and free platform features to beat.
- 03Find customer languageHow buyers describe the problem in reviews, forums and marketplace listings.
- 04Verify market factsPrices, mandates and dates checked against primary sources.
- 05Model the economicsCustomers needed for each income level, with the math shown.
- 06Test founder conditionsWhat only you can supply, and how much the odds fall without it.
- 07Set decision gatesDated checkpoints with what to do when you miss one.
Sources actually cited
Generated from the links in every assessment. Only categories we really use appear.
- Company, pricing & industry sites1,029
- Government & regulators78
- App marketplaces41
- Indie Hackers21
- Capterra17
- GitHub15
- Social posts14
- App Store5
- Google Play2
- Hacker News2
- G21
Go deeper
Compare. Calculate. Challenge the assumptions.
New every week
Recently assessed
30 Sept 2026
AccessCheck
Honest monitoring, evidence files and accessibility-statement reporting for the European Accessibility Act, sold white-label to EU web agencies in one market (not an overlay, never claims compliance)
30 Sept 2026
CertTrack
Per-event checks of vendors' certificates of insurance (COIs) for US event venues: each vendor and the host's one-day policy checked against the venue's requirements for that date, chased on a schedule and logged, flat price per venue
30 Sept 2026
DocChase
Chases records from sole traders and landlords for small UK accounting practices before each Making Tax Digital for Income Tax quarterly deadline, using no-login WhatsApp, SMS and email links, priced per MTD client
30 Sept 2026
DocPress
Developer API for German e-invoices: JSON in, then a validated ZUGFeRD/Factur-X PDF and XRechnung out, plus parsing of received e-invoices, ahead of the 2027–28 B2B mandate
Questions
Plain answers, no jargon.
Do I need to be technical to use this?
No. Every assessment is written in plain language, and the calculators only ask for things like your price and how many customers you can win. Hover any underlined term for a plain definition.
Are the odds predictions?
No. They are reasoned estimates from the evidence, like a weather forecast for an idea: “about 15 in 100” means about 15 of 100 similar attempts get there. Use them to compare and decide, not as a promise.
What is MRR?
Monthly recurring revenue: the money that comes in every month from subscriptions. $1k MRR means $1,000 a month.
Is it free?
Yes. All 38 assessments, the calculator, the comparisons and the research are free to read. During early access, applying for a human-reviewed brief of your own idea is free too. Paid extras may come later, but a free version will always exist.
Who writes the assessments?
People, using a fixed template and real sources. Every claim links to where it came from, and unknowns are listed instead of hidden.
What if my idea gets a “Stop”?
That is useful too: 37 of 38 ideas here are “Stop” as first described, and most then narrow to a smaller version that could work.
Stop guessing. Start deciding.
Explore free, or get a human-reviewed decision brief for your own idea: research, economics, risks and a 30-day plan with clear continue, narrow, pivot and stop gates.
One new assessment every week
Including the ideas we tell you to stop. One email, no hype.