Science of Apps - Launchpad · Framework

How to evaluate a startup:
a 30-point framework

To evaluate a startup, score six things in order: the problem (is it real and urgent?), the product (does it solve it, and can new users get value fast?), the market (is it large and reachable?), traction (are users staying and paying?), the team (can they execute?) and risk (what could kill it?). Early-stage investors weigh team and problem most; later, traction dominates. Score each 0–5 with evidence, not opinion.

By , FounderUpdated 8 min read

Share

Quick answer: How do you evaluate a startup?

  1. Six areas: problem, product, market, traction, team, risk.
  2. 30 points: five questions per area, each scored 0–5 with evidence.
  3. Early stage: team and problem weigh most; traction is thin.
  4. Later stage: retention, revenue and unit economics dominate.
  5. Product evidence: test the live product, not the pitch.

Key takeaways

  • Evaluate with evidence: every score should point to something you can show.
  • The product is the most-claimed and least-tested part of most evaluations.
  • Retention tells you more about product-market fit than sign-ups.
  • Weights change with stage; a seed startup isn't judged like a Series B.
  • Founders should run the same evaluation on themselves before investors do.

Every investor, accelerator and acquirer evaluates startups, and most founders only find out how after the "no". The questions are surprisingly consistent: is the problem real, does the product solve it, is the market big enough, are users staying, can this team execute, and what could kill it. This guide turns those questions into a 30-point startup evaluation framework you can run on any startup, including your own. It's our editorial framework, built on the outlines investors publish (Sequoia, Y Combinator and Dave Berkus), which we read on 17–18 November 2026. Jump to the framework, how weights change by stage or how to evaluate the product.

Areas every evaluation covers
6
Questions, scored 0–5 with evidence
30
Product areas in a full product test
22
Rule: evidence, not opinion
1

The 30-point startup evaluation framework

Score each question from 0 (no evidence) to 5 (strong, verifiable evidence). The point is not the total; it's finding the weakest area.

AreaThe five questionsEvidence to ask for
1. ProblemIs it real? Urgent? Frequent? Who has it most? Why now?Interviews, search demand, workarounds people already pay for
2. ProductDoes it solve the problem? Can a new user get value fast? Is it usable and accessible? Is it trustworthy? Is it hard to copy?A live product test, onboarding data, usability findings
3. MarketHow big? How reachable? Growing? Who else serves it? What's the wedge?Bottom-up sizing, channel tests, competitor map
4. TractionAre users active? Retained? Paying? Referring others? Growing efficiently?Cohort retention, revenue, referral rate, CAC payback
5. TeamDo they know the problem? Can they build? Can they sell? Have they shipped before? Do they learn fast?Track record, shipping speed, customer conversations
6. RiskWhat kills it: tech, market, regulation, platform, funding? Which is mitigated?A written risk list with mitigations

Table: A 30-point startup evaluation framework (Science of Apps editorial framework, built on Sequoia's and YC's published outlines)

What investors publish about evaluating startups

The best-known published outlines line up closely with the framework above.

Sequoia Capital article Writing a Business Plan by Team Sequoia
Sequoia's outline: purpose, problem, solution, why now, market, competition, model, team, financials, vision.Image: Sequoia Capital · Screenshot shown for commentary, 17 Nov 2026 · source
  • Sequoia's outline asks for company purpose, problem, solution, why now, market potential, competition, business model, team, financials and vision.
  • Y Combinator's seed deck advice (Aaron Harris) stresses clarity and concision: at seed stage there isn't much detail, so the narrative and the founders carry the evaluation.
  • The Berkus Method, from angel investor Dave Berkus, values a pre-revenue startup by adding up to $500,000 for each of five risk-reducing elements: a sound idea, a prototype, a quality management team, strategic relationships and product rollout or sales.
Dave Berkus's article The Berkus Method: Valuing An Early Stage Investment, dated March 25, 2012
The Berkus Method: value assigned to evidence that reduces risk.Image: Dave Berkus · Screenshot shown for commentary, 18 Nov 2026 · source

Notice the pattern: every method rewards evidence that reduces risk. A working prototype is worth more than a plan; a retained user is worth more than a sign-up.

How the weights change by stage

StageWeighs mostWeighs leastTypical evidence
Idea / pre-seedTeam, problemTractionInterviews, prototype, founder track record
SeedProblem, product, early tractionFinancialsActive users, early retention, first revenue
Series ATraction, market, unit economicsIdea noveltyCohort retention, revenue growth, CAC payback
GrowthEfficiency, market share, riskPrototypeMargins, net revenue retention, competitive position

Table: How startup evaluation weights shift by stage (editorial summary)

How to evaluate the product (area 2)

The product is where evaluations are weakest. Decks show screenshots; nobody signs up. Here's how to do it properly:

Diagram of 7 steps for how to evaluate the product (area 2): 1. Arrive as a stranger; 2. Sign up in a clean browser; 3. Reach the first moment of value; 4. Do the core task; 5. Hit the paywall; 6. Check accessibility and speed; 7. Write down every finding with a screenshot.
How to evaluate the product (area 2)
  1. Arrive as a stranger — find the product from a search, read the landing page and note what it does in the first five seconds.
  2. Sign up in a clean browser — on mobile too; count the fields and the friction.
  3. Reach the first moment of value — how many minutes and steps?
  4. Do the core task — the job the product is hired for, end to end.
  5. Hit the paywall — is the upgrade clear and honest?
  6. Check accessibility and speed — WCAG basics and page load on a mid-range phone.
  7. Write down every finding with a screenshot — evidence, not impressions.

That's a compressed version of what the Science of Apps Launchpad does when you list a product.

Science of Apps product investigation page with the headline Product intelligence that shows its work
A product investigation: every finding backed by a measurement or pinned to its screen.Image: Science of Apps · © Science of Apps · source
  1. Arrive

    First impression

    Discoverability on Google, AI assistants and social; landing page, pricing, trust and speed.

  2. Join

    The first minutes

    Sign-up, sign-in, recovery and onboarding to the first moment of value.

  3. Use

    The everyday job

    Home, core task flow, dashboards, search, microcopy and empty, loading and error states.

  4. Pay

    Value into revenue

    Paywall, upgrade, checkout and payments.

  5. Stay

    Coming back

    Notifications, account, settings and help.

  6. Every screen

    Throughout

    Accessibility to WCAG 2.2, and visual design and consistency.

Evidence for area 2

Your product, evaluated

List free. Once you're live, get the free baseline check, then the report that fits your stage.

Economy

See what new customers see

  • Store listing and homepage, screen by screen
  • Scorecard against 60 Laws of UX
  • Customers, competitors and the fixes worth most
≈ 40 pagesDelivered within 3 hours
Most popular
Premium

The full diagnosis and the plan

  • Every screen, every finding, every flow
  • All 60 laws with evidence and fixes
  • 8 strategy chapters and a 90-day roadmap
≈ 95 pagesDelivered within 24 hours

Researched from your live product and fact-checked against the evidence.

A worked example: scoring a seed-stage SaaS

An illustration, not a real company: a seed-stage scheduling SaaS for clinics.

AreaScore (0–25)Why
Problem20Interviews show clinics lose hours to no-shows; they pay for workarounds
Product11Works, but sign-up takes 9 fields and the first booking needs a 20-minute setup
Market15Reachable through clinic associations; sizing done top-down only
Traction940 clinics active; month-3 retention unknown
Team19Founder ran a clinic; technical co-founder shipped before
Risk12Depends on one calendar integration; no mitigation written
Bar chart: A worked example: scoring a seed-stage SaaS — Problem: 20; Product: 11; Market: 15; Traction: 9; Team: 19; Risk: 12.
A worked example: scoring a seed-stage SaaS

Table: Illustrative startup evaluation of a hypothetical seed-stage SaaS

The verdict writes itself: fix onboarding (product) and measure retention (traction) before raising.

Mistakes when evaluating a startup

Checklist: Mistakes when evaluating a startup — Scoring the pitch instead of the product; Counting sign-ups as traction; Top-down market sizing only; Ignoring risk; Using the same weights at every stage; Founders skipping self-evaluation.
Mistakes when evaluating a startup
  • Scoring the pitch instead of the product — sign up and use it
  • Counting sign-ups as traction — retention and revenue are the evidence
  • Top-down market sizing only — "1% of a huge market" isn't a plan
  • Ignoring risk — write down what could kill it and what's being done
  • Using the same weights at every stage — a pre-seed team isn't judged on revenue
  • Founders skipping self-evaluation — investors will find the weak area anyway

Evaluators also fall for the aesthetic-usability effect: a polished deck or product makes everything else look better than it is. Score each area separately, with evidence.

Ready to get evaluated?

How we built this framework

The six areas and their questions are Science of Apps' editorial framework, cross-checked against Sequoia's business plan outline, Y Combinator's seed deck guidance and the Berkus Method as published by Dave Berkus and the Angel Capital Association, all read on 17–18 November 2026. The worked example is illustrative, not a real company. Search volumes are Google Keyword Planner's global 12-month averages.

Key terms, defined

  • Startup evaluation: a structured judgement of a startup's problem, product, market, traction, team and risk.
  • Traction: evidence that users want the product: activity, retention, revenue and referrals.
  • Retention: the share of users who keep using the product over time.
  • CAC payback: months of gross profit needed to recover the cost of acquiring a customer.
  • Berkus Method: a pre-revenue valuation method that assigns up to $500,000 to each of five risk-reducing elements.
  • Product test: using the live product as a new user would and recording findings with evidence.

Frequently asked questions

How do you evaluate a startup?

Score six areas with evidence: the problem, the product, the market, traction, the team and risk. Ask five questions in each, score 0–5, and weight them by stage: team and problem matter most at pre-seed; retention, revenue and unit economics dominate from Series A.

What criteria do investors use to evaluate startups?

Most use a version of the same outline: the problem and why now, the solution, market size, competition, business model, traction, team and financials. Sequoia's published pitch outline covers almost exactly these. Investors add their own judgement of the founders and of risk.

How do you evaluate a startup for investment?

Start with the team and the problem, check the product yourself, verify traction from data rather than the deck, size the market bottom-up, list the risks, then decide what valuation the evidence supports. Pre-revenue methods such as Berkus and Scorecard turn qualitative evidence into a number.

How do you evaluate a startup's product?

Use it as a new user would: find it from search, read the landing page, sign up, reach the first moment of value, try the core task, hit the paywall. Note every point of friction with a screenshot. The Science of Apps Launchpad runs this kind of test across 22 areas once a product is listed.

What is a startup evaluation framework?

A structured list of questions and weights used to judge a startup consistently. Good frameworks cover problem, product, market, traction, team and risk, ask for evidence on each, and adjust weights by stage, so two evaluators would reach similar scores.

How do I evaluate my own startup?

Run the 30 questions honestly, ask someone who doesn't love the idea to score it too, and get outside evidence on the product: user interviews, analytics and an independent product test. The weakest area is your next priority.

Sources

  1. Sequoia Capital — Writing a business plan (checked 17 Nov 2026)
  2. Y Combinator — How to build your seed round pitch deck (checked 17 Nov 2026)
  3. Dave Berkus — The Berkus Method: valuing an early stage investment (checked 18 Nov 2026)
  4. Angel Capital Association — After 20 years: updating the Berkus Method of valuation (2024)
  5. Science of Apps — Product investigation: 22 areas (checked 18 Nov 2026)
  6. Science of Apps — Launchpad (checked 18 Nov 2026)
  7. Checked November 18, 2026. Figures that change — fees, salaries, prices — are reviewed every six months; if something is out of date, tell us.
Pass it on

Know someone who'd enjoy this?

Send it to them — one tap, no sign-up. Good ideas travel best between friends.

Keep exploring

Where to go from here

Join free

Get the ideas, not the noise.

Sign up with Google and we’ll send you short UX psychology tips by email — why people tap, stall and leave, and what to do about it. Unsubscribe anytime. Prefer something to keep? Get the free Laws of UX guide as a PDF.

Get the free Laws of UX PDFFree · no spam · unsubscribe anytime
Free, forever

Launch day ends. Your listing shouldn't.

The product half of any startup evaluation is the hardest to judge from a deck. List free on the Science of Apps Launchpad and get your live product tested across 22 areas, with evidence for every finding.

Free PDF · 74 pages · 2 editions

The Laws of UX, in one field guide

60 research-backed principles in 6 categories — what each one means, why it works on the human brain and how to use it on your own screens. Written by Sushant Singh, founder of Science of Apps, who has scaled consumer apps to lacs of users and spent 10+ years leading product across startups and digital businesses in the US, EU & India.

Pick your edition
Free PDF · 90 pages · 2 editions

100+ Subscription Paywalls, decoded

110 real paywalls from top apps, taken apart, rated out of 5 and decoded with the Laws of UX — plus the frameworks, copy and checklists to ship your own. By Sushant Singh, founder of Science of Apps.

  • 110 paywalls rated
  • 34 Laws of UX
  • 12 frameworks

Which paywall scored 4.8 out of 5, and which scored 2.6? It’s all inside.

Pick your edition
Free PDF · 188 pages · 2 editions

Unmistakable: the logo book

How the world’s greatest logos are made, chosen and kept — and how to make yours. 98 short lessons and 19 true stories for founders, designers and the curious. By Sushant Singh, founder of Science of Apps.

  • 98 lessons
  • 19 true stories
  • 60+ real logos

Why did a $35 sketch become the Swoosh, and why did Gap lose its nerve in six days? It’s all inside.

Pick your edition