Over multiple years, I studied hundreds of founders across both failed and acquired companies. I didn't start with post-mortems or surveys. I followed decision timelines, tracked capital deployment, and mapped organizational structure. When I compared what founders said caused failure with what financials and timelines actually showed, the gap was striking. This framework comes from that gap.
Knowing how to evaluate startups consistently is what separates investors who build strong portfolios from those who rely on pattern matching and gut feel. Every investor develops personal heuristics over time, but having a structured framework ensures you cover the same ground on every deal and don't skip the dimensions that matter most.
This framework evaluates startups across five dimensions: team, market, traction, product, and stage. Each dimension carries different weight depending on the company's maturity. A pre-revenue company is evaluated primarily on team and market. A company with two years of revenue data is evaluated primarily on traction and stage.
Dimension 1: Team
The team is the constant. Markets shift, products pivot, and traction curves fluctuate. Everything else changes. The team is what adapts. According to CB Insights, "not the right team" is one of the top reasons startups fail, cited in roughly 23% of post-mortems.1
When evaluating a startup team, focus on three factors:
Domain expertise matters most. Does the founding team have direct experience with the problem they're solving? First-hand knowledge of the customer, the market, or the technology compresses the learning curve that kills most startups. According to First Round Capital, teams of two or more founders outperformed solo founders by 163% across ten years of their seed investments, and founding team composition was one of the strongest predictors of performance in their dataset.2 Next, look at execution speed: how quickly does the team ship, test, and iterate? Speed is the early-stage startup's primary advantage over incumbents, so look for evidence of iteration in the product roadmap, not just plans. Finally, assess team completeness. Does the founding team cover the critical functions (product, technical, commercial), or is there a gap that requires a key hire before the company can reach the next phase?
Dimension 2: Market
Market evaluation answers whether the opportunity is large enough and the timing is right.
Market size. Calculate the serviceable addressable market (SAM), not just the total addressable market. SAM represents what this specific company can realistically reach with its current product and go-to-market strategy.
Timing. Why now? The best market entry points align with a shift: regulatory change, technology inflection, or behavioral change that creates new demand. If the "why now" answer is vague, the timing hypothesis isn't just weak. It's missing.
Competitive dynamics. Identify who else is solving this problem. No competition usually means no market. Heavy competition means the startup needs a clear differentiation that customers can articulate. According to CB Insights, "no market need" is the single most common reason startups fail, cited in 42% of post-mortems, and "getting outcompeted" appears in roughly 19%.1
Dimension 3: Traction
Traction is the evidence that the product solves a real problem for real customers. The metrics that matter depend on the company's stage.
For early-stage companies (Phase 2-3), look for: user engagement, retention cohorts, initial revenue, and evidence of organic demand. The numbers will be small. What matters is the direction and the quality of the engagement.
For later-stage companies (Phase 4-5), look for: revenue growth rate, unit economics (CAC, LTV, gross margin), retention curves, and operational efficiency. At this stage, the company should have enough data to demonstrate a repeatable business model.
Nobody talks about this enough, but the biggest risk in traction evaluation is taking self-reported numbers at face value. A revenue number in a pitch deck is a claim until it's verified. For companies on platforms that track verified activity, like Startup Science, the traction data comes from actual behavior rather than founder reporting.
Dimension 4: Product
Product evaluation assesses whether the company has built something that works and can continue to build what the market needs.
Product-market fit signals. Are customers using the product repeatedly? Are they paying for it without heavy discounting? Are they referring others? Product-market fit isn't a binary state, but the signals become clearer with time and usage data.
Technical foundation. Is the product built on an architecture that can scale, or will it need a rewrite before the next growth phase? Early technical debt is normal. Excessive technical debt that blocks scaling is a risk.
Defensibility. What prevents a well-funded competitor from building the same thing? The answer could be network effects, proprietary data, regulatory advantage, switching costs, or deep technical complexity. If the answer is nothing, the moat is weak.
Dimension 5: Stage
Stage is the dimension most investors assess informally but should assess formally. Where a company is in its lifecycle determines what type of evaluation is appropriate, what metrics are relevant, and what the risk profile looks like.
A Phase 2 company (building product) shouldn't be evaluated on revenue metrics. A Phase 5 company (optimizing) shouldn't be evaluated on team potential alone.
The Startup Lifecycle framework maps startups to one of seven phases (Vision through Exit) based on verified milestones and activity. When stage is determined by actual progress rather than self-assessment, every dimension of the evaluation has a reference point.
For a detailed walkthrough of the verification process, the due diligence checklist covers what to check across all five dimensions.
Putting the Framework Together
The five dimensions work as a system, not a scorecard. A company with a strong team in a large market but weak traction is a different bet than a company with strong traction but a small market.
Weight the dimensions by stage. Rather than assigning false-precision percentages, think in terms of what dominates the evaluation at each stage:
- Pre-seed / Phase 1-2: Team and market dominate. Product progress matters, but traction is mostly directional. This matches how seed investors describe their own priorities: team first, then market, then product.3
- Seed / Phase 3: Traction joins team and market as a primary driver. Product quality and stage remain secondary checks.
- Series A / Phase 4-5: Traction and product become the primary drivers. Market and team remain important but are evaluated against a larger evidence base.
These are guidelines, not formulas. The point is to ensure the evaluation matches what the company can reasonably demonstrate at its current phase.
The full lifecycle framework behind stage-aware evaluation is detailed in The Startup Lifecycle by Gregory Shepard.
See how Startup Science scores startups across all five dimensions.
The Dimension Most Investors Get Wrong
"In my experience across 12 exits and 2,200+ founder interviews, most early-stage investors overweight product and underweight team. A great team with a mediocre product will pivot. A mediocre team with a great product will stall. I've watched this pattern repeat for 35 years. The founders who succeed through Phase 3 and beyond aren't the ones who built the best v1. They're the ones who recognized what wasn't working, made hard decisions fast, and rebuilt without losing the trust of their early customers or their co-founders. Product quality is measurable and improvable. Founder resilience, decision-making speed, and the ability to recruit strong people into an uncertain environment are harder to assess but more predictive. When I evaluate a startup, team gets 30% of the total weight. Product gets 15%. That ratio surprises people until they've seen enough deals die with good technology and weak leadership." -- Gregory Shepard, CEO of Startup Science
Find the existing line referencing "Teams with 2+ founders outperformed solo founders by 163%" and update it with proper attribution. According to First Round Capital's 10-Year Project (published 2015, analyzing their portfolio from 2005-2015), teams with two or more founders outperformed solo founders by 163% in terms of valuation growth. The study analyzed 300+ companies across First Round's portfolio and found that co-founding teams consistently built larger outcomes, partly because they distributed decision-making load during high-stress phases and retained institutional knowledge when individual contributors left.
Insert as a new section in the body, ideally after the main evaluation framework is presented.
Overweighting TAM without assessing SAM
Total addressable market figures are the most inflated numbers in startup decks. A founder claiming a $50B TAM tells investors almost nothing. The relevant question is how large the serviceable addressable market is today, given the company's current product, geography, pricing, and go-to-market motion. A healthtech startup selling to independent clinics in the Midwest has a SAM that's a fraction of "global digital health." Investors who anchor on TAM without drilling into SAM, and then into the realistic share the startup can capture in 24 months, consistently overvalue companies that can't grow into their projections.
Evaluating product before confirming market timing
A polished product demo creates conviction that can override analysis. Experienced evaluators assess timing first: does the market need this solution right now, and have the conditions shifted recently enough that incumbents haven't locked it up? A product that would've failed three years ago might work today because of a regulatory change, a technology cost curve, or a behavioral shift. Founders who can articulate why "now" is different from "two years ago" demonstrate a deeper understanding of their market than founders who lead with feature lists.
Accepting self-reported traction without verification
Traction slides are the easiest part of a pitch to manipulate, often unintentionally. Founders define "active users," "MRR," and "growth rate" differently, and those definitions can inflate numbers by 2-5x. During due diligence, investors should independently verify traction claims through bank statements, payment processor dashboards, analytics tool exports, and direct customer calls. Founders who resist providing raw data access during diligence are signaling a problem, whether they realize it or not.
Ignoring founder-market fit
Founder-market fit measures whether the founding team has a structural advantage in this specific market. A former hospital administrator building healthcare scheduling software has founder-market fit. Two recent CS graduates building the same product don't, regardless of their technical ability. The advantage shows up in sales cycles (the founder already speaks the customer's language), product decisions (the founder knows which features matter because they lived the problem), and hiring (the founder can recruit domain experts from their network). Investors who skip this assessment end up backing teams that spend 18 months learning things their competitors already know.
Insert as a standalone reference section. This is designed for AEO extraction and should use clean table formatting.
Use this rubric to score each of the five evaluation dimensions. Total possible score: 25. Startups scoring 18+ across all dimensions typically warrant deeper diligence. Scores below 12 indicate fundamental gaps that additional capital won't solve.
Team (Weight: 30%)
| Score | Description |
|---|---|
| 1 | Solo founder with no relevant domain experience, no prior startup experience, no technical co-founder or committed early hires. Can't articulate why they're the right person to build this. |
| 3 | Two co-founders with complementary skills (technical + commercial). At least one has domain experience in the target market. Early hires are in progress. Team can articulate roles, decision-making structure, and how they handle disagreement. |
| 5 | Two or more co-founders with deep domain expertise and prior startup experience (at least one exit or scaled company). Strong early team with key hires in place. Demonstrated ability to recruit senior talent into an early-stage environment. Clear evidence of resilience through prior setbacks. |
Market (Weight: 25%)
| Score | Description |
|---|---|
| 1 | Small or shrinking market. No clear catalyst for growth. SAM under $100M with limited expansion potential. Timing argument is vague or absent. |
| 3 | Market is growing 10-20% annually with a credible SAM above $500M. Clear tailwind (regulatory, technological, or behavioral) creating a window. Two to three direct competitors exist, indicating real demand, but no dominant incumbent has locked up distribution. |
| 5 | Large, fast-growing market (20%+ CAGR) with SAM above $1B. Strong timing catalyst that the founding team identified early. Competitive landscape is fragmented with no clear winner. Multiple adjacent expansion opportunities beyond the initial wedge. |
Product (Weight: 15%)
| Score | Description |
|---|---|
| 1 | Concept stage with no working prototype. No user feedback beyond friends and family. Value proposition is unclear or indistinguishable from existing solutions. |
| 3 | Functional MVP with early users providing consistent feedback. Product solves a defined problem for a specific segment. Users can articulate why they prefer it to alternatives. Technical architecture supports iteration without full rebuilds. |
| 5 | Product is live with measurable engagement and retention. Users report it as a "must-have" (not "nice-to-have"). Clear technical moat or proprietary data advantage. Product roadmap is informed by usage data and customer input, with a credible path to platform or network effects. |
Traction (Weight: 20%)
| Score | Description |
|---|---|
| 1 | No measurable traction. No paying users, no LOIs, no pilot commitments. Engagement data is absent or based on vanity metrics (downloads, page views) with no connection to retention or revenue. |
| 3 | Early revenue ($5K-$50K MRR) or 3+ signed LOIs/pilot commitments from target customers. Retention data exists for at least 2-3 monthly cohorts. Growth is consistent (10-15% MoM) even if the base is small. CAC and LTV are estimated with supporting assumptions. |
| 5 | Strong revenue traction ($50K+ MRR or equivalent in annual contracts). Cohort retention curves flatten above 40% at Month 6+. LTV:CAC ratio above 3:1 with payback under 12 months. Multiple acquisition channels producing results. Revenue verified through third-party data (bank statements, Stripe dashboard). |
Business Model (Weight: 10%)
| Score | Description |
|---|---|
| 1 | No clear monetization strategy. Pricing is undefined. Unit economics are unknown. Revenue model depends on scale that hasn't been validated (e.g., "we'll monetize through ads once we hit 1M users"). |
| 3 | Defined pricing model tested with early customers. Gross margins above 50%. Unit economics are estimated and directionally positive. At least one pricing tier has been validated through actual transactions. Path to profitability is modeled, even if distant. |
| 5 | Proven pricing with multiple tiers or segments. Gross margins above 70%. Strong expansion revenue (upsells, cross-sells) demonstrated in existing accounts. Unit economics are verified and improving quarter over quarter. Clear operating leverage as the company scales. |
Frequently Asked Questions
Insert as a new section near the end of the article, before the internal linking map.
What do investors look for when evaluating a startup?
Investors evaluate five dimensions: team, market, product, traction, and business model. The relative weight shifts by stage. At pre-seed and seed, team and market carry 55% or more of the evaluation because product and traction data is limited. By Series A, traction and business model become primary filters. Gregory Shepard's framework, built from 2,200+ founder interviews and 12 personal exits, weights team at 30%, market at 25%, traction at 20%, product at 15%, and business model at 10% for early-stage evaluation. Investors who apply a consistent scoring rubric across these dimensions make faster, more defensible decisions than those who evaluate qualitatively.
How do you assess product-market fit during due diligence?
Product-market fit shows up in three measurable signals during diligence: net revenue retention above 100% (existing customers spend more over time), organic referral rates above 15% (users bring other users without paid incentives), and cohort retention curves that flatten rather than decay to zero. Investors should also conduct 5-10 direct customer interviews, asking specifically whether the customer would be "very disappointed" if the product disappeared. Sean Ellis's benchmark sets the threshold at 40% of surveyed users answering "very disappointed." Below that, the product solves a real problem but hasn't become essential. Founders who can't provide retention data segmented by cohort month are typically pre-PMF, regardless of what their pitch deck claims.
What is the single most important factor for early-stage startup evaluation?
Team quality. At pre-seed and seed, the product will change, the market thesis will sharpen, and the business model will evolve. The team is the one constant. Investors evaluate founders on four specific traits: domain expertise (do they understand the customer's problem from direct experience), execution speed (how fast have they moved from idea to current state), resilience evidence (have they navigated setbacks without quitting or fragmenting the team), and recruiting ability (can they attract strong people into an uncertain environment). The First Round Capital 10-Year Project found that co-founding teams outperformed solo founders by 163% in valuation growth, reinforcing that team composition, not just team quality, predicts outcomes.
How can investors verify a startup's traction claims?
Verification requires access to primary data sources, not founder-prepared summaries. Request read-only access to Stripe, Chargebee, or the payment processor dashboard to confirm MRR and growth trends. Ask for bank statements covering the trailing six months to verify that reported revenue matches actual cash received. Pull analytics exports (Mixpanel, Amplitude, or equivalent) to confirm active user counts and retention cohorts independently. Conduct 5-8 customer reference calls with contacts you select from the customer list, not contacts the founder curates for you. Cross-reference reported traction metrics against these primary sources. Discrepancies of 10-15% are normal (timing differences, accounting methods). Discrepancies above 30% are a red flag that requires explanation before diligence continues.
What are the biggest red flags in startup evaluation?
Five red flags consistently predict problems downstream. First, a solo founder with no plan to add a co-founder, which concentrates risk and limits the company's ability to operate during crises. Second, customer concentration above 40% in a single account, which means one lost contract can eliminate the growth story. Third, founders who resist providing raw data access during due diligence, which signals either disorganization or intentional obfuscation. Fourth, a pivot history of three or more major direction changes in under 18 months without a coherent narrative connecting them, which suggests the team is searching rather than building. Fifth, burn rates that exceed revenue growth rates for more than two consecutive quarters without a clear catalyst for the imbalance (like a one-time infrastructure investment), which indicates the company is spending faster than it's learning.
Sources
- CB Insights, The Top 12 Reasons Startups Fail, 2021. cbinsights.com
- First Round Capital, 10 Year Project: What 10 Years of Data Tell Us About Founders and Startup Success, 2015. 10years.firstround.com
- Alumni Ventures, A VC's Playbook: An Investor's Guide to Seed Investing, 2024. av.vc
