How Venture Capitalists Make Investment Decisions: Process and Criteria
A practical look at how VC firms move from thesis fit and diligence to return math, partner advocacy, and the final investment-committee decision.
Venture capital firms decide which companies to invest in by asking two different questions: could this become a venture-scale company, and is it the right investment for this specific fund? A startup must usually clear both tests. Strong founders, a large market, differentiated technology, and traction are not enough if the company falls outside the fund’s mandate or cannot produce a meaningful return at the proposed price and ownership.
The decision normally moves through seven stages: thesis fit, initial screening, meetings, due diligence, return and portfolio analysis, an investment memo, and final partner or investment-committee approval. The exact process varies by firm. Some partners can approve their own checks; other firms vote or require broad consensus.
The largest academic study of the process surveyed almost 900 venture investors and found that VC work spans sourcing, selection, valuation, deal structure, portfolio support, exits, and internal firm organization—not a single spreadsheet or “gut feel.” The study is available as an NBER working paper, with an accessible overview from Stanford Graduate School of Business.
The decision in one line: thesis fit gets a company reviewed; evidence and venture-return potential get it diligenced; conviction and partnership governance get it approved.
How VCs decide which companies to invest in
VCs reject most opportunities quickly because the fund’s constraints narrow the field before deep company analysis begins. The fund may invest only in a particular stage, sector, geography, business model, or check-size range. Its portfolio may already contain a close competitor. Its ownership target may not work with the round. None of those answers says the startup is bad.
Once an opportunity fits, the decision becomes an evidence problem. The deal team tests the founders, customer pain, market, product, technology, traction, go-to-market motion, economics, and risks. It then asks whether plausible outcomes could matter to the fund after future dilution.
Finally, the decision becomes an organizational problem. A partner or deal lead must turn the evidence into a recommendation, expose the weak points, and build enough conviction among the people who control the firm’s capital.
The seven-stage VC investment decision process
1. Thesis and mandate screen
The first screen is usually fit, not quality. The investor checks the opportunity against the fund’s investment thesis: stage, sector, geography, business model, check size, ownership target, and any exclusions.
This is why broad, untargeted fundraising performs poorly. A cybersecurity seed fund and a generalist growth fund can reach opposite conclusions about the same company without disagreeing on its quality.
2. Initial screen and partner interest
Opportunities arrive through referrals, outbound research, events, inbound applications, and systematic deal sourcing. At the initial screen, an investor looks for a sharp problem, a credible team, a plausible market, and enough evidence to justify more time.
A positive screen also needs internal ownership. Someone must be willing to become the deal’s champion: arrange meetings, frame the case, coordinate diligence, and defend the recommendation.
3. First meetings and an evidence map
Early meetings test whether the story survives questions. The deal team should leave with three lists:
- What it currently believes.
- What it doubts.
- What evidence would change its mind.
This evidence map is more useful than a long, generic checklist. A seed investor may need to test whether customers feel the problem urgently. A Series B investor may accept that the problem exists and focus instead on repeatable acquisition, retention, and operating leverage.
4. Due diligence
VC due diligence converts the open questions into workstreams. Depending on the company and stage, those may include customer references, market analysis, product and technical review, founder references, financial analysis, legal review, and cap-table inspection.
Good diligence is not an exercise in collecting documents. It should resolve the few assumptions that can change the decision. If a company’s case depends on enterprise expansion, cohort behavior and customer calls matter more than polishing a generic market-size slide.
5. Return case and portfolio fit
VCs invest from a portfolio, so they must translate company potential into fund outcomes. They test the entry price, likely ownership, future dilution, capital needs, exit scenarios, follow-on reserves, and overlap with existing investments.
The question is not only “could this company be valuable?” It is “could our stake become valuable enough to matter to this fund?” The venture capital method and the fund’s portfolio strategy provide deeper treatments of that math.
6. Investment memo and partner debate
The deal lead turns the case into an investment memo. A useful memo names the decision requested, evidence, assumptions, return logic, risks, disconfirming facts, and unresolved questions. Its job is not to sell the company. Its job is to make the decision inspectable.
Partner debate often begins well before the formal meeting. Skeptical questions may trigger more diligence, a change in terms, a smaller check, or a pass.
7. Investment committee, terms, and final approval
The firm applies its approval rules: individual partner authority, a vote, a supermajority, or consensus. If approved, the team still needs acceptable terms. Price, ownership, governance rights, liquidation preferences, and other provisions can turn an attractive company into an unattractive deal—or resolve enough downside to proceed.
The two-pass scorecard VCs use
A practical evaluation separates company quality from fund fit. Collapsing both into one score hides why the deal should proceed or stop.
Pass one: Is this a venture-backable company?
| Dimension | Core question | Useful evidence | Common weak signal |
|---|---|---|---|
| Founding team | Can this team learn, recruit, and execute through uncertainty? | Relevant insight, pace of learning, references, complementary skills | Prestige without evidence of judgment or speed |
| Market and timing | Can the market support a large outcome, and why now? | Bottom-up customer count, spend, adoption drivers, regulatory or technical change | A large top-down market report |
| Customer problem | Is the pain urgent enough to change behavior or budget? | Customer calls, usage, paid pilots, short sales cycles, strong retention | Enthusiastic survey responses with no commitment |
| Product and technology | Does the product create a material advantage? | Demonstrated workflow improvement, technical feasibility, adoption, proprietary data or know-how | Novelty presented as defensibility |
| Traction and economics | Is demand real and improving? | Cohorts, retention, revenue quality, gross margin, payback, expansion | Cumulative sign-ups or one-off revenue |
| Go-to-market | Is there a credible path to repeatable distribution? | Defined buyer, working channel, sales productivity, referral loops | “We will hire salespeople” |
| Defensibility | Can the company preserve value as competitors respond? | Workflow depth, network effects, switching costs, brand, data advantage | A feature lead that can be copied |
| Key risks | What could break the case? | Named failure modes, tests, mitigation, honest unknowns | A risk slide that restates strengths |
Team quality often matters heavily, particularly early. Stanford’s summary of the major VC survey reports that the average respondent evaluated roughly 200 companies a year and invested in about four, and that team qualities ranked above product or technology for many investors. That does not mean “team always wins.” At later stages, weak retention or poor economics cannot be rescued by founder charisma.
Pass two: Is this the right investment for this fund?
| Dimension | Decision question |
|---|---|
| Thesis fit | Does the company match the fund’s explicit mandate and edge? |
| Check and ownership fit | Can the fund deploy the intended check and reach a useful stake? |
| Outcome materiality | Could plausible outcomes return meaningful capital to this fund? |
| Follow-on requirements | How much reserve capital may be needed to protect or increase ownership? |
| Price and terms | Does the entry valuation leave room for the required return, and are the terms acceptable? |
| Portfolio fit | Does the deal create concentration, conflict, or unwanted correlation? |
| Ability to win and add value | Can the firm earn an allocation and materially help after investing? |
The distinction matters. A company can pass every operating-quality test and still fail because the round is too small for the fund, the price compresses the return case, or an existing portfolio company creates a conflict. Conversely, a perfect mandate fit does not compensate for weak customer evidence.
What evidence matters at each stage
The criteria stay recognizable across stages, but the standard of proof changes. Pre-seed investing is largely a judgment about people, insight, and the speed of learning. By growth stage, the company has enough history for investors to test repeatability and efficiency.
| Dimension | Pre-seed / seed | Series A | Growth |
|---|---|---|---|
| Team | Founder insight, learning speed, complementary skills | Evidence the team can recruit and operate a company | Leadership depth, executive gaps, organizational performance |
| Customer problem | High-quality discovery, design partners, early use | Retention, references, repeat usage, willingness to pay | Durable demand across segments and markets |
| Product / technology | Feasibility, differentiated insight, fast iteration | Reliability, adoption, roadmap, technical scalability | Architecture at scale, security, resilience, technical debt |
| Traction | Early engagement, pilots, initial revenue where relevant | Cohort retention, growth quality, repeatable use cases | Predictable revenue, expansion, churn, segment performance |
| Unit economics | Directional logic and major cost drivers | Early payback, gross margin, sales efficiency | Proven contribution economics and operating leverage |
| Market | Credible wedge into a large or expanding market | Evidence the wedge can broaden | Market share, adjacency potential, international expansion |
| Go-to-market | Founder-led learning and a plausible initial channel | Early channel repeatability and a defined buyer | Scalable sales/marketing system and channel mix |
| Principal risk | Problem and product risk | Repeatability and market risk | Execution, competition, capital efficiency, and exit risk |
How VCs evaluate technology and innovation
Technical novelty is only one input. Investors normally need to understand five things:
- Feasibility: Can the system do what the company claims?
- Customer value: Does the technology improve cost, speed, quality, access, or an otherwise important outcome?
- Adoption: Can customers integrate it without unacceptable workflow, security, or behavior change?
- Scalability: Will performance and economics hold as usage grows?
- Defensibility: Does the advantage deepen through data, distribution, workflow, intellectual property, or accumulated know-how?
At seed, expert references and a working prototype may carry substantial weight. At growth stage, the same claim needs production evidence: reliability, customer retention, gross margins, implementation time, and a roadmap that survives larger competitors.
“Innovative” is not a decision criterion by itself. The technology must create a valuable and durable business advantage.
How investment committees make the final decision
There is no standard VC investment committee. Origin Ventures spoke with 15 funds and grouped the approaches into three broad models.
| Model | How it works | Strength | Failure mode |
|---|---|---|---|
| Partner checkbook | A general partner can approve a deal within an agreed mandate or check-size limit | Fast; preserves individual conviction | Weak shared ownership when a deal struggles |
| Majority or scored vote | Partners vote, sometimes after completing a scorecard | Clear decision rule and recorded views | Scores can create false precision or become ceremonial |
| Consensus or supermajority | Most or all decision-makers must support the deal | Broad ownership and multiple perspectives | Slow decisions and a bias toward unobjectionable deals |
The formal rule tells only part of the story. The deal’s champion usually socializes the opportunity, answers objections, and recruits support before the meeting. A strong meeting rarely compensates for weeks of weak internal advocacy.
Approval thresholds may also change. A small seed check might sit within one partner’s authority, while a larger initial investment or a difficult follow-on may require broader support. A firm can demand more conviction when the decision consumes more capital or creates a larger portfolio concentration.
Corporate venture capital can add another layer. Research summarized by Stanford GSB found that many CVC units use a two-tier process, with an approval group outside the investment team. That structure can add strategic criteria and slow the decision relative to an independent fund.
For founders, the practical questions are straightforward:
- Who is leading the deal internally?
- Who must approve it?
- What questions remain open?
- When is the next decision point?
- Could a different check size or deal term change the answer?
An investor who cannot explain the next step may still be interested, but the deal is not yet close to a decision.
What goes into a VC investment memo
The investment memo forces a deal team to convert enthusiasm into a falsifiable case. Formats vary, but a useful one-page version answers eleven questions.
| Memo block | Question to answer |
|---|---|
| Decision requested | Approve diligence, issue a term sheet, invest, reserve capital, or pass? |
| Company and round | What is the company raising, at what stage, and for what use? |
| Thesis fit | Why does the deal belong in this fund? |
| Why now | What changed in technology, regulation, behavior, or distribution? |
| Team | Why is this team unusually suited to the problem? |
| Market and wedge | Which initial customer and use case create a path into a large outcome? |
| Product / technology | What is differentiated, valuable, feasible, and defensible? |
| Evidence | Which customer, usage, revenue, retention, or technical signals support the case? |
| Return case | What ownership can the fund reach, what dilution is plausible, and which outcomes matter? |
| Risks and open questions | What evidence contradicts the case, and what remains unknown? |
| Recommendation | Proceed, proceed with conditions, continue diligence, or pass—and why? |
Three disciplines make a memo useful.
First, label facts, estimates, assumptions, and unknowns. “Customers save four hours per week” is not a fact unless the team has measured it. “The market will consolidate around three vendors” is an assumption and should be treated as one.
Second, include disconfirming evidence. If two customer references praise the product but refuse to expand their contracts, that belongs near the positive evidence—not in an appendix.
Third, connect each open question to a decision. “Validate security readiness before a term sheet” is stronger than “security is a risk.” It tells the team what to test and what happens if the result is weak.
Worked example: from interesting startup to an IC recommendation
Consider a fictional seed-stage company, Northstar Workflow, selling approval software to finance teams. The assumptions below are illustrative; they are not benchmarks.
Initial screen
The positive signals are credible:
- The founders previously managed the same approval process as buyers.
- Six design partners use the product weekly.
- Four customers pay, and two have expanded usage to a second team.
- The product replaces a visible mix of email, spreadsheets, and manual follow-up.
The company appears venture-backable, but the fund-fit answer is not settled. The round is small relative to the fund’s normal initial check, and the fund already owns a business selling adjacent finance software.
Diligence questions
The deal team identifies three questions that can change the recommendation:
- Is the problem urgent? Interview paying users and two churned pilots. Look for budget ownership and evidence that the product changes cycle time or error rates.
- Is expansion repeatable? Separate founder-driven exceptions from a consistent second-team use case.
- Is the portfolio overlap manageable? Compare buyers, product roadmaps, data access, and conflict obligations with the existing investment.
Return and risk logic
The base case assumes the company wins a narrow finance workflow but expansion is slow. The upside case requires a repeatable move into adjacent approval processes. The downside case is not “growth is lower”; it is that the product remains a useful feature with limited pricing power.
The fund should not hide those branches inside one forecast. Each branch needs an observable sign:
| Case | What must be true | Early signal |
|---|---|---|
| Upside | The workflow expands across teams and customers | Multiple customers add a second use case without bespoke implementation |
| Base | Core finance use remains valuable but narrow | Strong retention with modest account expansion |
| Downside | The product is easy to replicate or too light to own budget | Usage persists but willingness to pay stalls |
Recommendation
Continue diligence with conditions. Northstar clears the company-quality screen, but the deal lead should not request a term sheet until the fund resolves portfolio conflict and confirms that expansion is a pattern rather than two anecdotes.
That conclusion is more useful than a blended score of 8.2 out of 10. It states what the team believes, what can still break the case, and what evidence unlocks the next decision.
What founders should do with this framework
1. Target mandate before polishing the pitch
Research the firm’s current stage, sector, geography, check range, portfolio, and stated thesis. Venture Capital Careers’ companies directory is a useful starting point; verify the firm’s current website before outreach because mandates change.
2. Replace adjectives with evidence
“Massive market,” “world-class team,” and “proprietary technology” create questions, not conviction. Translate each claim:
- Massive market → customer count × credible annual spend.
- World-class team → specific insight, execution speed, recruiting ability, and references.
- Proprietary technology → measured customer value plus a reason the advantage compounds.
3. Make the venture-return case legible
Founders do not need to reverse-engineer a fund’s exact model, but they should understand why the company could become large enough for the investor’s portfolio. Connect the initial wedge to expansion, capital needs, and the kind of outcome the strategy can support.
4. Ask about decision ownership
After substantive interest develops, ask who is leading the deal, which questions remain, who else must engage, and what the next decision point is. This is not an attempt to control the process. It is basic pipeline clarity.
5. Diagnose the “no”
A rejection can mean:
- Mandate no: wrong stage, sector, geography, check, or conflict.
- Evidence no: a critical claim is unproven.
- Return no: the ownership, price, dilution, or outcome does not work for the fund.
- Risk no: diligence surfaced an issue the firm cannot accept.
- Conviction no: no partner is willing to champion the deal.
The right response depends on the category. More traction may solve an evidence no. It will not solve a mandate no, and lowering the valuation may not solve a weak customer problem.
How aspiring VC investors can practice the decision
Learning the vocabulary is not enough. Practice turning incomplete evidence into a recommendation.
- Write the one-line fit screen. State the fund stage, sector, geography, check, ownership, and exclusion logic. Pass quickly when the company does not fit.
- Build the two-pass scorecard. Evaluate company quality first, then fund and deal fit. Do not average the two into a meaningless total.
- Name the disconfirming evidence. Identify the one or two facts most likely to invalidate the case.
- Test outcome materiality. Use transparent assumptions to check whether a plausible stake could matter to the fund after dilution.
- Make a decision request. Recommend a pass, another meeting, targeted diligence, or an investment—then state the conditions.
The strongest case-study answers distinguish what is known from what is merely plausible. They also spend time efficiently. A two-hour analysis should not pretend to have the certainty of a six-week diligence process.
To build pattern recognition, compare how firms describe their theses and portfolios in the Venture Capital Careers companies directory, then read the firms’ own investment pages and “why we invested” posts. When you can explain why the same startup fits one fund and not another, your analysis is becoming fund-specific.
If you are applying for investment roles, browse current opportunities on the Venture Capital Careers job board. In interviews, the memo structure above gives you a cleaner way to discuss sourcing, case studies, and investment judgment than a list of generic criteria.
The central habit is intellectual honesty: make the best decision available from the evidence, show what would change it, and resist turning conviction into certainty.
Frequently asked questions
Do VCs care more about the team or the idea?
At the earliest stages, team quality and founder insight often carry more weight because product and market evidence are limited. As a company matures, retention, economics, distribution, and execution become harder evidence. A strong team can earn more time to investigate; it cannot indefinitely compensate for a weak market or product.
Why do strong startups still get rejected?
VC decisions are fund-specific. A strong startup may be outside the mandate, too early or late, too small for the required check, in conflict with the portfolio, priced beyond the fund’s return case, or unable to offer enough ownership. A rejection is not always a verdict on company quality.
How long does a VC investment decision take?
There is no universal timeline. It depends on stage, check size, competition for the deal, diligence complexity, partner schedules, and the firm’s approval model. A smaller check within one partner’s authority can move quickly; a regulated company or a large investment requiring consensus can take longer.
Who makes the final decision at a VC firm?
It may be an individual general partner, a small investment committee, or the full partnership. Some firms vote; others require a supermajority or consensus. Founders should ask their internal contact who owns the recommendation and who approves it.
How are corporate VC decisions different?
Corporate VC teams may evaluate strategic fit as well as financial return and can require approval from executives outside the investment team. That additional layer can change the criteria, timeline, and people who must support the deal.
