How to Evaluate a Startup Offer: A 5-Input Decision

Evaluating a startup offer means converting a vague feeling about "salary plus some equity" into one comparable number. The scorecard has five inputs: cash gap vs market, equity expected value, benchmark deviation against the Index Ventures grid, round-and-founder signal, and 90-day reversibility. Anything below the walk-away threshold is a no.

Why "competitive salary + meaningful equity" tells you nothing

Standout sees the same pattern every week. A senior tech professional reads a one-line offer summary, takes a friendly Zoom from a founder, and signs inside 72 hours. The math underneath the headline gets done six months later, at the wrong end of the cliff.

The base rates make the case for doing the math now. A Correlation Ventures sample of ~27,000 venture-backed investments found 64% failed to return principal (Source: Correlation Ventures study (cited widely; original sample n≈27,000)). Harvard's Ghosh study on 2,000+ VC-backed companies puts the full-liquidation rate at 30-40% (Source: Shikhar Ghosh / Harvard Business School). Those are the deals where every dollar of employee equity expires worthless. Carta's 2025 dataset shows the median Series A round dilutes existing shareholders by 18% in one closing (Source: Diego Serebrisky / Latin American VC (citing Carta 2025 data on 2,005 US software startups)), with one in ten rounds diluting by 30%+ (Source: Carta 2025 / Serebrisky summary).

Hot take number one: the modal outcome on a startup equity grant is zero. Any model that doesn't start from that assumption is a marketing brochure for the company, not an offer evaluation.

The scorecard fixes that. Each input either compares to a published benchmark or forces a verifiable question to the founder. The pattern Standout candidates report after running their last three offers through it is consistent: one of the three would have scored a hard no, and the candidate signed it anyway.

Input 1: Cash gap

The cash gap is the dollar amount of base salary the offer leaves on the table versus the closest like-for-like alternative. Equity has to beat the cash gap on expected value, or the offer is a paid cut plus optionality.

Index Ventures' Rewarding Talent data anchors senior tech roles at seed-stage US startups at an average $185,000 base, up 68% across three years (Source: Index Ventures , Rewarding Talent). Big-tech bands sit higher; growth-stage scale-ups split the difference. The right comparator is the offer the candidate has, or could realistically generate inside the next 30 days, not "what FAANG would pay at L6" in the abstract.

If the annual gap is under 20% of total cash comp, the cash side passes. If the gap is over 35% with nothing in equity to justify it, the cash side hard-fails on its own.

Input 2: Equity expected value

Most candidates compute equity wrong by a factor of two to four. The standard mistake: multiply percentage of the company by the most recent post-money preferred valuation, treat that as the grant's "value," done. That ignores three things.

First, the strike price. The 409a fair market value used to set the option strike is a separate appraisal from the preferred share price. For early-stage startups, common typically prices at 25-60% of the latest preferred (Source: Qapita , 409A vs Preferred Price), because preferred carries liquidation preferences and protective rights common does not.

Second, dilution between today and exit. Carta's 2025 medians: Seed 19.5%, Series A 18%, Series B 14%, Series C 10%, Series D 7.5% (Source: Diego Serebrisky / Latin American VC (citing Carta 2025 data on 2,005 US software startups)). The modal Seed/Series A band is 20-24%, with ~10% of rounds diluting by 30%+ (Source: Carta 2025 / Serebrisky summary). A grant taken at Series A will see at least one further dilution event before any liquid exit, usually two.

Third, the probability the equity converts. The Correlation Ventures base rate is 64% of deals returning less than capital invested (Source: Correlation Ventures study (cited widely; original sample n≈27,000)); Ghosh puts full liquidation at 30-40% (Source: Shikhar Ghosh / Harvard Business School). Any expected-value calculation has to be probability-weighted before it hits the scorecard.

The corrected formula:

``` Expected equity value = (% fully diluted) × (preferred price × 409a discount) × (1 − cumulative future dilution) × (probability of exit at preferred valuation or above) × (vested fraction at exit) ```

Hot take number two: do this math out loud the next time a founder presents the equity number. If the founder cannot give a current 409a price, a current preferred price, and a fully-diluted share count, the grant is not a number. It is a story. Postpone the decision until it is a number.

Worked example. Senior IC, 0.25% over four years at a Series A company. Latest preferred at $10/share; latest 409a at $4/share (40% of preferred, mid-band of the 25-60% range (Source: Qapita , 409A vs Preferred Price)). Future Series B + C dilution strips ~24% combined at Carta medians (Source: Diego Serebrisky / Latin American VC (citing Carta 2025 data on 2,005 US software startups)). Probability of preferred-or-better exit: a generous 25%, above the Correlation Ventures base (Source: Correlation Ventures study (cited widely; original sample n≈27,000)). Vested fraction at four years: 100%.

On a 100M fully-diluted cap table, the grant is 250,000 options at $4 FMV: a notional $1M at today's common valuation. After dilution adjustment, ~$760K. After probability weighting, ~$190K risk-adjusted over four years, or ~$47K/year. If the cash gap is over $47K, the equity is not making up the difference. That is the math the headline hides.

Input 3: The Index Ventures benchmark grid

Most candidates have never seen Index Ventures' Rewarding Talent grid, which is the single most useful public benchmark for whether a grant is fair. The relevant page sizes Series A option grants as a percentage of base salary, not as a percentage of company (Source: Index Ventures , Rewarding Talent: Equity for All). That reframe is the load-bearing correction on this entire topic.

A senior engineer joining a Series A startup at $200K base should expect option value of ~$100K at grant. Against a common stock at 40% of preferred, that converts to a percentage of company in the low tenths of a percent, not the 1-2% that internet folklore keeps circulating. Marketing, sales, and operations roles get sized at roughly half the engineering equivalent at the same level.

The deviation score is actual grant ÷ benchmark. Between 0.7× and 1.3×, the grant is in the fair zone. Below 0.5× with no make-good on cash or refresh schedule, the grant is structurally light and the company is using the title to substitute for compensation. Above 1.3× usually signals an early grant or a hiring premium worth taking.

Hot take number three: most negotiation advice tells candidates to "ask for more equity." The Index Ventures grid lets the candidate ask for a specific number with specific math underneath it. The number gets respect; the vague ask doesn't.

Input 4: Round and founder signal

The same dollar grant means different things at different rounds; the round itself is the signal. Carta's medians: founders retain 56.2% after Seed, 36.1% after Series A, 23% after Series B (Source: Carta , Founder Ownership Report 2026). By Series C, the median employee option pool (16.8%) exceeds median founder ownership (16.1%) (Source: Carta , Founder Ownership Report 2026). The cap table has flipped, and later-stage offers should be benchmarked against the pool, not the founder stake.

Questions the round itself should answer:

• Who led? A top-decile lead means diligence has been done; the candidate is consenting to the same thesis as the investor. A no-lead party round or an undisclosed lead means the company shopped a deck and accepted whatever closed. Score it down.
• How much was raised, and what is it for? 18 months of runway funding one concrete milestone is healthy. Indefinite runway to "scale the team" without naming the milestone is a round raised to keep the lights on.
• Founder track record. A second-time founder with a prior outcome (acquisition, IPO, even a clean wind-down) is a different risk profile than first-timers. The base rates shift, but they don't reverse.

Standout's view from running matches with hiring teams at recently-funded US startups: these questions take fifteen minutes to research and another fifteen to ask in the offer call. The candidates who reach signed offer in under five weeks treat the diligence as their job, not the recruiter's.

Hot take number four: a candidate who can't answer "what kills this company in the next 18 months" after a single offer call doesn't have enough information to sign.

Input 5: 90-day reversibility, the part candidates skip

Reversibility is the cost of being wrong fast. Most offer evaluations assume the candidate gives the role four years to play out. The base rate says otherwise: a meaningful fraction of senior hires leave inside the first year, usually because something invisible from outside became visible from inside.

Standard vesting is four years with a one-year cliff: 25% vests at 12 months, then 1/48 of the grant monthly for the next 36 (Source: Carta , Vesting Explained). Leaving before the cliff means walking with zero vested equity. A two-year cliff is non-standard and should be pushed back on or compensated in cash.

Other reversibility lines to read:

• Signing-bonus clawback: standard at 12 months. Past 18 months is a tax on optionality; counter it.
• Non-compete and non-solicit: largely unenforceable for California employees by statute, but companies still draft them. They cost in litigation even when they fail in court.
• IP assignment scope: the wider the IP capture, the harder it is to ship a side project or take adjacent contract work after departure.
• Acceleration on change of control: double-trigger (acquisition + involuntary termination) is the candidate-friendly default. No acceleration is the founder-friendly default. Push for at least single-trigger on involuntary termination, double-trigger on acquisition.

The under-discussed reversibility line is AMT on early exercise. Exercising ISOs before sale pulls the spread between strike and current 409a FMV into Alternative Minimum Taxable Income (Source: Startup Law Blog , ISO vs NSO Complete Guide (2026)). The One Big Beautiful Bill Act (July 4, 2025) reverted the AMT exemption phase-out to 2018 levels ($500K single, $1M MFJ) and doubled the phase-out rate to 50¢ per dollar above threshold (Source: Startup Law Blog , ISO vs NSO Complete Guide (2026)). A senior engineer in California who early-exercises a meaningful grant can produce a five-figure AMT bill before the company has done anything wrong.

Running the scorecard end-to-end

Senior backend engineer, $190K base + 0.25% over four years at a Series A AI startup, $20M raised, top-decile lead, two-time founder.

• Cash gap: market reference at a comparable SF growth-stage company is ~$240K base (Source: Index Ventures , Rewarding Talent). Gap = $50K/year, ~21% of cash side. Above the 20% pass threshold by one point. Borderline pass.
• Equity EV: 0.25% of a 100M-share fully-diluted cap table = 250K options. Strike at 409a $4 vs $10 preferred (Source: Qapita , 409A vs Preferred Price). Future Series B+C dilution ~24% (Source: Diego Serebrisky / Latin American VC (citing Carta 2025 data on 2,005 US software startups)). Outcome probability 25%, above the Correlation Ventures base (Source: Correlation Ventures study (cited widely; original sample n≈27,000)). EV over four years: ~$190K, or ~$47K/year. Equity is neutralizing the cash gap, not beating it. Borderline pass.
• Index Ventures deviation: senior engineer benchmark is 50% of base (Source: Index Ventures , Rewarding Talent: Equity for All). Target $95K/year option value; actual $47K/year. Deviation 0.5×, at the floor. Negotiate the grant up to 0.35% before signing.
• Round + founder: top-decile lead, defined milestone, second-time founder with prior outcome. Strong pass.
• Reversibility: standard 4-year/1-year cliff, no acceleration, 12-month signing bonus clawback. Push for double-trigger acceleration. Borderline pass.

Verdict: four borderline, one strong. The math says counter to 0.35-0.40% and double-trigger acceleration before signing, with a walk-away if the counter is fully refused. The same offer evaluated on "good company, sounds exciting" reads as a clear yes. On the scorecard, it has a specific gap and a specific ask that closes it.

The five mistakes that flatter every offer

Ranked errors that show up most often in offers Standout candidates run past us:

1. 1Anchoring on share count, not percentage fully diluted. Shares are meaningless until divided by the fully-diluted count. The question is what % of the company the grant represents, and how that % dilutes by exit (Source: Diego Serebrisky / Latin American VC (citing Carta 2025 data on 2,005 US software startups)).
2. 2Pricing options at the preferred share price. Common is not preferred. The 409a typically prices common at 25-60% of the latest preferred (Source: Qapita , 409A vs Preferred Price). Multiplying % of company by the preferred valuation overstates the grant by 2-4× at early stage.
3. 3Treating a fresh round as derisking. A new round dilutes existing shareholders 14-19% per Carta medians (Source: Diego Serebrisky / Latin American VC (citing Carta 2025 data on 2,005 US software startups)) and raises the valuation the eventual exit has to clear. New money is not new safety.
4. 4Skipping cliff and clawback fine print. Leaving at month 11 means zero vested equity (Source: Carta , Vesting Explained). A clawback past 18 months is a real cost too. Read both before signing.
5. 5Counting headline total comp without probability adjustment. $1M face at common is not $1M of EV. After dilution, AMT, and the 30-40% liquidation base rate (Source: Shikhar Ghosh / Harvard Business School), the risk-adjusted number is usually a fifth of the face.

How Standout fits

Standout matches US tech professionals with hiring companies directly. Candidates do not apply; we match the candidate with a company, and on consent we introduce the candidate to the founder for a direct conversation (Source: Standout). Free for candidates, placement-fee-only on the company side; first matches typically arrive within hours of profile completion (Source: Standout).

Three things to keep clear:

• Standout is all-roles, not engineering-only (product, design, data, ML/AI, DevOps, marketing, sales, ops, CS, BD) (Source: Standout).
• Standout is US-only as of Q2 2026 (Source: Standout).
• Standout is not a job board. Matches arrive based on profile + founder conversations (Source: Standout).

The reason the scorecard matters more inside Standout than outside is the comparison set. Running five inputs against one offer in isolation produces a yes/no the candidate has to defend against no counterfactual. Running the same inputs against three concurrent offers produces a relative ranking, a negotiation point on the front-runner, and a walk-away that is credible because it is collateralized by the second-place offer. The matching model exists to produce the second offer, not just the first.

Adjacent reading: let recruiters come to you, YC startup hiring, and how Standout's matching works.

FAQ

Is a 1% equity grant at Series A a good offer?

For most senior individual contributors, no. The Index Ventures grid sizes Series A engineering grants at 33% of base salary for an IC and 50% for a senior, which translates to roughly 0.1-0.4% of the company at typical Series A valuations (Source: Index Ventures , Rewarding Talent: Equity for All). A 1% grant at Series A is usually only normal for the first non-founder hire or an executive role.

How do I calculate the real value of startup equity?

% fully diluted × 409a price (typically 25-60% of preferred (Source: Qapita , 409A vs Preferred Price)) × (1 − cumulative future dilution per Carta medians: 19.5/18/14/10/7.5% across Seed-D (Source: Diego Serebrisky / Latin American VC (citing Carta 2025 data on 2,005 US software startups))) × probability of preferred-or-better exit (generous benchmark 25%, above the 36% return-principal rate from Correlation Ventures (Source: Correlation Ventures study (cited widely; original sample n≈27,000))).

What's the difference between 409a value and preferred price?

The 409a is an IRS-prescribed appraisal of common stock fair market value, used to set the option strike price. The preferred price is what the most recent investor paid for preferred shares with liquidation preferences and protective rights. For early-stage startups, common typically prices at 25-60% of preferred (Source: Qapita , 409A vs Preferred Price).

What happens to my unvested equity if I leave before the 1-year cliff?

It returns to the option pool. Standard vesting is four years with a one-year cliff: 25% of the grant vests at the 12-month mark, 1/48 monthly after that (Source: Carta , Vesting Explained). Pre-cliff, zero equity has vested, and the entire grant is forfeited on departure.

Should I early-exercise my ISOs to start the capital gains clock?

Only if the AMT math works. Exercising an ISO before sale pulls the spread between strike and current 409a fair market value into AMTI. The 2025 OBBBA lowered the AMT exemption phase-out to $500K single / $1M MFJ and doubled the phase-out rate (Source: Startup Law Blog , ISO vs NSO Complete Guide (2026)), which raises AMT exposure on the same spread compared to pre-2025 rules. Run the calculation before deciding.

[Get matched with the next startup offer worth running through this scorecard.](https://standout.work) Standout introduces tech candidates directly to founders at vetted US tech companies. Free for candidates. First matches within hours.