The Pitch-Deck Review: What a Venture Interview Actually Grades

Hand a candidate a real pitch deck, give them a few minutes, and ask what they think. That is the entire exercise, and it is graded on something almost no candidate expects going in: not whether they reach the right verdict, but whether they find the one or two things that actually determine the outcome, take a position on them, and say what would change their mind.

A candidate who reads back ten balanced considerations, five for and five against, and closes with "I'd want more diligence" has failed the exercise. Every one of those ten points is already visible on the slides, so naming them proves only that the candidate can read. What an interviewer cannot get from the deck itself is judgement, and judgement is the entire point of the job this exercise is standing in for.

This is the venture equivalent of the private equity modelling test, and it is scored on the opposite thing. A modelling test rewards getting the mechanics right against a knowable answer. A deck review rewards committing to a view before the mechanics exist to check it against, because at pre-seed and seed there is no dataset coming. Waiting for more evidence is not caution here. It is passing on everything.

What Actually Happens in the Room

One detail is worth naming precisely because it rarely gets asked: whether the deck under review is a company the fund actually backed, or one it has no connection to at all. Both happen, and interviewers treat the distinction as beside the point. What is being tested does not change either way, which is itself informative.

What Is Being Graded: Judgement, Not Verdict

The deck review looks like a comprehension test on the surface: read the deck, summarise it, note the strengths and weaknesses. It is not graded that way. One interview-prep account puts the actual criterion plainly: "Interviewers watch whether you grasp how hard a business is to scale and whether you ask sharp questions." The same account adds that showing you can quickly find the gaps in a pitch builds instant rapport, not enthusiasm for the company itself.

That framing matters because it names what the exercise is actually a proxy for. A junior investor's real job is mostly finding reasons to say no efficiently, then backing the rare company where the reasons do not hold. A deck review compresses that entire skill into a few minutes and watches whether a candidate can do it live, under mild pressure, without notes.

A strong answer sounds different from a comprehensive one in a specific way. It names the single metric or claim the whole thesis actually rests on, states plainly whether the candidate believes it, and stops there rather than hedging across every slide equally.

Read the Deck in the Order It Is Actually Read

Founders build a deck to be read start to finish. Working investors almost never read it that way, and a candidate who mirrors the founder's intended order rather than the investor's actual habit is already behind before the questions start.

Multiple independent accounts of how investors actually triage decks converge on the same handful of slides getting disproportionate attention regardless of where they sit in the deck. One founder-facing guide, drawing on data across tens of thousands of shared decks, found financials carry above-median attention despite appearing in under half of decks reviewed, and recommends opening on traction rather than burying it mid-deck.

A separate account of the same behaviour states it more bluntly: "Their eyes immediately jump to three slides: Traction, Market Size, and Team. If the story those slides tell isn't compelling, they close the file."

A candidate walking into a review can use the same triage the interviewer already runs, rather than working through the deck slide by slide as if presenting it for the first time:

1. Find the traction slide first, and check whether the metric on it is the one the business actually needs to prove, or merely the one that looks best.
2. Find the team slide next, and check whether the founders' specific background explains why they, and not someone else, should win this particular market.
3. Find the market-size slide, and check whether the number is built bottom-up from a customer count, or asserted from a single external report.
4. Scan everything else, competition, product, business model, only after those three, since they matter far less to the verdict than founders assume.
5. Name the one slide that would change your answer if the number on it were different, before the interviewer asks you to.

TechCrunch's own reporting on deck structure makes the same point from the founder's side: there is no universally correct slide order, and the strongest decks lead with whichever slide is genuinely the strongest, because investors decide early and read the rest looking for reasons to change their mind rather than form it from scratch.

State a Position With Its Falsifier Attached

Taking a position is necessary but not sufficient. "I'd back this" and "I'd pass on this" are both still just verdicts, and a verdict without a stated falsifier reads as a guess dressed up as confidence.

The detail that actually separates a prepared candidate from a rehearsed one is naming the single fact that would flip the answer. Consider a seed-stage vertical-software deck claiming strong month-over-month growth from a small existing customer base.

A weak answer states a verdict and stops: "The growth looks strong, I'd want to dig deeper, but this seems promising." A strong answer commits and names its own exit: "I'd back this if the growth is coming from new logos rather than expansion inside three accounts, and I'd pass if one customer is driving most of the curve, because that's not growth, it's one relationship."

Both answers used the same slide. Only one of them tells the interviewer what evidence would actually move the candidate, which is the part of the exercise that cannot be faked by reading more decks beforehand.

Stage Changes the Question

The single biggest mistake a well-prepared candidate makes is answering every deck the same way. What a deck review is actually testing shifts hard by stage, because what there is to underwrite shifts hard by stage.

Pre-Seed and Seed: Underwriting the Founders, Not the Financials

At pre-seed and seed, financial history is not thin. It is usually absent entirely, which means a candidate reaching for a spreadsheet is reaching for the wrong tool. Kruze Consulting's own framing of early-stage valuation puts it directly: investors at this stage are "largely underwriting the team, market, and product vision, not mature financials," because round size, traction and sector move the outcome far more than any model would.

A strong pre-seed or seed answer therefore spends almost no time on numbers, because there are barely any numbers to spend time on. It spends its time on whether the specific founders in front of the fund have an unfair reason to win this specific problem, and whether the market is large enough that winning it matters at venture scale even before any revenue exists to measure.

Series A: Repeatable Growth, or Founder-Driven Growth

By Series A, some history exists, and the question sharpens into something checkable rather than purely a bet on people. The central thing to establish is whether the growth on the traction slide would survive the founder stepping back from sales entirely.

A deck showing revenue climbing steadily while every named customer win traces back to a warm introduction from one founder is not showing repeatable growth. It is showing one very good salesperson with a company attached, and a candidate who treats that curve as evidence of a scalable go-to-market has missed the actual test embedded in the slide.

Series B: Where the Deck Review Starts to Look Like Growth Equity

By Series B, the exercise changes character almost entirely. Enough operating history now exists that the review stops resembling a bet on people and starts resembling a growth-equity underwrite: cohorts, payback periods and retention replace narrative as the evidence that actually matters.

Founder-facing guidance on what Series B investors expect converges on a consistent, if loosely sourced, set of benchmarks: commonly cited figures put healthy CAC payback under roughly 18 months and LTV-to-CAC above roughly 3:1 for a venture-backed software business at this stage, alongside a cohort-retention picture the founders can produce on demand.

None of these numbers is authoritative in the way a firm's own published rule would be, but the shift they describe holds everywhere: from a story about potential to a claim about efficiency that has to survive being multiplied.

What a Strong Answer Sounds Like Versus a Weak One

Adjectives do not teach this skill. Worked contrasts do.

The pattern across every row is the same. The weak column describes the slide. The strong column interrogates it, names a specific absence, and attaches a falsifier to the conclusion. Nothing in the strong column requires outside information the candidate does not have; it requires only reading the same slide more carefully than the weak answer bothered to.

The Classic Failure Modes

Most weak deck-review answers fall into a small number of repeatable patterns, and naming them in advance is the fastest way to stop making them under pressure.

• Reaching for a model when every input is a guess. A candidate who builds a discounted-cash-flow sketch on a pre-seed deck with no revenue is not demonstrating rigour. They are applying a tool built for a different kind of company, and an interviewer who has seen the trick before will ask the one question it cannot survive: where did that growth-rate assumption actually come from?
• The balanced, both-sides answer. Covered above and worth repeating precisely because it is the most common failure of all: a list of considerations with no conclusion reads as an inability to decide, not as thoroughness.
• Pitching a business that is safe, profitable, and cannot return a fund. This is the failure candidates least expect, because the underlying business genuinely looks good.
• Confusing "will this work" with "if it works, how big." These are different questions, and a deck can pass the first while failing the second.

Why "Safe and Profitable" Is Still a Fail

Venture math explains why the third failure mode is a real fail rather than a matter of taste. Peter Thiel's own framing in Zero to One states the mechanism directly: "the biggest secret in venture capital is that the best investment in a successful fund equals or outperforms the entire rest of the fund combined."

A fund built on that arithmetic cannot afford to back a company whose best realistic outcome is comfortable rather than enormous, no matter how well it is run.

"Lifestyle: too small for VC, but will generate enough annual cashflow to be a great business to own and operate... VC Fundable: large enough to justify and provide a return on a VC investment and the founder is willing to exit." — Fred Wilson, Union Square Ventures, drawing the line a deck review is quietly testing for

The fourth failure mode sits right beside the third and is easy to miss because it sounds like the same question asked twice. "Will this work" is about execution risk: can this specific team actually build and sell the thing. "If it works, how big" is about outcome size: does success here even matter to a fund that needs a small number of enormous results.

A candidate can correctly answer that a company will probably work and still be right to pass on it, because a probable, modest win is not what the fund's own math requires.

The hedge (fails the exercise)

• Reaches for a model when every input on the page is a guess
• Lists balanced considerations and stops at I would want more diligence
• Praises a well-run business without naming its ceiling
• Answers will this work without ever asking if it works, how big

The committed answer (passes)

• Names the single metric the whole thesis rests on and says plainly whether it holds
• States a position with a specific, checkable falsifier attached
• Says outright when a good business plan is not a fundable one
• Separates execution risk from outcome size, and answers both

The Market-Size Trap Inside the Deck

Of every slide in a venture deck, the total-addressable-market slide is the one most worth a direct, specific challenge, and it is also the one candidates most often let pass unquestioned because the number looks authoritative.

A survey of thirty investors published by Pear VC found that funds distrust a top-down TAM figure standing alone, treating it as, at best, a sanity check against a bottom-up build rather than a number worth citing on its own.

What Named Firms Actually Run

The mechanics above are general. What each named fund actually runs is specific, and worth stating plainly.

Two things generalise across all five. First, every firm naming a specific mechanic, Creandum's video, Antler's memo, Picus's analytics test, wants a specific kind of candidate to self-select in before it spends partner time on them.

Second, none of the five publishes an exact time allotment for the live portion of its process. Firms will name what they test far more readily than they will name how long a candidate gets to do it, which means the pressure itself is part of the exercise rather than an accident of scheduling.

How to Prepare, in Order

Preparation time for this exercise is scarce, and it rewards a specific kind of practice far more than general reading. In rough order of return on time invested:

1. Pull three or four real decks (a friend's fundraise, a public deck archive, your own portfolio-company exposure) and force yourself to state a position and a falsifier for each, out loud, in under five minutes. Recording yourself and listening back is uncomfortable and works better than rehearsing silently.
2. Practise the same read at every stage, not just one. Run a pre-seed deck, a Series A deck and a Series B deck through the same five minutes, and notice how the question you should be asking changes each time.
3. Rebuild one TAM slide bottom-up from a deck that only shows a top-down number, so the habit of interrogating a market-size claim is automatic rather than something you remember to do mid-interview.
4. Write down your own falsifier before you check whether you were right. The skill being tested is naming what would change your mind, not eventually reaching the correct verdict.
5. Read the named-firm mechanics you are actually applying to, since a candidate who prepares a five-minute deck pitch for a fund that runs a written memo round instead has prepared for the wrong exercise entirely.
6. Rehearse saying "I'd pass" out loud. Candidates who have only ever practised sounding enthusiastic about a deck are the ones who default to a hedge the moment the room pushes back, because they have never had to defend the less comfortable answer.

The Bottom Line

The pitch-deck review is not a comprehension test wearing a time limit, and it is not a smaller version of the private equity modelling exercise with the leverage removed. It is a compressed rehearsal of the actual job: reading an incomplete story, deciding what matters inside it, and being willing to say so before anyone hands you the data that would make the decision easy.

Stage changes what that story is made of, from founders and conviction at pre-seed to cohorts and payback at Series B, but the underlying test never changes. A candidate who can name the one fact that would flip their answer, at any stage, on any deck, has shown the interviewer the only thing the exercise was ever built to reveal.

That is a rehearsable skill, not a personality trait. Nobody arrives at their first review already comfortable saying "I'd pass" to a founder's ten-slide pitch. The candidates who sound comfortable saying it have simply done it more times before it counted.