A pitch deck audit is a structured, skeptical read of your deck that finds the weak assumptions, missing evidence and unanswered questions an investor will find anyway, ideally days before the meeting instead of during it. The useful kind returns a ranked list of what to fix, not a pat on the back. Below you get a checklist you can run yourself tonight, a way to prioritize the fixes, what outside review costs, and how to rehearse the answers.
TL;DR
- An audit is worth it only if it attacks the deck: assumptions, metric quality, the ask and the story, not fonts and colors.
- Run the checklist below in order: problem, market, traction, business model, financial logic, the ask, then narrative flow.
- Fix in priority order: clarity and proof of demand first, design last.
- Reading a deck is only half the work. Rehearse your answers to the hardest questions out loud.
- Our Tactical Audit is $25 per audit, and Founder Pro is $49 per month.
Table of Contents
- What a pitch deck audit actually is
- The pitch deck audit checklist
- Run the three-minute stranger test
- How to critique your own deck for weak assumptions
- How to prioritize improvements
- What to look for in a useful audit or critique
- How much does a pitch deck review cost?
- Prepare for the questions, not just the slides
- Perspective
- Put the audit to work: Dialectic
- A simple audit routine for the week before your meeting
- FAQ
- Sources
What a pitch deck audit actually is
Think of it as a pre-mortem. You assume the investor passes, then work backward to the reasons. Most passes come from a short list: the problem is vague, the market number is a top-down guess, the traction is hard to interpret, the unit economics are missing, or the ask does not connect to a milestone.
A good audit does three things:
- Finds fragile claims. Any sentence that collapses if one assumption turns out wrong.
- Finds gaps. What an investor will want to know that the deck never says.
- Predicts the questions. The specific, uncomfortable things a partner will ask, so you are not improvising.
It is different from a design review and different from friendly feedback. Friends tell you the deck looks clean. An investor tells you, or more often does not tell you, why they are not moving forward.
One point of perspective helps here. Sequoia has said that what it valued in the business plans it received was the ideas, the clarity of thinking and the scope of ambition, not the slides themselves. And a deck on its own rarely closes a round: a pitch deck can help get meetings, but no one invests from seeing only the deck, as one founder who raised with several decks put it. So audit the thinking behind the slides, then audit the slides.
The pitch deck audit checklist
The list below follows the areas that TechCrunch’s review of where founders go wrong with decks covers: the problem, solution, customer, market, team, traction, business model, go-to-market, competition, financials, the ask and the overall narrative. Score each item pass, weak or fail. Be harsh. An investor will be.
Problem and customer
- Can you name the specific buyer, not “small businesses” or “enterprises”?
- Is there evidence the problem is painful enough that people already spend money or time on it?
- Does the problem slide say why it matters now? TechCrunch also recommends checking for a clear “why now.”
Solution and product
- Does the solution follow directly from the problem, in plain words?
- Could someone outside your industry say what the product does after one slide?
- Is there proof that it works: a demo, a pilot, a number from real users?
Market size
- Is the market built bottom-up (customers times price) or only pasted from a report?
- Do you state the share you can realistically obtain, and why?
- Are the assumptions visible so an investor can challenge them instead of dismissing the whole number?
Team and founder-market fit
- Does each bio answer “why you, for this problem”?
- Are the gaps in the team acknowledged, with a plan to fill them?
Traction and metric quality
This is where decks quietly lose credibility. The checklist from VC Advisory is a useful standard: each core metric should come with its definition, period, population and source. Keep actual, contracted, pipeline and forecast revenue separate, and state retention cohorts with their denominators.
In practice, ask yourself:
- Does “users” mean sign-ups, active users or paying customers?
- Is the growth rate measured over a period long enough to mean something?
- Is any pipeline revenue sitting in the same chart as booked revenue?
- If retention is shown, can you say which customers it covers?
Business model and unit economics
- Is pricing stated, with a reason it is set there?
- Are CAC, payback and gross margin shown, or at least estimated with a clear method?
- Do you know which assumption, if wrong, breaks the model?
Go-to-market and competition
- Is there a repeatable channel, or a list of channels you might try?
- Does the competition slide include the real alternative, which is often a spreadsheet, an agency or doing nothing?
- Is the moat something that gets stronger with time, not just a feature list?
Financial logic
Investors do not expect your five-year forecast to be right. They expect it to be reasoned. Show the key drivers and a range instead of false precision, and tie spending to measurable milestones. If one slide says you will hit a number and no other slide explains how, that is a hole.
The ask and use of funds
- Is the amount appropriate for your stage?
- Does the money buy a specific milestone, such as a revenue level or a product launch, that justifies the next round?
- Does the runway math hold up?
Narrative flow
- Does each slide’s headline carry the point, so the deck stands alone without you narrating?
- Does the order build an argument, from problem to proof to ask?
Run the three-minute stranger test
Before anything else, hand the deck to someone who knows nothing about your company and give them three minutes. A good bar, suggested in a guide on auditing a deck after investor rejection, is that a stranger should be able to summarize the business after skimming it for that long.
Then ask them to write down, without help:
- Who the customer is.
- What you sell and how you make money.
- Why it is a big opportunity.
- How much you are raising and what it buys.
If any of the four is wrong or blank, you have found your first fix. This test costs nothing, which is why it belongs before any paid review.
How to critique your own deck for weak assumptions
The hardest part of self-audit is that you know too much. You fill in gaps without noticing. Three habits help.
Underline every claim, then ask “says who?” Each statement about the market, customers or growth should have a source, a calculation or an honest label such as “our estimate.” Anything left with no answer is a weak assumption.
Write the investor’s rebuttal. For each slide, write the single sharpest objection in one sentence. “Your TAM assumes every clinic buys.” “Your retention is two months old.” If you cannot write a rebuttal, you are not being critical enough.
Rank assumptions by damage. Not all weak spots are equal. Ask what happens to the business if each one is wrong by half. The ones that break the model go to the top of your fix list, and those are the ones to rehearse hardest.
How to prioritize improvements
A long list of fixes is useless if you do them in the wrong order. Use this sequence:
| Priority | What to fix | Why it comes first |
|---|---|---|
| 1 | Customer, problem and proof of demand | If these are unclear, nothing downstream matters |
| 2 | Traction and metric definitions | Ambiguous numbers erode trust in every other number |
| 3 | Business model and unit economics | This is where diligence questions land |
| 4 | The ask and use of funds | Ties the round to a milestone |
| 5 | Narrative order and headlines | Makes the deck work without you in the room |
| 6 | Design and polish | Helps, but cannot rescue a missing buyer or missing evidence |
The same guide makes the point that audit criteria should favor storytelling and clarity over design, because a beautiful slide with no clear buyer or proof of demand will still fail in diligence. Design belongs on the list, just at the bottom.
What to look for in a useful audit or critique
Whether you do it alone, trade decks with a founder friend or use a service, judge the output by these standards:
- It is adversarial. Feedback that mostly praises is comforting and useless. You want the objections.
- It is specific. “Strengthen your traction slide” is not feedback. “Your 40% growth figure covers six weeks and one customer” is.
- It is ranked. You should know which three things to fix first.
- It covers the substance. Assumptions, evidence and the ask, not only layout.
- It prepares you to speak. The deck is read once; the questions come live.
How much does a pitch deck review cost?
Prices vary widely by who does the work. A review from a former investor or fundraising advisor is a custom service, usually quoted per project, and it can run from a short feedback session to a multi-week engagement. Free options include feedback from founder peers and mentors, which is good for the stranger test but rarely adversarial. Software sits at the low end of cost and gives you something you can repeat after every revision.
For reference, here is what we charge, as published on our site:
| Option | Price | Best for |
|---|---|---|
| Tactical Audit | $25 one-off per audit | A single deck or memo before one meeting |
| Founder Pro | $49 per month | Founders iterating through a full raise |
| Venture & Studio 5 Seats | $149 per month | Studios and teams with several founders |
Because an audit is a loop, not a one-time event, the cost question is really about how many rounds of revision you expect before you pitch.
Prepare for the questions, not just the slides
A polished deck that you cannot defend is worse than a plain one you can. Investors move quickly from the slide to the follow-up, and the follow-ups cluster in predictable places.
Business model. “What happens to margin if your largest customer leaves?” Know your cost structure and your concentration.
Financial logic. “What has to be true for this forecast?” Name the two or three drivers and your range, not a single point.
Traction. “How is this metric defined, and over what period?” Have the definition ready before you are asked.
The ask. “What will you have achieved when this money runs out?” Answer with a milestone and a date range.
Practice answering each in under 60 seconds, out loud, with someone interrupting you. Writing answers does not train the skill of defending them under pressure.
Perspective
Most feedback founders get is polite. A friend, a mentor or a general-purpose AI tool tends to find what is good about your deck and gently suggest a few tweaks. A partner in a room does the opposite: they hunt for the one assumption that makes the whole case fall over.
We built Dialectic around that gap. Our approach is adversarial on purpose: partner-style questioning, scoring for how fragile your claims are, and a structured list of blind spots with the tactical pivots that turn each weakness into something you can defend. We think a deck is only ready once it has survived a hostile read.
We also think the audit should end in speech, not in a document. That is why we pair the written teardown with a real-time boardroom rehearsal where your defense is graded. Our aim is for the hardest question in your meeting to be one you have already answered.
D
Put the audit to work: Dialectic
If you want the adversarial read without hunting for a former investor, upload your deck or memo to Dialectic. You get scrutiny in the style of a seasoned VC partner, a fragility and risk score, a structured list of weaknesses and blind spots, and actionable feedback on what to change.
The Tactical Audit is $25 per audit, which suits a single deck before a single meeting. If you are iterating across a raise, Founder Pro at $49 per month lets you rework and re-test, and Venture & Studio 5 Seats at $149 per month covers a team. To see what the output looks like on a real document, read the Nexus take-rate teardown, where a forensic CFO persona red-teams a board memo.
A simple audit routine for the week before your meeting
- Day 1: Run the three-minute stranger test with two people. Note every misunderstanding.
- Day 2: Go through the checklist above and mark each item pass, weak or fail.
- Day 3: Get an adversarial read, from a skeptical advisor or from a tool built for it. Collect the objections.
- Day 4: Fix the top three problems by the priority table. Redefine any ambiguous metric.
- Day 5: Rehearse the ten hardest questions out loud, timed.
- Day 6: Run the deck past the stranger once more. Confirm the summary is now correct.
- Day 7: Rest. Re-read only the ask and the use of funds.
This routine keeps you from the common trap of polishing slides at the last minute while the real exposure, an unsupported number, goes untouched.
FAQ
What is the difference between a pitch deck audit and a pitch deck review?
The words overlap, and most providers use them loosely. In practice, a review often means general feedback on flow and design, while an audit implies a systematic check of claims, evidence and likely objections against a checklist. Whichever label is used, ask what you will receive: a ranked list of fixes is the signal that it is the useful kind.
How long should a pitch deck audit take?
A self-audit with the checklist takes a couple of hours, plus time to fix what you find. A tool-based audit returns results in a single session. Human reviewers usually need a few days to schedule and return feedback, so build that into your timeline before your meeting.
Should I audit the deck or the business first?
The business, then the deck. Slides are only a summary of your reasoning, and weak reasoning shows through good design. If the audit exposes a gap such as unclear unit economics, fix the underlying numbers before rewriting the slide.
Can I use a general AI chatbot to audit my deck?
You can, but general tools tend to default to supportive feedback unless you prompt them hard to be critical. A purpose-built tool scores fragility and asks partner-style questions by design. Whichever you use, treat the output as a list of questions to answer, not a verdict.
Leave a Reply