The pitch deck for investors, slide by slide

A pitch deck does one job: it gives an investor enough clarity and conviction to continue the conversation. It does not close the round, document every claim, or prove that risk has disappeared. When founders ask one slide to do all three, they produce a dense document that nobody can scan and a live pitch that nobody can follow.

The deck should work when you speak over it and when a partner forwards it without you. Y Combinator's seed fundraising guide calls for a coherent leave-behind and says there is no fixed format or order. That is sound advice. Your story may need a different sequence, but the investor's questions do not change. Every slide below earns its place by answering one of them.

Use this as an editing test, not a rigid template. A company with strong usage and little revenue should foreground usage. A biotech company will need a different proof slide from a marketplace. Keep the questions, then choose the strongest evidence your company can honestly provide.

Stage changes the evidence, not the burden of explanation. Before launch, customer research and technical feasibility may carry the case. After launch, investors will expect behavior from the product and sales process to replace some of those early signals. A later seed company cannot keep presenting interview enthusiasm as if no operating data exists.

Build one source deck, then make deliberate versions for the meeting and for forwarding. The live version can rely on your demonstration and spoken transitions. The forwarded version may need a sentence of context on a chart or product screen. Do not maintain two unrelated stories; discrepancies between numbers and claims create avoidable doubt.

Slide 1 answers "What does this company do?"

The first slide should let an investor repeat what you do, for whom, and why the customer cares after one reading. Put the company name, a plain one-sentence description, and one piece of context that prevents a predictable misunderstanding. That context might be "sold to hospital finance teams" or "software, not a marketplace." Use a product image only if it makes the sentence easier to understand.

Sequoia's guide to writing a business plan tells founders to define the company in one declarative sentence and warns against listing features instead of communicating the mission. I agree with the sentence test, but "mission" often tempts founders into fog. "We make commerce intelligent" says almost nothing. "We reconcile marketplace payouts for finance teams" gives the investor a customer, a job, and a product category.

A useful sentence pattern is:

[Company] helps [specific customer] accomplish [important job] by [distinct mechanism].

Do not force every word into that syntax. Read the sentence aloud, remove the adjectives, and ask someone outside your company what they think you sell. If their answer differs from yours, fix the sentence rather than adding a paragraph.

The opening slide is not the place for a full vision, a founder biography, three logos, a financing history, and a slogan. One precise claim creates room for the investor to listen. If you have a striking traction fact that changes how the rest of the deck will be read, add it as a small proof point. Otherwise save evidence for its own slide.

Slide 2 answers "Whose problem is painful enough to buy a solution?"

The problem slide should identify a specific customer, the moment the problem occurs, what that customer does now, and the cost of leaving it alone. "Small businesses struggle with operations" is a market category wearing a problem costume. Describe an observable workflow: who opens which file, waits for whose approval, loses which sale, or pays which bill.

Customer quotes can help, but a quote must reveal behavior rather than admiration. "I love the idea" proves nothing. "Our controller spends every Monday matching transfers to orders" identifies a recurring job and a possible buyer. Remove names or identifying details unless you have permission to use them.

Distinguish the user, buyer, and beneficiary when they differ. An employee may use the product, an operations leader may approve it, and finance may control the budget. A deck that calls all three "the customer" hides the sales motion. Investors will find that gap during questions, so resolve it on the slide.

The strongest problem slides also show the current alternative. That alternative may be a competitor, a spreadsheet, an assistant, an internal team, or simply tolerating the pain. Name it without ridicule. If capable customers keep choosing the status quo, there is usually a reason: switching costs, trust, regulation, habit, or a problem that is irritating but not urgent.

Avoid a theatrical story about one invented customer. A composite persona can make a workshop easier to follow, but it weakens an investor deck when it looks like evidence. Use a real workflow and actual observations from customer conversations. Keep the interview count and methodology in your notes or appendix unless those details establish unusual depth.

Slide 3 answers "What have you built, and why is it meaningfully better?"

The solution slide should show the product completing the job introduced on slide 2. Keep the same customer and the same unit of work. If the problem is a controller reconciling payouts, do not jump to a generic analytics dashboard. Show the input, the action, and the result in the product's real sequence.

A short demo often beats a slide in a meeting, but the document must still survive after the demo ends. Use two or three annotated screens, a simple workflow, or a before-and-after example. Each label should explain a consequential action, not point at interface furniture. "Flags orders whose payout does not match" is useful; "smart insights" is not.

Founders regularly confuse a product description with a defensibility claim. "We use AI" describes an implementation choice. It does not explain why customers choose you, why they stay, or why another team cannot reproduce the result. State the advantage at the level the customer experiences: less manual review, access to previously unavailable supply, a faster approval path, a lower failure rate, or a workflow that fits an existing constraint.

Walk one record through slides 2 and 3 as an editing exercise:

1. Name the person and the triggering event.
2. Show the current action and its measurable cost.
3. Show the product taking the same input.
4. Show the changed result and how you know it changed.
5. Remove any feature that does not affect that record.

This sequence catches the common failure where the problem concerns money but the solution talks about convenience, or the problem concerns speed but the proof measures clicks. The claim and the evidence must use the same unit.

Slide 4 answers "Why can this company win now?"

The timing slide should name a change that makes the product possible, necessary, or economical now, then connect that change to customer behavior. A rising category label is not an answer. Neither is a row of news headlines. Explain the causal chain.

Sequoia's pitch outline asks "Why now?" and frames the issue plainly: if the opportunity is open, why has nobody built the winning company already? Treat that as a challenge, not a ceremonial slide. A credible answer might involve a regulatory change, a distribution channel that recently opened, a cost curve crossing a practical threshold, a new buyer mandate, or an input that has become available.

Write the claim in two sentences. The first says what changed and when. The second says what customers can or must do differently because of it. Then attach evidence: a policy date, a procurement requirement, price history, platform adoption, or behavior in your own pipeline. If you cannot attach evidence, the change may be a hope rather than a timing advantage.

Do not use "people are more comfortable with technology" as a substitute for research. It is broad, hard to test, and rarely specific to one company. Likewise, technical availability alone does not create demand. A new model, sensor, or API matters only if it changes product performance, cost, distribution, or customer willingness to act.

Some companies do not need a separate timing slide. If the shift is inseparable from the problem or market, place it there and use this slot for stronger proof. The question still needs a crisp answer somewhere in the deck.

Slide 5 answers "What evidence says customers want this?"

The traction slide should present the strongest evidence that the company is moving from founder belief toward repeatable customer behavior. Choose metrics that match your stage and business model. Revenue can be decisive, but a pre-revenue company may have stronger evidence in retained use, paid pilots, technical milestones, supply commitments, or a sales pipeline with clearly defined stages.

Start with a chart or small set of numbers that has dates, units, and a readable denominator. "200% growth" invites suspicion if the base was two customers. "42 weekly active teams out of 55 paying teams" tells the reader what happened and among whom. If you use annual recurring revenue, separate contracted, live, and collected revenue. Investors will ask, and mixing them destroys trust.

Traction and projections are different kinds of claims. Traction reports observed behavior. A projection states what you expect under assumptions. Never draw one continuous line that makes forecast months look observed. Use a clear divider, a different visual treatment, and labels. Better yet, keep a speculative five-year curve out of an early deck unless it teaches the investor something about the economics.

Explain inflections instead of smoothing them away. If growth jumped after one large contract, say so. If a cohort is small, show the count. If usage is seasonal, use a comparison that respects the season. Clean numbers are less persuasive than numbers whose limitations you understand.

Pre-launch founders still need evidence. Show what you tested: signed design partnerships, deposits, letters with meaningful terms, completed regulatory work, repeated interview patterns, or a prototype benchmark. A waitlist based on a broad giveaway is weak evidence of willingness to pay. Name it accurately rather than promoting it to "demand."

Slide 6 answers "Can this become a large business?"

The market slide should show how many plausible customers exist, what they could spend, and which portion you can reach first. Build the estimate from customer behavior whenever possible. A percentage of a large research report looks impressive but says little about your route to revenue.

Sequoia's presentation guidance recommends calculating a new market through customer count, how that count changes, and the value of each customer. It also warns against displaying a huge third-party market number without the details behind it. That warning matters because market arithmetic reveals whether the customer definition in slide 2 is real.

Use a checkable model such as:

Initial market = eligible accounts x realistic annual contract value
Reachable revenue = accounts in chosen segment x expected penetration x annual contract value

Put the source and year beside each input. If the annual contract value comes from five pilots, label it a pilot average. If customer count comes from a government dataset, define the industry and size filters. An investor should be able to disagree with an assumption without rebuilding the whole calculation.

Separate total market, reachable market, and the first target segment. These are not three decorative circles. The total market answers how large the opportunity could become. The reachable market reflects your product and sales constraints. The first segment identifies where you can win enough customers to learn and expand.

Do not add unrelated future products merely to inflate the total. A credible expansion path starts with an adjacent buyer, workflow, geography, or pricing unit that follows from what you have already learned. Put detailed segment tables and sensitivity cases in the appendix.

Slide 7 answers "How does usage turn into durable revenue?"

The business model slide should state who pays, what they pay for, how pricing scales, and what it costs to deliver the product. Use current facts first and assumptions second. If pricing is unsettled, say what you charge today, what you are testing, and what evidence will decide the next version.

Show the pricing unit because it shapes both growth and customer incentives. Per seat, per transaction, per location, usage based, subscription, and take rate models behave differently. Explain why your unit grows with customer value rather than creating a reason to limit use. If the buyer signs an annual contract but usage varies monthly, make that relationship clear.

Early decks often display mature margins before the company has mature operations. That is acceptable only when the assumptions are visible. Separate current gross margin from a modeled margin at scale. List the major variable costs, including service work that founders currently perform for free. Hidden human labor does not become software because it sits behind an interface.

For marketplaces, distinguish gross merchandise value from revenue. For financial products, distinguish money moving through the product from fees the company keeps. For pilots, distinguish a one-time implementation payment from recurring product revenue. These distinctions may feel basic, yet blurring them is one of the fastest ways to lose an investor's confidence.

Unit economics belong here when the data is meaningful. A young company with five customers should not pretend it has a stable customer acquisition cost. Show observed sales effort, onboarding work, churn or retention signals, and contribution margin. Put cohort tables and alternative pricing cases in the appendix.

Slide 8 answers "How will you acquire customers repeatedly?"

The distribution slide should identify the first narrow path to customers, the evidence that it works, and the constraint that could stop it from scaling. "Content, partnerships, and sales" is a list of possibilities. A plan names a buyer, an entry point, an owner, a cycle, and a conversion event.

Start with how the first customers actually arrived. Founder relationships are legitimate at the beginning, but classify them honestly. Then explain what was repeatable inside those wins: the same professional community, the same trigger event, a channel partner, an integration, an outbound message, or users inviting colleagues. The repeatable element matters more than the channel label.

Match the motion to contract size and buying friction. A product with a low monthly price cannot normally support months of founder sales and custom onboarding. A regulated enterprise purchase rarely closes through an unattended checkout page. If your current economics and sales motion disagree, show which one you are changing and why.

Include a simple funnel only when every stage has a definition. "Lead" should not mix a newsletter subscriber with a buyer who requested a security review. Show counts for the same period and identify where founder judgment changes the stage. A tiny honest funnel teaches more than a giant pipeline assembled from weak signals.

Distribution advantages can come from access, product behavior, or economics. Give evidence rather than adjectives. If a partner can distribute the product, state the partner's incentive and current commitment. If users invite other users, show the observed invitation and activation pattern. Detailed channel experiments belong in the appendix.

Slide 9 answers "Why will customers choose you over every alternative?"

The competition slide should explain the decision a customer makes, including the choice to do nothing. Name direct competitors, indirect tools, internal work, and the status quo. An empty quadrant does not make your company unique; it usually tells the investor that you searched for axes that isolate your dot.

Compare the few criteria that customers actually use. Those might include implementation time, workflow coverage, required behavior change, accuracy under a specific condition, procurement fit, or total cost. Cite customer interviews, lost deals, win notes, and product tests. Avoid subjective rows such as "easy," "modern," or "best."

Your advantage and your moat are not identical. An advantage explains why you win a deal now. A moat explains why that advantage can persist as competitors react. The moat may come from data rights, network structure, distribution access, regulatory approvals, accumulated workflow depth, switching costs, or cost structure. A feature competitors can copy belongs in the first category, not automatically the second.

The popular advice to omit competitor names because mentioning them gives them credibility is wrong. Investors will search after the meeting, and customers already know their alternatives. Omitting a known competitor makes your research look shallow. Name the company, state where it is strong, and explain the customer or use case where your approach wins.

Do not claim "no competitors." The status quo always competes for budget and attention. If nobody pays to solve the problem today, connect that fact to slides 2 and 4: explain what changed enough to create a purchasing event now.

Slide 10 answers "Why is this team suited to this problem?"

The team slide should connect each founder's relevant evidence to the company's hardest work. Job titles, school names, and employer logos cannot make that connection by themselves. State what each person built, sold, researched, operated, or learned that reduces a specific execution risk.

Keep biographies short. "Led billing operations across eight countries" is stronger than a paragraph of seniority claims if cross-border billing is central to the product. A technical founder might show unusual access to a dataset or years spent with the infrastructure. A commercial founder might show buyer relationships or repeated sales in the category. Use facts you can support.

Address missing capabilities directly when they matter. If the company needs a regulatory leader, manufacturing expertise, or a second technical founder, say what the current team covers and what the next hire must own. Investors do not expect an early team to be complete. They do expect founders to recognize the work.

Advisors belong only when they do specific, ongoing work. A long row of impressive names raises questions about permission and substance. Put an advisor in the main deck if her contribution changes the company's ability to execute, and describe that contribution. Otherwise move advisors to the appendix or omit them.

Team placement can move earlier when founder-market fit is the main source of conviction, especially before product evidence exists. Do not move it forward from vanity. Move it because understanding the founders changes how the investor interprets the opportunity.

Slide 11 answers "What are you raising, and what will it prove?"

The ask slide should state the amount you are raising, the stage or instrument if decided, the operating period the plan covers, and the milestones the financing is designed to reach. Investors should not have to infer that you are fundraising from a deck sent for a fundraising meeting.

Tie spending categories to milestones rather than presenting a pie chart of departments. "Hire four engineers" describes an expense. "Complete the compliance release and move five pilots into paid deployments" describes what the expense is meant to achieve. Include the two or four milestones that would change the company's risk, not a backlog of ordinary tasks.

Be careful with runway precision. A plan built on uncertain hiring dates and revenue assumptions does not become credible because it says "18 months." Show the main assumptions behind the period: starting cash, planned hiring, expected revenue contribution, and a reasonable buffer. Keep the monthly cash model in the appendix or diligence materials.

Do not turn this slide into a negotiation over valuation. If you have set terms and intend to share them, do so clearly in the conversation and appropriate documents. The deck's job is to connect capital to the next body of evidence. Prior funding, current cash, and material commitments may belong here as brief context, with the capitalization table reserved for diligence.

Y Combinator's fundraising guide advises founders to leave a meeting with an attempted close or absolute clarity on next steps. The slide can support that behavior, but it cannot do it for you. End the conversation by asking about the investor's decision process, remaining questions, and a specific next action.

The appendix answers "Can you support the claim without derailing the pitch?"

The appendix should hold material that strengthens a specific claim but is too detailed, conditional, or audience dependent for the main narrative. It is not a storage unit for slides you were afraid to delete. Every appendix slide should answer a question you expect a serious investor to ask.

Good appendix material includes cohort retention, market assumptions and sources, product architecture, security or regulatory detail, sales pipeline definitions, pricing tests, financial scenarios, capitalization history, customer research methods, product roadmap detail, and biographies. The exact set depends on your company. A health company may need clinical and regulatory depth; a marketplace may need separate supply and demand cohorts.

Keep diligence documents out of the deck. Customer contracts, employee agreements, board consents, bank statements, detailed capitalization records, legal correspondence, and personally identifiable data belong in a controlled data room when requested. The deck makes the case. The appendix handles predictable depth. The data room supports verification. Treating those as one document creates needless disclosure and makes the pitch unreadable.

Run a final boundary test on every slide:

• Keep it in the deck if the investor needs it to understand or believe the core case.
• Move it to the appendix if it answers a likely question but interrupts the core case.
• Move it to diligence if it verifies a claim and contains sensitive or exhaustive detail.
• Delete it if it repeats another slide or exists only because you already designed it.

Export the deck to PDF, then read it once without speaking. Check that every chart has units and dates, every acronym is defined, every projection is marked, and the ask is explicit. Send it to a founder who was not in the drafting sessions and ask her to write down what the company does, who buys, why now, the strongest evidence, and what you are raising. If those answers are wrong, another round of visual polish will not save the deck.

Sisters gives women founders a place to get honest feedback on a deck and ask fundraising questions of women who have already done the work. Bring the version that still feels uncomfortable; that is usually where a sharp reader can help most.