--- title: "Talk like a CEO: 3 tools, 5 examples" date: 2026-10-04 updated: 2026-10-04T00:00:00Z url: https://www.evallens.io/blog/how-to-talk-like-a-ceo author: "Yaroslav Volovyi" column: "Anonymous Unicorn" reviewer: "Vladislav Starodubov" --- # Talk like a CEO: 3 tools, 5 examples Picture minute four of a seed pitch. The founder is still on the history of the industry: how it started, what went wrong in 2019, why the incumbents are slow. Every sentence is true. Then the investor leans in: "Sorry. What do you actually do?" The founder said nothing false. They said all of it in the wrong order. **To talk like a CEO, fix the structure of what you say before you work on your voice.** Dig to the root of the problem with the 5 Whys. Sort your reasons so none overlap and none are missing (MECE). Then say the answer first and put the proof under it (the Pyramid Principle). Most advice on how to speak like a CEO covers executive presence: your voice, your posture, your filler words. Those matter a little. The architecture of what you say matters far more: what you put first, what you group together, and how deep you dug before you opened your mouth. The CEOs you find convincing organized their thinking before the meeting, and the confidence you hear comes from that. ## The short version Three tools from the consulting and manufacturing world do most of the work. Each one has a job. ```text DIG 5 Whys → depth what is true at the root SORT MECE → completeness nothing doubled, nothing missing LEAD Pyramid → delivery answer first, then reasons, then proof ``` You run them in that order when you prepare. You speak them in reverse. > Think from the bottom. Speak from the top. ![Think from the bottom, speak from the top: prepare with 5 Whys, MECE and the Pyramid, then speak answer first](https://wlnkmhxeuwvnyojiksfu.supabase.co/storage/v1/object/public/media/photos/talk-like-a-ceo/think-vs-speak.e2ca0021.webp) ## Speak in the order your listener needs We think in one direction: we gather facts, notice a pattern, form an argument, and reach a conclusion. Most founders speak in that same order, because it's how they discovered the idea. They walk the listener through their whole journey and arrive at the point at the end, like a mystery novel. Listeners need the reverse. They want the conclusion first, so they know where they're going, and then the reasons, so they can check the road. Give them the journey first and they spend the whole time guessing the destination. Paul Graham made a sharp version of this point years ago: good ideas are most of writing well, but an alarmingly small part of speaking well, because a person in an audience thinks less carefully than the same person reading ([Paul Graham](https://paulgraham.com/speak.html)). A reader can flip back a page. A listener can't. Structure is what carries your idea across that gap. And the listener is short on time. When Harvard's Michael Porter and Nitin Nohria tracked how 27 CEOs spent roughly 60,000 hours, the takeaway was that attention is the scarcest thing an executive has ([HBR](https://hbr.org/2018/07/how-ceos-manage-time)). Investors behave the same way: the average pitch deck gets a few minutes, not a few hours ([DocSend](https://www.dropbox.com/resources/docsend-pitch-deck-research)). Nobody has time to wait for your point. To talk like a CEO, do the hard thinking before the meeting, so your listener doesn't have to do it during. ## Step 1: Dig. Ask why until the answer stops changing. **The 5 Whys is a root-cause method: you ask "why" about a problem, then ask "why" about the answer, and keep going until you reach a cause you can fix.** It comes from Toyota. Sakichi Toyoda used it in the 1930s, and Taiichi Ohno later built it into the Toyota Production System, calling it the basis of Toyota's scientific approach ([Wikipedia](https://en.wikipedia.org/wiki/Five_whys)). Ohno's classic example is a machine that stopped. Why? A fuse blew from overload. Why? The bearing wasn't lubricated enough. Why? The oil pump wasn't pumping enough. Why? The pump shaft was worn. Why? There was no strainer, so metal scraps got in ([5xWhys](https://5xwhys.com/articles/5-whys/history-toyota/)). Replace the fuse and the machine stops again next week. Add a strainer and it doesn't. Most founders stop at the fuse. They pitch the first answer, which is a feature, when the fifth answer is the reason anyone pays. Run the same drill on a pitch claim (an illustrative example, not a real company): ```text Claim: "We automate payment reconciliation." Why 1 → Finance closes the books late every month. Why 2 → Two people spend the last week matching payments by hand. Why 3 → Bank exports and the ERP use different reference formats. Why 4 → Volume doubled this year. The finance team didn't. Why 5 → A late close delays the board pack, and the CFO owns the board pack. Root: "We get the CFO's board pack out on time." ``` "We automate reconciliation" is what you do. "We get the board pack out on time" is why the CFO signs. The second one is your headline. ![5 Whys example: from the claim 'we automate payment reconciliation' down to the root 'we get the CFO's board pack out on time'](https://wlnkmhxeuwvnyojiksfu.supabase.co/storage/v1/object/public/media/photos/talk-like-a-ceo/five-whys.c97b5896.webp) ### Use it in customer discovery, too The same drill works on customers, with one rule borrowed from Rob Fitzpatrick's *The Mom Test*: ask about specific things they did in the past, not opinions about a hypothetical future, and talk about their life instead of your idea ([Unusual Ventures](https://www.unusual.vc/rob-fitzpatricks-mom-test/)). "Why did you rebuild that spreadsheet last month?" gets you a root. "Would you use a tool that fixes this?" gets you a polite yes. ### When the 5 Whys breaks Two caveats, so you don't over-trust it: - **Real problems often have more than one cause.** The method assumes a single chain. Former Toyota executive Teruyuki Minoura called it "too basic" for that reason, and different people asking the same whys can land on different roots ([Wikipedia](https://en.wikipedia.org/wiki/Five_whys)). If a why has two honest answers, follow both. - **Five is not the number.** Stop when another "why" stops producing new information. Sometimes you get there in three. Check the root against something you observed, not something you deduced. ## Step 2: Sort. Make your reasons MECE. **MECE stands for "mutually exclusive, collectively exhaustive": the parts of your answer don't overlap with each other, and together they cover the whole question.** It was developed by Barbara Minto at McKinsey and sits underneath her Pyramid Principle ([Wikipedia](https://en.wikipedia.org/wiki/MECE_principle)). It sounds like consultant jargon. In practice you follow two plain rules: you never say a point twice, and you never forget one. To your listener, that discipline reads as clarity. Most investors have seen this "why us" slide many times: ```text Before • Great team • Strong technology • AI-powered • Experienced founders • Fast execution ``` "Great team" and "experienced founders" are the same point. "Strong technology" and "AI-powered" are the same point. And nothing on the list explains how you'll find customers, and that is often the question the investor came in with. Five bullets, three ideas, one hole. ```text After • Why this problem one named buyer, a pain they pay to fix today • Why our approach the one mechanism a rival can't copy quickly • Why we win it our edge in reaching that buyer first ``` Three buckets, no overlap, and together they answer "why should this company win?" with nothing left hanging. ![MECE before and after: overlapping 'why us' bullets versus three buckets with no overlaps and no gaps](https://wlnkmhxeuwvnyojiksfu.supabase.co/storage/v1/object/public/media/photos/talk-like-a-ceo/mece-sort.c1a1bc55.webp) ### Two MECE maps every founder should keep **Your deck.** Investors already carry a mental checklist. Sequoia's well-known business-plan outline is a good stand-in: purpose, problem, solution, why now, market, competition, business model, team, financials, vision ([Sequoia, via mirror](https://dantes.io/resource/writing-a-business-plan-sequoia-capital-6a7120eece2ccb511596ab82)). If one of those is missing, the listener notices the gap before they notice your strengths. **Your objections.** Sort every pushback you've heard into a few buckets that don't overlap, for example: price, trust, fit and timing. Each objection should land in exactly one bucket. When a new one fits nowhere, your map has a hole. Better to find it before the next call than during it. ### When MECE breaks MECE is a check, not a template. Don't bend reality to get three tidy boxes. If the truth has two reasons, say two. Symmetry that hides the real reason is worse than an uneven list that names it. ## Step 3: Lead. Say the answer first. **The Pyramid Principle says you persuade best when you state your conclusion first, then group two or three supporting arguments under it, then put the evidence under each argument.** Barbara Minto, the first woman with an MBA that McKinsey hired, developed it in the 1960s while editing the firm's reports, and published it as *The Pyramid Principle* in 1985 ([Wikipedia](https://en.wikipedia.org/wiki/Barbara_Minto)). She wasn't alone in reaching it. The US military teaches the same idea as BLUF, "bottom line up front" ([Wikipedia](https://en.wikipedia.org/wiki/BLUF_(communication))). In 2004, Jeff Bezos banned slide decks from Amazon's leadership meetings in favor of written memos, arguing that bullet points let people gloss over weak ideas ([Management Today](https://www.managementtoday.co.uk/why-amazon-banned-powerpoint/leadership-lessons/article/1689543)). Different worlds, same rule: the structure forces the thinking. ```text ANSWER We get the CFO's board pack out on time. / | \ REASON 1 REASON 2 REASON 3 The pain is Our matching We reach CFOs monthly and is built for through their owned by the messy bank auditors CFO formats | | | PROOF PROOF PROOF ``` When you open with the answer, the listener stops guessing and starts checking. Every sentence after it either supports the top or gets cut. ![Pyramid Principle example: the answer on top, three reasons under it, proof under each reason](https://wlnkmhxeuwvnyojiksfu.supabase.co/storage/v1/object/public/media/photos/talk-like-a-ceo/pyramid.0268d4b0.webp) ### The first two sentences Y Combinator's Michael Seibel gives founders the same advice: when you pitch, start by saying what your company does in about two plain sentences. Skip the long setup about the problem ([Y Combinator](https://www.ycombinator.com/library/4b-how-to-pitch-your-company)). If an investor has to ask "so what do you do?", the opening already failed. A simple test: after your first sentence, could the listener repeat what you do to a colleague? After thirty seconds, could they say why they'd believe you? If not, move the answer up. ### Opening a harder conversation For a board update, a pivot or a bad-news call, Minto's opener is useful: **Situation, Complication, Question, Answer (SCQA)**. Say what's stable, what changed, the question that creates, and your answer, in four short sentences. Then go straight into the pyramid. ```text Situation We grew to [40] customers on the self-serve plan. Complication Churn doubled after we raised prices in [month]. Question Do we roll the price back? Answer No. We add a cheaper starter tier and keep the core price. ``` ## One workflow, three rooms The same three steps work whether you're talking to an investor, a customer or your own team. What changes is the answer at the top and the buckets underneath. | Room | Lead with (the answer) | Sort into (MECE buckets) | Dig for (the root) | |---|---|---|---| | Investor | What you do, and the one number that proves it's working | Problem, solution, market, team, traction, why now | Why this matters now, not five years ago | | Customer | The outcome they get, in their words | Their alternatives, your difference, proof, price | The pain they already pay to fix | | Team | The decision, and what it means for them this week | Options you considered, why this one, what changes | Why we're doing this now | **With customers, your pitch has to survive being retold.** Gartner reported in March 2026 that 67% of B2B buyers now prefer a buying experience without a sales rep at all ([Gartner](https://www.gartner.com/en/newsroom/press-releases/2026-03-09-gartner-sales-survey-finds-67-percent-of-b2b-buyers-prefer-a-rep-free-experience)). Your champion will repeat your pitch in a room you're not invited to. A pyramid can be retold in one breath. A journey can't. April Dunford's *Sales Pitch* adds a useful warning here: complex deals are lost to the buyer's indecision more often than to a competitor ([Shortform](https://www.shortform.com/summary/sales-pitch-summary-april-dunford)). A clear answer at the top is how you fight "let's revisit next quarter." If selling feels unnatural to you, take heart from Pete Kazanjy. As he puts it in *Founding Sales*, sales skill isn't inborn, it's learned ([Founding Sales](https://www.foundingsales.com/introduction)). The three steps above are the learnable part. ![Three rooms, one method: the same structure for investors, customers and your team](https://wlnkmhxeuwvnyojiksfu.supabase.co/storage/v1/object/public/media/photos/talk-like-a-ceo/three-rooms.a05990c9.webp) ## Five rewrites: before and after Below, the same founder speaks before and after the three steps. (All examples are illustrative. Brackets mark the places for your real numbers.) **1. The opening** > Before: "So, back in 2023 my co-founder and I were working at a logistics company, and we kept noticing that..." > > After: "We help mid-size finance teams close the month in [3] days instead of [8]. [12] companies pay for it today." **2. The product** > Before: "We're an AI-powered platform that leverages machine learning to streamline financial operations." > > After: "We read your bank exports and your ERP, match every payment automatically, and flag the few we can't for a human." **3. "How big is the market?"** > Before: "It's a [$40B] market, and if we get just 1%..." > > After: "Our first buyer is [5,000] mid-size companies in [region] with this exact ERP. At [$X] a year, that's [$Y] before we expand." **4. "This is expensive."** > Before: "Well, we have a lot of features, and our support is great, and we can maybe discuss a discount..." > > After: "Is it the price itself, or that you're not sure it pays back? If it's payback, one late close costs you [N] hours of two people's time. We pay for ourselves in [month]." The fourth one is MECE at work: before answering, you find which bucket the objection is in. **5. The ask** > Before: "So yeah, we'd love to hear your thoughts and stay in touch." > > After: "We're raising [$X] to reach [milestone] by [date]. The next step is a call with our first customer. Want the intro?" ## Before your next meeting: a ten-minute checklist 1. **Write your answer in one sentence.** If it needs a comma and an "and," it's two answers. Pick one. 2. **Run five whys on that sentence.** Stop when the answer stops changing. Use the root as your headline. 3. **List your reasons, then cut the overlaps.** Two or three that don't repeat each other beat five that do. 4. **Check for the missing bucket.** For investors: problem, solution, market, team, traction, why now. For customers: alternatives, difference, proof, price. 5. **Put one piece of proof under each reason.** A number, a name or a quote. An adjective doesn't count. 6. **Sort your likely objections into buckets.** Prepare one answer per bucket, not per objection. 7. **Say it out loud in thirty seconds.** If the listener couldn't repeat it, move the answer higher. 8. **Write down your ask.** Exact amount, exact next step, exact date. Entering a pitch competition? Pair this with our [pitch competition prep checklist](https://www.evallens.io/blog/how-to-prepare-for-our-pitch-competition), which covers what to put on each slide. ## The stranger test Your first listener often never met you. Someone, or something, reads your deck cold, in a few minutes, with no chance to ask what you meant. One VC describing AI-first screening wrote: "The deck is a form we extract, not a narrative to interpret" ([Venture Curator](https://www.venturecurator.com/p/vc-ai-pitch-deck-screening)). EvalLens shows you that cold read before it counts. A panel of six AI reviewers analyzes a deck ([how EvalLens evaluates pitch decks](https://www.evallens.io/blog/how-evallens-evaluates-pitch-decks)), and one of them looks only at pitch quality: structure, clarity of delivery and narrative. The deck is read as ten sections (problem, solution, market, business model, traction, team, roadmap, financials, ask, and anything else), and a section that's missing or thin gets flagged, which is MECE applied to your slides. The [evidence-based report](https://www.evallens.io/product/evidence-based-reports) then lists the questions a reviewer should ask you live, sorted into critical, important and optional, which is your objection map, written for you. AI prepares the analysis. A human makes the final call, every time. EvalLens has run more than 1,000 evaluations, and the rubric rewards the same habit in every one of them: the answer on the first slide and the proof right under it. ![How a stranger reads your deck: ten sections, six AI reviewers, live questions, and a human makes the final call](https://wlnkmhxeuwvnyojiksfu.supabase.co/storage/v1/object/public/media/photos/talk-like-a-ceo/stranger-test.dc9e8647.webp) Your listener will forget how confident you sounded. They will remember whether they could repeat what you said to their partner on Monday. Want to see how your own deck reads to a stranger? **Try the live demo.** Running a program or a fund and want this read on every deck? **Book a demo.** ## Common questions **What does it mean to talk like a CEO?** It means leading with your conclusion, backing it with two or three reasons that don't overlap, and doing the root-cause thinking before the meeting. Your structure carries most of the weight. Voice and confidence help, but a listener remembers what they can repeat, and they can only repeat what you put in order. **What is the Pyramid Principle in a pitch?** The Pyramid Principle is Barbara Minto's rule: put the answer first, then two or three supporting arguments, then the evidence under each one. In a pitch, your first sentence says what you do and why it matters. Every slide after it proves that sentence or gets cut. **What is an example of the 5 Whys in a pitch?** Start from your claim, "We automate payment reconciliation," and ask why the customer cares. Books close late. Matching is manual. Bank and ERP formats differ. Volume doubled. A late close delays the CFO's board pack. The root, "we get the board pack out on time," becomes your headline. **How do I know if my pitch is MECE?** Run two checks. Do any two points say the same thing in different words? Merge them. Is there an obvious question your points leave open, like how you'll reach customers? Add a bucket for it. When nothing overlaps and nothing is missing, your structure is MECE. **What is BLUF, and should founders use it?** BLUF means "bottom line up front," the US military's version of answer-first writing. Founders should use it in investor updates, emails and the first minute of a pitch: state the result or the ask in the first sentence, then give the detail. It's the Pyramid Principle in its shortest form. **How many "whys" should I ask?** Ask until the answer stops changing. Five comes from Toyota as a rule of thumb, and you'll often get there in three or four. If one "why" has two honest answers, follow both chains, and check the root against something you observed rather than something you assumed. **Should I pitch investors and customers differently?** Use the same method with a different answer on top. Investors need to hear why this business is big and already working. Customers need to hear the outcome they get, in their own words. The digging and the sorting stay identical, so you prepare once and adapt the headline. **How do I stop rambling when an investor interrupts?** Answer the question in one sentence first, then offer the detail: "Short answer: yes, [12] paying customers. Want the breakdown?" Let them pull more instead of you pushing it. An interruption often means the listener lost the thread, so give them the top of the pyramid again.