# Deck arcs: frameworks in their authors' words, and seven real decks How a deck holds together from the first slide to the last: what the storytelling frameworks actually say, seven real decks that pair a large vision with a small entry point, and how titles, transitions, the opening and the close carry the thread. Use it after picking a backbone in [storytelling-frameworks.md](storytelling-frameworks.md), when you write the sequence of slide titles. Quotes stay under 25 words; "..." marks a cut. Where the original is gone or hard to reach, the link goes to its Internet Archive copy, marked "archived". [inference] marks our reading, not the author's. ## 1. The frameworks, in the authors' words - **Andy Raskin, strategic narrative** ([archived](https://web.archive.org/web/2024/https://medium.com/the-mission/the-greatest-sales-deck-ive-ever-seen-4f4ef3391ba0), 2016; written about a sales deck). Five elements "in precisely this order": name a big, relevant change in the world; show there will be winners and losers; tease the promised land, "a new future state, not your product or service"; features as magic gifts against the obstacles; evidence you can make the story come true. Against opening on the problem: "you put them on the defensive". Without customers, "Product demos are the next most effective evidence". - **Nancy Duarte, what is and what could be** ([TED transcript](https://www.ted.com/talks/nancy_duarte_the_secret_structure_of_great_talks/transcript), 2011; [HBR](https://hbr.org/2012/10/structure-your-presentation-li), 2012). Establish "what is", set it against "what could be", and make the gap as big as possible; the middle travels back and forth between the two; the end is a call to action and the world as "a new bliss". The audience is the hero: "You're not Luke Skywalker, you're Yoda." For decks read without a presenter: [Slidedocs](https://www.duarte.com/resources/books/slidedocs/). - **Sequoia** ([Writing a Business Plan](https://www.sequoiacap.com/article/writing-a-business-plan/)): "define your company in a single declarative sentence", then problem, solution ("Explain your eureka moment"), why now, market, competition, business model, team, financials, and the vision: "If all goes well, what will you have built in five years?" - **Aaron Harris, YC** ([seed deck template](https://www.ycombinator.com/library/2u-how-to-build-your-seed-round-pitch-deck); [podcast](https://www.ycombinator.com/blog/aaron-harris-on-fundraising-and-meeting-with-investors), 2019; [blog, archived](https://web.archive.org/web/2020/http://www.aaronkharris.com/advice-on-pitching), 2014). Start from where you are now: jumping straight to the giant vision is "a huge mistake"; "There has to be a continuous line from one thing to another"; "make it really big but believable". If you make a large transition, "be very clear about it and explain why". Slides "should be reentrant", so a listener who looks up from a phone can "pick up the thread". - **Geoff Ralston, YC demo day** ([guide](https://www.ycombinator.com/blog/guide-to-demo-day-pitches/), 2016): "begin your story as narrowly as you can and then expand"; "Don't bury the lead" (Justin Kan, as Ralston quotes him); listeners keep "3 or 4 key points", so close by listing them; build the story first and the slides after, or the talk "will stutter from slide to slide". - **Kevin Hale, YC** ([How to Pitch Your Startup](https://www.ycombinator.com/library/6q-how-to-pitch-your-startup), 2019): do not start with a technology, start with a problem; you need a storytelling version for non-experts. - **Michael Seibel, YC** ([How to Pitch Your Company](https://www.michaelseibel.com/blog/how-to-pitch-your-company)): "Start with the name of your company and what it does"; "There's no need to set up the problem"; the insight gives "more of an aha moment" than the description; "There's no need to dance around the ask." - **Paul Graham** ([How to Present to Investors](https://paulgraham.com/investors.html), 2006): "Say what you're doing as soon as possible, preferably in the first sentence"; the problem gets no more than "a tenth of your time". - **Jared Friedman, YC, hard tech** ([video](https://www.youtube.com/watch?v=C1DlZWfI6rk), 2019): the ask as a step in a staircase, where no step should be too large. - **Reid Hoffman, LinkedIn's Series B** ([archived](https://web.archive.org/web/2024/https://www.reidhoffman.org/linkedin-pitch-to-greylock/), 2013, on a 2004 deck): open with what investors "must believe in order to want to be shareholders"; "Steer into your investors' objections"; the wedge, "start by transforming one business"; and "If an investor sequenced through the titles, they'd be able to get a sense of the flow of the argument." - **Barbara Minto, the pyramid** ([site](https://www.barbaraminto.com/)): situation, complication and question identify the reader's question; each point summarizes the group beneath it. The book was not opened. - **Story Spine (Kenn Adams) and Pixar's rules (Emma Coats)** ([Adams](http://www.aerogrammestudio.com/2013/06/05/back-to-the-story-spine/), 2013; [Coats, archived](https://web.archive.org/web/2015/http://storyshots.tumblr.com/post/25032057278/22-storybasics-ive-picked-up-in-my-time-at-pixar), 2012): "Once upon a time... Every day... But, one day... Because of that... Until finally... And, ever since then". Coats, rule 7: "Come up with your ending before you figure out your middle." Her rule 4 restates Adams's spine without credit. - **Trey Parker and Matt Stone** (mtvU's *Stand In*, 2011; [unofficial upload, auto-captions](https://www.youtube.com/watch?v=vGUNqq3jVLg), no official one found): between beats the word should be "therefore or but", never "and then", because "that gives you the causation between each beat". - **Mike Maples Jr., Floodgate** ([Book Bite](https://nextbigideaclub.com/magazine/5-qualities-hugely-successful-start-ups-according-venture-capitalist-bookbite/50610/), 2024): "An inflection is a specific and powerful new change event"; a breakthrough needs "a radical and unconventional insight about the future". [inference] Opening = the inflection, middle = the insight, end = the future pulled into the present. - **James Currier, NFX** ([23 rules](https://www.nfx.com/post/23-rules-storytelling-fundraising), 2021): "Give Your Best Points Right Away"; "Don't Always Start Your Narrative With 'The Problem'"; four templates (industry story, customer story, hero founder, company journey to insights); slide titles "are the core structure of your story"; the listener must be able to retell it, and "typically, they have about 30 seconds". ## 2. Seven real decks: a big vision, a small entry | Deck | Order | The bridge from wedge to vision | The close | |---|---|---|---| | Tesla, "Secret Master Plan", 2006, an essay ([archived](https://web.archive.org/web/2016/https://www.tesla.com/blog/secret-tesla-motors-master-plan-just-between-you-and-me)) | vision first; raises its own objection ("Are we really in need of another high performance sports car?") and answers it | "enter at the high end of the market ... and then drive down market as fast as possible" | the plan restated in five lines | | LinkedIn Series B, 2004 ([archived](https://web.archive.org/web/2024/https://www.reidhoffman.org/linkedin-pitch-to-greylock/)) | vision first; opens on revenue, the objection of the day | "a great recruiting business with an option for more" (Greylock's thesis, as Hoffman tells it) | a three-point summary, then slide 1 repeated | | UberCab, Dec 2008, pre-launch ([archived PDF](https://web.archive.org/web/2020/http://garrettcamp.com/files/UberCab_Dec2008.pdf)) | wedge first ("Focus on SF/NYC to begin") | "Extend infrastructure to other LBS applications"; worst case stated: "Remains a 10 car, 100 client service in SF" | progress to date and next steps | | Coinbase seed, Aug 2012 ([PDF, third-party host](https://media.genppt.com/pitch-decks/coinbase/coinbase-pitch-deck-2012.pdf)) | wedge first | "Coinbase: A Hosted Bitcoin Wallet", then traction, then "Coinbase: A New Payment Network" | slide 1's title repeated: it closes on the wedge | | Joby Aviation, SPAC, Feb 2021 ([SEC](https://www.sec.gov/Archives/edgar/data/1819848/000119312521053631/d135823dex992.htm)) | wedge first, mission last | "Pre-service revenues from government contracts de-risk commercialization" | the mission on the last slide | | Aurora, SPAC, Jul 2021 ([SEC](https://www.sec.gov/Archives/edgar/data/1828108/000119312521215991/d326572dex992.htm)) | vision first | trucking is "the tip of the spear", moving the core platform into adjacent verticals | financials and valuation | | Matterport, SPAC, Feb 2021 ([SEC](https://www.sec.gov/Archives/edgar/data/1819394/000119312521031520/d42860dex992.htm)) | vision first | "Take your building online" (slide 8), then "Tomorrow our data will increase the value of every building" (slide 9) | "Investment highlights" again (slides 4 and 33) | The pattern [inference; seven survivors]: no order dominates, four vision-first and three wedge-first. All seven state an explicit bridge, one causal sentence or a pair of titles from today to tomorrow, and a reason to start small (the unit cost of new technology, a first market, "tip of the spear", certification risk). The three SPAC decks are late-stage, so their transfer to a seed deck is our inference. ## 3. The thread: titles, transitions, opening and close - **Titles that tell the story on their own** (Hoffman, Currier, Minto). LinkedIn's slides 3 to 9 go from people search "1.0" on flat directories to "2.0" on networks, then eBay, PayPal and Google as the "2.0" of their categories, then LinkedIn: the repeated "2.0" is the thread. Label titles ("Early Adopters", "Technology") read in sequence form no sentence [inference]. - **Transitions.** Join blocks with "therefore" or "but" (Parker and Stone); the middle is a chain of "because of that" (Adams). Test [inference]: read the titles aloud joined by "therefore" or "but"; where only "and then" fits, a cause is missing. - **Opening.** A change in the world (Raskin, Currier); something the audience already recognizes, then the turn (Duarte); the thesis or the best point on slide 1 (Hoffman, Ralston, Currier); what the company does in the first sentence (Seibel, Graham; Coinbase's slide 1 is "Your hosted bitcoin wallet"). Techstars puts 75% of preparation into the opening (E23 in [library-investor-advice.md](library-investor-advice.md)); Papermark sees the largest reader drop-off between slides 1 and 2 ([library-investor-studies.md](library-investor-studies.md)). - **A close that returns to the opening.** LinkedIn repeats slide 1 after the summary; Coinbase repeats slide 1's title; Matterport repeats its highlights; Tesla restates the plan in five lines; Adams's "And, ever since then" mirrors "Every day"; Ralston lists the 3 or 4 points to keep. ## 4. Does a narrative itself move investors? No study found measures arc, slide order or a callback in a seed deck. What exists is weak and mixed: - Villanueva 2012, PhD thesis, field experiment with 188 active angels: storytelling's effects "operate in opposite directions, effectively cancelling each other out" ([abstract, archived](https://web.archive.org/web/2016/https://conservancy.umn.edu/handle/11299/144342)). - Hsu, Chan, Yang & Motro 2025, conference abstract (226 angels, then 98 randomized): narrative raised investment "for entrepreneurial angels but not for non-entrepreneurial angels" ([abstract](https://api.openalex.org/works/doi:10.5465/amproc.2025.14743abstract)). - Clarke, Cornelissen & Healey, *Academy of Management Journal*, experiment: the type of language had limited effects, while "using gestures to depict and symbolize business ideas has strong positive effects" ([abstract](https://api.openalex.org/works/doi:10.5465/amj.2016.1008)). - Manning & Bejarano, *Strategic Organization*, Kickstarter, comparative and not causal: successful campaigns use narratives "more coherently" ([abstract](https://api.openalex.org/works/doi:10.1177/1476127016648500)). - Martens, Jennings & Jennings 2007, *Academy of Management Journal*, IPO prospectuses: narrative framing affects capital raised "net of previously emphasized factors" ([abstract](https://api.openalex.org/works/doi:10.5465/amj.2007.27169488)). - Garud, Schildt & Lant 2014, *Organization Science*, theory: expectations set by stories "can also serve as the source of future disappointments" ([abstract](https://api.openalex.org/works/doi:10.1287/orsc.2014.0915)). - Kalvapalle, Phillips & Cornelissen 2024, *Academy of Management Annals*, review of 252 papers: the communicative process of the pitch "remains undertheorized and underexamined" ([abstract](https://api.openalex.org/works/doi:10.5465/annals.2022.0066)). The cleanest test of "tell it as a story" nets to zero; coherence goes with success only in crowdfunding, and without causality. ## 5. Where the sources contradict each other 1. **Open with the problem** (Harris's template, Hale; Graham, capped at a tenth of the time) versus no need (Seibel) versus a change in the world (Raskin, Currier) versus the "what is" the audience recognizes (Duarte). 2. **Tension resolved at the end** (Duarte, Raskin, Adams) versus the answer first (Minto, Hoffman, Seibel, Graham, Ralston, Currier). 3. **Where the vision goes:** first (Hoffman, Raskin), last (Sequoia, Duarte), grown out of a narrow start (Graham, Ralston, Harris), or alternated throughout (Duarte; Liz Wessel, E45 in [library-investor-advice.md](library-investor-advice.md)). 4. **How big:** "Make It Big" (Currier) versus jumping straight to the giant vision is "a huge mistake" (Harris), and stories that set expectations can become the source of disappointment (Garud et al.). 5. **Where the proof goes:** at the end (Raskin) versus at the start (Ralston, Currier, Harris 2014) versus the main objection in the first slides (Hoffman). 6. **The last slide:** a call to action (Duarte), evidence (Raskin), the 3 or 4 points (Ralston), the ask and where the money gets you in a year (Harris's template), the five-year vision (Sequoia), slide 1 repeated (LinkedIn), the ask as a small step (Friedman). 7. **The hero:** the audience, with the founder as guide (Duarte, Raskin), versus founders who come across as formidable (Hale), and "hero founder" as one of Currier's four templates. 8. **Self-standing slides** ("reentrant", Harris 2014) versus one continuous line (Harris 2019, Parker and Stone, Hoffman). 9. **Length:** 20 to 25 slides (Hoffman) versus 10 to 14 plus an appendix (Currier) versus about 10 sets of at most 3 slides (Harris). [SKILL.md](../SKILL.md) sets 10 to 12 for a VC deck, and the workflow wins ([library.md](library.md)). Limits: Minto's book was not opened; Parker and Stone come from an unofficial upload with auto-captions; three of the seven decks are late-stage SPAC decks; the Coinbase deck is a third-party copy.