Deck teardown · Media kit

Facebook's 2004 media kit, scored: 76/100

The Vintane TeamAll deck teardowns
Teardown · Media kit (2004)TheFacebook's first media kit (Spring 2004) — the ad-sales proposal then-CFO Eduardo Saverin used to pitch advertisers in New York that April, roughly two months after launch, when the company was still called “TheFacebook.”
76/ 100
Fundability score

“The most famous “pitch deck” in the canon was never a pitch deck at all. It's a media kit selling ~$80K display buys against 70,000 college users — and it leads with the best-shown traction in our set, then skips every slide a capital raise lives on: no ask, no team, no dollar-sized market.”

Of every “famous pitch deck” founders are told to study, Facebook's is the one that isn't a pitch deck. The artifact that circulates — saved by a New York advertiser Eduardo Saverin met in April 2004 — is an ad-sales media kit: it exists to sell display inventory, not to raise a round. We still ran it through Vintane's fundability rubric — the same five silent judgments a partner makes while skimming — because founders copy it anyway, and because it's a perfect illustration of a deck that wins on traction and loses on everything a raise is underwritten on.

Executive summary

Facebook's media kit scores 76 because it does the single hardest judgment better than almost any deck in our set, and skips three others entirely. What it does: it shows traction instead of claiming it — 70,000 real users across 20 named colleges, roughly two months after launch, with the campus-penetration story a partner can verify in a sentence. The thesis underneath — a targeted, verified, highly engaged college audience you can reach by profile (by school, by dorm, by interest) — is durable; it's essentially the company Facebook became. But as a template for a founder pitching investors, three slides are missing or soft: there is no capital ask (it asks advertisers for ~$80,000, not investors for a round, and ties neither to any milestone), the audience is described in inconsistent units (70,000 “users” headlined, an ad buy that “reaches thousands,” and a 2005 follow-up proposal that switches to 1.9M “monthly unique visitors”), and the market is asserted as a college count — 20 now, 200 in six months — never built into a dollar figure. Read as what it is, it's a brilliant sales document. Read as the pitch-deck template it's held up to be, it's all proof and no ask.

For the canonical example of a deck that pairs shown traction WITH a clear ask, see our Buffer seed deck teardown — it scores 81 on the same rubric.

The 5 silent judgments

The deductions a partner makes while skimming — the ones that almost never make it into the “pass” note. Ordered worst first.

01An ask that never gets made (to an investor)High

The kit asks advertisers for ~$80,000 in display commitments — not investors for capital. There is no raise amount, no use of funds, no milestone the money buys. Even the ad ask isn't tied to an outcome the buyer can bank. As a fundraising template this is the biggest gap: a reader finishes unsure what decision they're being asked to make. (In context, correct — it was never meant to raise a round.)

02An audience described in three different unitsHigh

The headline is 70,000 users; the ad buy “reaches thousands”; the 2005 follow-up proposal reports 1.9M monthly unique visitors. Users, reach, and unique visitors are three different denominators, and a buyer — or a partner — can't reconcile what they're actually getting. The rubric flags exactly this: a number whose definition quietly changes.

03A market asserted by college count, never sized in dollarsMedium

The growth claim is “20 colleges now, 200 in six months” — a real bottom-up by headcount, which is more than many decks manage. But it never converts to a dollar market: no ad-spend-per-student, no reachable-audience revenue, no TAM a partner can underwrite. The opportunity is counted in logos, not dollars.

04Traction shown, not toldStrength, near-zero deduction

This is where the kit wins. 70,000 real users across 20 named schools, roughly two months after launch, with campus penetration a partner can sanity-check instantly. It's arguably the best-shown early traction in our entire set — a number on the page where most decks of this age put an adjective. Almost nothing to deduct.

05A one-line story that became literally trueStrength, near-zero deduction

“A targeted, verified, deeply engaged college audience you can reach by profile.” It's clear, it's durable, and the targeting-by-profile thesis is the business Facebook actually built. The story holds with the sound off. Minimal deduction.

5 ranked fixes

Each judgment maps to one concrete fix — a rewrite or reframe you can apply today. Ordered by impact.

01Add a capital ask with a milestone-to-dollar bridgeHigh

If this were a raise: state the amount, the runway it buys, the one metric it's built to hit (e.g. colleges live, or weekly actives per campus), and why hitting it de-risks the next round. End on the decision, not the inventory.

02Pick one audience metric and define it onceHigh

Lead with a single number — say monthly active users — defined one way, and hold it across every slide and every follow-up. Stop mixing “users,” “reach,” and “unique visitors.” Consistent denominators are what let growth read as growth instead of a redefinition.

03Size the market bottom-up, in dollarsMedium

Keep the college count, then convert it: students per campus × campuses × realistic ad spend (or ARPU) = a reachable-revenue number you can defend under the first question. Beachhead first; keep the global ceiling in the appendix.

04Keep leading with traction you can showLow (protect the strength)

The 70,000-users-at-20-colleges open is the best thing in the kit — don't bury it behind setup. Whatever else changes, the first thing on the page should stay a number a partner can verify.

05Make the story hold between updatesMedium

The jump from 70,000 users (2004) to 1.9M unique visitors (2005) should read as one curve, not two vocabularies. Report the same metric, defined the same way, every time — so the next investor update compounds the last one instead of resetting it.

What this means for your deck

Facebook's media kit is proof that shown traction and a durable one-line story will open almost any door — and proof that neither is the same as a fundable deck. It got meetings because it put a verifiable number on the page and told a story you could feel in one line. It would lose points in a raise because it never asks, never sizes the dollar opportunity, and never pins its audience to one definition. Copy the first two moves exactly. Then add back the three this kit never needed: one consistent metric, a market in dollars, and an ask tied to a milestone.

Frequently asked questions

What did Facebook's pitch deck score on Vintane's rubric?

76 out of 100. It posts one of the highest traction-and-story marks in our set — 70,000 real users across 20 colleges, two months after launch — but loses ground on three judgments: there's no capital ask, the audience is described in inconsistent units, and the market is never sized in dollars.

Was Facebook's “pitch deck” actually a VC fundraising deck?

No — and that's what makes it unusual. The famous artifact is an ad-sales media kit that then-CFO Eduardo Saverin used to pitch advertisers in spring 2004, asking for roughly $80,000 in targeted display buys against 70,000 users at 20 colleges. It's the most famous “pitch deck” in the canon that was never a fundraising deck at all.

Why does it score 76 if Facebook became one of the most valuable companies on earth?

Fundable and flawless are different things. The kit wins the two hardest judgments — traction shown, not told, and a story that turned out literally true — but as a pitch-deck template it has no capital ask, mixes “users,” “reach,” and “unique visitors,” and counts its market in colleges instead of dollars. Those are the gaps a partner marks silently.

What does Facebook's deck do better than almost any in our set?

It shows traction instead of describing it. 70,000 real users across 20 named colleges roughly two months after launch, with campus penetration a partner can verify at a glance — a number on the page where most decks of that age put an adjective.

What should founders copy from it — and what should they not?

Copy the discipline: open on traction you can actually show, and a one-line story you can follow with the sound off. Don't copy the omissions — add a capital ask tied to a milestone, one consistently-defined audience metric, and a market sized in dollars, not logos.

Every deck below is scored on the same five silent judgments.

Mixpanel pitch deck teardown — scored 80/10080/100 — the strongest efficiency proof in our set after Shopify (a two-year revenue curve and a ~6-month sales payback on the page), held back by no market slide at all and an ask that never asks.BuzzFeed pitch deck teardown — scored 78/100A sharp contrarian thesis and elite founder-market fit in viral distribution — the gaps are a market sized top-down from total ad spend and traction shown as one-off hits, not a repeatable curve.Brex pitch deck teardown — scored 81/100Fast, legible traction and elite founder-market fit — the gaps are a market sized top-down from total card spend and an ask that lists plans instead of defending the number.Buffer pitch deck teardown — scored 81/100A rare deck that shows real revenue on the page — the gaps are ask clarity and a defended market, not the story.Moz pitch deck teardown — scored 81/100The most transparent deck in our set — a profitable business with its whole funnel on the page and four slides naming its own risks — held back by a market sized top-down from category ad-spend instead of its own ~$93 ARPU, and an ask that routes $6–7M to founder secondary.Intercom pitch deck teardown — scored 74/100A sharp, opinionated vision that reads beautifully — held back by traction told, not shown, and a story slightly ahead of the metrics.Spotify pitch deck teardown — scored 74/10074/100 — wins the hardest judgment outright (traction shown, not told, as real architecture on the page), then skips the ask, the market, and the fundable story, because it's an engineering talk, not a pitch deck.Front pitch deck teardown — scored 79/100Crisp positioning and real Series-A traction — the gaps are a top-down market and metrics that need one consistent definition.Airbnb pitch deck teardown — scored 77/100Legendary clarity and a problem anyone grasps in one line — the gaps are an undersized market and traction shown too thin for the story.Uber pitch deck teardown — scored 78/100A problem quantified to the dollar and a market slide most Series-A decks still can't match — held back by a vague ask and a headline outcome that swings from $20M to $1B on the same slide.LinkedIn pitch deck teardown — scored 80/100The greatest concept pitch ever written — a story that never wobbles and an ask already de-risked by over-delivery — held back by a market slide that's 73% of a category it drew itself, and the flattering framing that kept $0 revenue off the headline.Square pitch deck teardown — scored 80/100One of the most-passed-around decks in startup history — and it earns it: real traction on the page and arguably the strongest early team in the canon. The gaps are a top-down market slide and numbers that only reconcile once you open the appendix.DoorDash pitch deck teardown — scored 82/100The most decisively traction-led deck in the canon — real orders plotted as a live curve, and a story so stable it reappears word-for-word in the S-1 seven years later — held back only by a market shown as a behavioral proxy instead of a bottom-up build, and no ask slide at all.Dropbox pitch deck teardown — scored 76/100The clearest value prop ever put on a slide — 'Storage is a mess' answered by 'It just works' — carried by a live demo where a traction number should be. It wins the clarity-and-story judgment outright, but the seed deck shows no metric a partner can trace, sizes the market top-down as everyone with a computer, and never makes an ask.WeWork pitch deck teardown — scored 72/100The most-studied raise in startup history — the deck that sold Goldman, JPMorgan and Harvard on a ~$5B valuation. The narrative is one of the best ever written and the traction is genuinely shown, which is why it scores as high as it does. But it sizes its market top-down from a 5%-penetration fantasy, stamps 'Proven, Profitable' on unit economics shown only at maturity, and relabels a lease-arbitrage business 'space as a service' to borrow a tech multiple — the three judgments that, five years later, the IPO failed.Coinbase pitch deck teardown — scored 79/100The deck that had to teach the category before it could sell the company — traction genuinely shown (20% daily signups, $65K in five weeks) and a wedge-to-payment-network story still true 13 years later, held back by a market slide that borrows the whole Bitcoin network's $2M/day volume instead of a bottom-up build, a headline that conflates that with Coinbase's own number, and no ask or team slide.Mint pitch deck teardown — scored 77/100A pre-launch deck that wins the slides most seed decks fail — an elite team (PayPal, PGP, and Intuit's own founder on the masthead), a market built bottom-up to a named number, and the clearest referral-economics slide in the canon — then fails the one judgment it can't buy its way out of: there is no traction, because the product hadn't launched. The projection standing in for it leans roughly two-thirds on a targeted-advertising line that never became the business, and an exit slide names Mint's acquirers before it has a single user. Prescient — Intuit, listed on that slide, bought Mint two years later — but on paper, a plan where a number should be.YouTube pitch deck teardown — scored 75/100The purest example in the canon of traction told, not shown — a deck whose slide literally titled 'Metrics' contains no metric, only the sentence 'Launched June 11th. Has already overtaken all previously existing competitors.' It wins the one judgment that mattered most: a one-line story — 'the primary outlet of user-generated video content on the Internet' — so stable it still describes the company twenty years later, carried by three PayPal builders a partner couldn't help but trust. But the market slide has no number, there is no ask slide, and the real growth charts — 100,000 videos a day, a top-10,000 site in two months — lived in Sequoia's own memo, not the founders' slides. It raised, and Google bought it for $1.65B thirteen months later. On paper, a team-and-timing bet the numbers were never put on the page to make.Peloton pitch deck teardown — scored 71/100Peloton's seed deck is the purest test of the rubric in our entire set: one of the clearest product visions ever drawn — a three-circle concept anyone grasps in seconds, and roughly 40% of its slides spent showing the exact leaderboard-and-live-class experience the real company later shipped — attached to almost none of the things a partner actually underwrites. There is no traction, because there was no product and no company yet. The market slide is a single top-down $10.5B fitness-equipment number that sizes the wrong business — it never counts the subscription revenue that became ~$4B of Peloton a decade later. There is no business-model slide, no team slide, and no ask. By the founder's own account, roughly 400 venture firms passed and he pitched on the order of 3,000 angels before the money came together. It scores 71 — the lowest in our set — and the company IPO'd at about $8B anyway. The most instructive deck we've run: copy the product clarity exactly, then put on the page every fundable thing Peloton left off.Tinder pitch deck teardown — scored 73/100The most beloved problem slide in the canon — fear of rejection, told through one guy at a party — attached to almost nothing a partner underwrites: no market, no ask, no team, no traction. A masterclass in story, and a lesson in everything a story can't carry.Canva pitch deck teardown — scored 78/100A genuinely prescient product vision and a rare real proof point in Fusion Books — held back by a market framed as 'everyone designs' that partners couldn't put in a box, and an ask carried by vision more than Canva's own numbers.Robinhood pitch deck teardown — scored 79/10079/100 — the clearest one-line wedge in the canon and a team partners trust on sight, held back by a market sized by population, a waitlist for traction, and an ask that leads with "free."Shopify pitch deck teardown — scored 83/100The highest score in our set and one of only two decks already profitable when it raised (the other is Moz) — recurring revenue growing ~150% a year, strong retention, ~$130M in GMV, all on the page. Held back only by a market sized top-down when it had its own ARPU to build it up, and a headline GMV that dwarfs the company's own revenue.Revolut pitch deck teardown — scored 76/100A problem priced to the dollar — the hidden FX markup every traveler pays — held back by traction told not shown, a market sized by headline volume, and a 'zero fees' hook that never shows where the money comes from.Snapchat pitch deck teardown — scored 75/100The other famous ‘pitch deck’ that was never a pitch deck. Snapchat's 2014 deck sells brands on a disappearing-content platform, not investors on a round — and it leads with two of the hardest judgments won outright (700 million Snaps and 1 billion Story views a day, and a one-line philosophy that every rival later copied), then skips every slide a raise lives on: no ask, no dollar-sized market, and metrics counted in views and screenshots.Foursquare pitch deck teardown — scored 75/100The clearest consumer story in the canon paired with elite founder-market fit (Dodgeball → Google) and a business model proven by observed behavior — local merchants already rewarding “mayors” unprompted — held back by no market slide at all, no ask, and traction shown as anecdote instead of a plotted curve.Netflix pitch deck teardown — scored 73/10073/100 — wins the one judgment most decks fail outright (a story that never shifts, proven over fifteen public years), then skips the ask, the market, and the traction curve, because it's a culture manifesto, not a pitch deck.Mattermark pitch deck teardown — scored 79/100Mattermark — 79/100. The rare deck that pitched investors on a tool investors already used — and showed its own MRR curve ($125K MRR, 377% CAGR) the way it sold showing everyone else's. Held back by a market that balloons from a credible $3B wedge to “every knowledge worker,” a $10M-ARR projection with no ask to anchor it, and 54% of revenue riding on a 30,000-seat niche.The 5 silent judgments — the rubric every teardown usesStart here: what VCs judge silently, and the tell that fails each one.

Run your own deck through the same rubric

Paste your deck and last 10 founder updates for your own fundability score — the exact rubric used in the teardown above. Free, about 30 seconds, no signup.

Run your own deck through the same rubric — free, ~30 sec, no signup →

Free score, no payment needed · your deck stays private