Altitude Development PartnersIn the 1830s, the men who quarried granite in New England split it with a row of slim iron wedges. They drilled a line of shallow holes across a block the size of a delivery van, dropped a wedge into each, and tapped down the row, tap, tap, tap, until a stone a hundred men couldn't lift fell open along a clean seam. The wedge never out-muscled the granite. It found the one line where the rock already wanted to break, and put all of its force there.
For the first years of building my company I was the other guy, swinging a sledgehammer flat against the cliff and wondering why it kept winning. You can be smart, work for years, and build something people like, and they will tell you they like your work, right before they don't pay you.
The founders who break into a market just found the seam. Investors say the word so often it is basically wallpaper: you need a wedge. Here is what it means, why most wedges aren't wedges at all, and how to build one that can pry open a market.
A wedge is a strategy for winning a large market by first capturing a tiny sliver of it, or a big chunk of a small market next door. It's just a sharp edge and the right place to strike. In business, a sharp product and the right initial market. Miss either one and you are holding a chisel with no wall, or a wall with no chisel.
The clearest working definition is Lenny Rachitsky's: the right tool plus the right place to strike. You shave 99% of your grand vision off and aim the remaining 1% at a single point, one buyer, one painful job, one place you can win, because concentrated force is the only kind that breaks anything. Sarah Tavel of Benchmark puts it best: don't be "a blunt instrument trying to chop into a market by being everything for everyone." Be "the sharp blade with extraordinary focus." Everyone nods at this but nobody does it, because shaving 99% off your dream is like very scary.
The fastest way to believe in wedges is to notice how ridiculous the great ones looked at the start.
Airbnb: an air mattress for a sold-out city. Started as: two broke designers in San Francisco in 2008, buried in credit-card debt, three air mattresses on the living-room floor. When rent came due, they hand-folded around a thousand cereal boxes, Obama O's and Cap'n McCain's, and sold them at the political conventions for $40 a box, clearing roughly $30,000. Paul Graham later said he funded them largely because of the cereal: anyone who could move $40 novelty cornflakes could survive almost anything. The wedge was never "reinvent travel." It was an air mattress on the night every hotel in town sold out.
DoorDash: a binder full of canceled orders. Started as: four Stanford students in 2013 with a one-page site and their own cell number. They noticed a binder by the register of a University Avenue macaron shop, page after page of orders stamped CANCELED, because the owner had no way to deliver them. They put up PaloAltoDelivery.com, listed a few menus, and posted their number. The first order was shrimp pad thai, and they delivered it themselves, ducking out of class. That binder itself was the seam. How much smaller than that can you get?
PayPal: the wedge they didn't plan. Started as: encryption for beaming money between Palm Pilots, a sentence that ages like warm milk. The real wedge arrived as a complaint. An eBay seller had screenshotted the PayPal logo to make her own "we accept PayPal" button. Instead of a cease-and-desist, PayPal built her a nicer button and paid $10 per referral. Every auction listing became a billboard, and by 2002 roughly 70% of eBay auctions took PayPal. Sometimes the market tells you where your edge is, and the trick is listening instead of suing.
Stripe: seven lines of code. Started as: two brothers, 19 and 21, for whom getting paid online was harder than writing the software. They wrote about seven lines of code that let any developer accept a card. The wedge was developers, a group every bank treated as a rounding error. The business became a slab of global financial infrastructure.
None of these sounded like a billion-dollar company on day one. They sounded like a hobby, a stunt, or a feature. A good wedge is supposed to look too small. If it looked big, it wouldn't be sharp.
Because every startup has a fixed supply of money, time, and attention, and can either smear it across a continent or pile it onto one square foot. A narrow idea is easier to explain ("drop-in audio for Silicon Valley Twitter" lands; "a new kind of social network" does not), easier to balance in one zip code (which is why Uber launched as overpriced black cars in San Francisco, not a global app where riders tap the button and watch nothing arrive), and easier to learn from (Airbnb could fly to New York and shoot its listings by hand, because at the time New York was the whole company). A wedge is a promise small enough to keep. Most struggling startups make a promise to everyone and keep it for no one.
Not every narrow idea is a wedge. Run it past five impostors. If any of these sounds like your business, you don't have a wedge.
The blunt instrument is everything for everyone on day one. Zaarly launched as a "reverse Craigslist" for requesting anything, learned you cannot be excellent at delivering anything, and narrowed to home remodeling the hard way. Breadth feels safe but it is the most dangerous move on the board.
The island is a niche so peculiar nothing you learn there travels. The test: will winning niche #1 make niche #2 easier? If not, then nothing you do will help you scale the business.
The feature wins its market and captures no power, nothing a bigger company couldn't clone over a long weekend. A wedge gets you in the door; a moat stops everyone strolling in behind you (the companion piece: What is a moat and how to get one).
The vitamin is pleasant and easy to skip. If your beachhead client has no urgent reason to buy right now, focus won't save it, and less familiar clients definitely won't buy. A wedge has to be a painkiller.
The borrowed wedge snares the most ambitious founders: you notice that "start premium like Tesla" worked and reverse-engineer a wedge from it. A real wedge comes from something you uniquely see, a pain you have felt, a binder of canceled orders only you walked past. Copy the shape and you get the shape without the edge.
So is your business a blunt instrument, an Island, a feature, a vitamin, or a borrowed wedge? If none of the above, you might have a real wedge.
Say you have a real one. You can still waste it, and the most common place is the expansion, when the thing that made you sharp starts to make you stuck. Frank Rotman at QED studied more than a hundred companies and found the same delusions every time. "Customers love us, so the next product will be easy," except only a fraction want it and only a fraction of those buy, so your warm base is colder than you think. "Parity is fine for v2," no, if it is merely as good, your own customers go shopping. "We'll launch once it's perfect," congratulations, you now behave like the incumbent you meant to kill.
There is another trap: the wedge works so well the market files you under one heading forever. Amazon escaped "the online bookstore" but there are many companies that never did. The fix is to know, on day one, which direction you intend to swing. Tesla is the cleanest swing ever planned in advance. In 2006, Musk wrote it down: a low-volume, faintly absurd sports car funds a cheaper sedan, which funds a car for everyone. Roadster, Model S, Model 3. The Roadster was the wedge that paid for the swing.
Each example of a successful wedge above made a promise narrow enough to keep, to a group small enough to reach, and each captured something on the way in that made the next move easier. The impostors fail the opposite way: too broad to keep the promise, too isolated to travel, too thin to leave anything behind. The through-line is scope. The founders who broke into a market sounded small, specific, and focused on the business principles that could be applied in different parts of the market.
So here is where this leaves you, and me. If you have been at it for years and it still isn't working, the problem is usually not your effort, your intelligence, or even your product. Make sure you haven't been swinging a sledgehammer at the whole cliff. Swinging harder is not the answer. Stop, find the one seam where this particular market wants to break, and put everything there, knowing the slice will look small, knowing friends will ask when you'll do the "real" version, knowing it will feel like surrendering the dream when it is the only path that reaches it.
Pick the rock you eventually want to split. Design the smallest, sharpest wedge that can start the crack, and be honest, with a checklist if you have to, about whether what you are holding is a wedge or a wish. Get that right, and the granite does the thing no amount of raw force could ever make it do.
It falls open along the line you chose.