Two kinds of narrowness, often confused. A wedge is narrow in the problem it solves. A beachhead is narrow in the people it solves it for. You need both, and they are chosen by different criteria.
Chris Dixon's formulation: attack a narrow, overlooked problem first, win users with something simple, then expand into the full standalone product from that position. His examples are worth holding onto because they are unglamorous. Instagram started as filters and sharing, a gap the iPhone camera left open. OpenTable sold reservation terminals to restaurants long before it was a consumer booking site.1
Dixon's own warning is the useful half of the essay: most proposed wedges are what he calls dinky features. A dinky feature is also narrow and also simple, and it goes nowhere, because winning it leaves you holding nothing the next step needs.
This is the single idea to take from this lesson. A wedge is only a wedge if winning it produces an asset the expansion requires and that you could not have bought directly. Four kinds show up in practice:
| Bridge asset | What it buys you | Example |
|---|---|---|
| Accumulated data | The expansion starts full instead of empty | Years of client history that a new CRM would otherwise ask you to import |
| Workflow position | You are already in the moment where the next decision is made | Sitting in the meeting means you are there when the follow-up is decided |
| Distribution | The next product ships to an installed base, not to strangers | OpenTable's terminals in restaurants before the consumer site |
| Trust and reference | The expensive, risky purchase is now from a known party | Being the tool that never lost anyone's data for a year |
Notice what is not on that list: a list of email addresses. Signups are not a bridge asset. They decay, they belong to whoever emails best, and the incumbent can buy the same names. This is exactly why Lesson 2 insisted that an engineering-as-marketing tool is a channel and not a wedge, no matter how many people sign up for it.
Four questions. A real wedge answers all four. Failing question three is the common case and is fatal.
Geoffrey Moore's beachhead is a different cut. You do not pick a small segment because you want a small business; you pick it because dominating one small segment produces references, word of mouth and a complete solution, and those are what let you take the next segment. He calls the sequence bowling pins: the first pin has to knock down the second.2
Moore's selection criteria, compressed:
Independent agents and small agencies in France score unusually well on the first criterion. This is a market with trade press, a national trade body, franchise networks, and a 12,000-person annual trade show. Word travels. That cuts both ways: a product that embarrasses an agent in front of a client travels just as fast.
You can have a good wedge aimed at a bad beachhead, and the reverse. Writing them as one sentence forces both to be explicit:
For [narrow group of people] who [share one urgent reason to buy], we solve [one narrow overlooked problem], which accumulates [named bridge asset], which makes [the expansion] materially easier than selling it cold.
Fill that in for the document's proposal before writing the review and keep it. Most of the review writes itself from the gaps in that sentence.
Listen to or read Geoffrey Moore on finding your beachhead. It is Moore on his own framework, which beats every summary of it, including this one.
Want to pressure-test a wedge sentence you have written? Paste it to your teaching agent and ask it to attack question three specifically.