Specialization
Специализация
Uniqueness Position Depth → 3 · Drill rig
deep competence in a narrow niche
Moats it grows into
- → #17 Counter-positioning
Reached through Counter-positioning: the niche has to demand something the generalist incumbent cannot offer without breaking its own economics, a price, a simplification, a policy its core customers would punish. Depth alone is competence a rival can hire.
Summary
Specialization is the choice to serve one narrow job better than anyone who serves many. On the depth ruler it is a position, not a moat: depth of competence is what you do, not what stops a rival from doing it too. It becomes a moat only when the niche is one the generalist cannot enter without damaging its own business, and that is Counter-positioning.
How it works
A narrow target changes what you can build. The generalist has to keep its median user satisfied, so every screen, default and price is a compromise. You serve one job, so you can ship the workflow only that job needs, use its vocabulary, price by its economics and sell through the channels its people already read. Learning compounds too: every customer teaches you about the same problem, where a generalist's customers teach it about a hundred.
That is an edge, not a wall. Competence can be hired, and a generalist can add a vertical edition on top of its platform in a quarter. What it cannot do is serve the niche where doing so contradicts its own model: a cheaper tier that cannibalises its enterprise plan, a simpler product that embarrasses its flagship, a data policy its main customers would refuse. Aim the specialization at that contradiction and the position holds; aim it at mere focus and you are a feature waiting to be added.
Why it holds — or only takes
Being deep in a niche is a position: it decides who you sell to and what you build, and a rival can decide the same thing tomorrow. Nothing in depth itself raises their cost of following you. The one moat it reaches is Counter-positioning, and only when the niche's demands are something the incumbent cannot meet without hurting its own business, its pricing, its channel, its flagship product. Where the generalist can simply add a vertical, specialization is a segment, and the generalist's distribution wins in the end. Category leadership is not mapped: a niche is rarely a new category, it is a slice of an old one.
Do you have it
- Your customers describe their job in words a generalist's marketing does not use, and your product uses those words.
- Prospects arrive from the niche's own forums, conferences and newsletters, not from generic search.
- You have said no to a feature request because it served people outside the niche.
- The generalist's sales team refers the smallest, oddest accounts in your niche to you.
- Your pricing follows the niche's economics, per bed, per boat, per case, not per seat.
How to dig it
- Pick a job the generalist handles badly because it is too small, too regulated or too odd to be worth a product decision.
- Rebuild the core workflow around that job end to end; delete every screen the niche never opens.
- Price on the unit the niche counts in, and sell through the one channel its people already trust.
- List the choices the incumbent cannot copy without a price cut or a policy change; that list is the moat, the rest is focus.
How it erodes
- The generalist ships a vertical edition with your top ten features and bundles it into the plan your customers already pay for.
- The niche grows until it is worth a product decision at the incumbent, and its distribution beats your depth.
- The niche shrinks, is regulated away or its job gets automated, and depth in it is worth nothing.
- An AI layer on a generalist product learns the niche's vocabulary and workflow from prompts, closing the competence gap in weeks.
Solo, without capital
This is the most natural position for one person: pick a job you know from the inside, build only its workflow, and sell to the twenty forums where it is discussed. Counter-positioning is solo-reachable and rents to nobody, so the moat, when it appears, is yours. What you cannot do is defend a niche the incumbent decides to want. The only safe niche is one it would lose money serving, so choose it for the contradiction, not for the size.
In nature, in business
- Nature
- pandas
- Business
- trade schools
Cases
- Veeva Systems — Built a CRM only for life-sciences sales reps, on top of a generalist platform, with the regulated call-reporting the generalist would never build for one industry.
- Toast — Sells point-of-sale only to restaurants, with the kitchen screens, table flows and tipping that a general retail till treats as edge cases.
- Giant panda — Eats almost nothing but bamboo, a food no rival wants, with a false thumb evolved for gripping the stalks, and depends entirely on the bamboo surviving.
Combines with
- Craftsmanship — Narrow scope is where extra care pays: a single job done with more precision than the generalist can afford to give it.
- Completeness — Be the one-stop shop for the niche: every tool its job needs in one place, none of the tools it does not.
- Prestige — A niche has its own authorities; their endorsement reaches every buyer in it, where a generalist's badge reaches none of them.
Fights with
- Versatility — Every job you add to please a new segment dilutes the fit the niche pays for, and the generalist notices.
- Diversification — Depth means betting on one stream; spreading across several is a hedge against exactly that bet.