Specialising does not make you better, only harder to replace

Niching down does not make you better at the work. It makes you harder to substitute, and that is the only mechanism by which anyone charges more without losing volume. The size of the effect is measurable: on Malt's European platform, generalist business consulting averaged €415 to €462 a day, while named specialists in project management and agile coaching averaged €795 to €1,074.

Everything else usually written about this — the positioning language, the personal brand, the confidence to say a bigger number — sits downstream of that one fact and mostly distracts from it.

This is a question about substitutes, not about focus

The debate is normally framed as a trade-off between reach and depth. Go broad, serve more people. Go narrow, serve fewer but better. That framing makes the decision impossible, because it puts volume on one side and quality on the other and gives you no way to compare them.

The real mechanism is simpler. Your price is bounded by what the buyer's next-best option costs them. If someone looking for business consulting can name eleven people who plausibly do it, your price is capped by the cheapest credible one of the eleven. If someone looking for a person who has restructured distributor margin in a family-owned industrial supplier three times can name two, the cap moves. Nothing about your competence changed. The buyer's alternative got worse.

Specialisation is not a claim about how good you are. It is a claim about how short the buyer's list is when they go looking for someone else.

This is also why rate advice given on its own tends to fail. A rate is not something you decide; it is something the market ratifies or refuses. Experts underprice less often because they misjudge their worth than because they are reading, accurately, a substitute set that is large. Telling them to charge more without changing that set is asking them to ignore correct information.

What the day-rate data actually shows

Average freelancer day rates by category — Malt × Boston Consulting Group, “Freelancing in Europe 2022”, n = 3,334, fieldwork July–September 2021.

Category	France	Spain	Germany
Business consulting	€415	€446	€462
Project management / agile coaching	€982	€795	€1,074
Tech & data	€746	€410	€652
Marketing & communications	€546	€200	€615
Arts & design	€311	€217	€419

Take the first two rows and divide. France: €982 ÷ €415 = 2.4. Germany: €1,074 ÷ €462 = 2.3. Spain: €795 ÷ €446 = 1.8. Same platform, same pool of buyers, same fieldwork window — and the named specialism is worth between 1.8 and 2.4 times the generic label.

€612 a day separates generalist business consulting from named project management and agile coaching in Germany — €462 against €1,074, on the same platform in the same period.

Now the caveats, because they matter and they do not dissolve the finding. This is platform data, so it describes freelancers who list on Malt rather than the whole market. People self-select into categories, so the gap partly reflects who chooses to call themselves what. And it compares averages across categories, not the same person before and after narrowing — which makes it consistent with the substitutes argument rather than proof of it. That is correlation, and I would rather say so.

What survives the caveats is still substantial. Buyers on one platform pay roughly twice as much for a described capability as for an undescribed one, and the pattern repeats in three countries with different labour markets and different absolute price levels. Spain's narrower gap of 1.8 is itself informative: the lower the general price level, the less room a specialism has to open up.

The same shape shows up in DACH hourly rates

Average hourly rates by sector — Freelancer-Kompass 2026, freelancermap, n = 5,412, DACH region, fieldwork November 2025 to February 2026.

Sector	Average hourly rate
Consulting / management	€121
Finance / accounting / legal	€114
SAP / ERP	€113
Data and analytics	€108
HR / coaching	€106
IT infrastructure	€100
Marketing and communications	€94
Engineering	€91
Software and web development	€91
Design / media / creative	€89
Research and analysis	€89
Overall average	€103

Top to bottom is €121 against €89: a spread of €32 an hour, or €121 ÷ €89 = 1.36 times. A much tighter band than Malt's 1.8 to 2.4, and the difference is instructive rather than contradictory. These are broad sectors, not specialisms. Nobody is priced by being in engineering. They are priced by how replaceable they are within it, and a sector average flattens exactly the variation this article is about.

Two other figures from the same study are normally reported separately, and they are the same fact from opposite ends of the table. 70% of respondents reject projects because the hourly rate is insufficient. 43% have no secured project utilisation for the coming months. Read together they describe a market in which a large share of people are simultaneously refusing work at the offered price and short of work — which is what a crowded substitute set looks like from the inside. The overall average was €103, down from €104 in 2025, the first decline since the study began. Just under 1%. The direction is the part worth noticing.

The honest counter-argument: a narrow market is a fragile market

The fear is legitimate, and most writing on this subject waves it away — which is why the writing fails to persuade anyone who has been in business fifteen years.

A narrower niche is a smaller market. Fewer buyers means a single technology shift, a single regulatory change or one bad year in one sector removes a larger share of your demand at once. If you specialise in an industry that consolidates from forty buyers to twelve, your addressable market did not shrink by 10%. It shrank by 70%, and you found out in the same quarter as everyone else who serves it.

So the conclusion is not "niche as hard as possible". There is a point past which further narrowing stops reducing your substitutability and starts concentrating your risk. Where that point sits depends on three things you can estimate: how many buyers exist in the niche, how often they buy, and whether the thing you specialise in is a durable business problem or a passing tool.

That last distinction does most of the work. Someone who specialised in a particular reporting platform and became its acknowledged authority had a genuinely small substitute set — until the platform was superseded, at which point the specialism became a liability, because the expertise and the obsolescence were the same asset. Someone who specialises in a recurring business problem — pricing, succession, distributor margin, cash conversion — has a substitute set that stays small while the underlying problem keeps returning. Both narrowed. Only one narrowed into something that renews itself.

Buyer, problem or method — three niches with three different risks

This is the distinction most content on the subject misses entirely, and treating the three as interchangeable is how people niche themselves into the wrong corner.

Niching on the buyer — Swiss family manufacturers, dental practices, independent wine importers — makes referral work. Buyers in a defined group know each other, meet at the same associations, and ask each other who they used. Your acquisition cost falls because other people are doing your qualifying. What it does not automatically do is shrink the substitute set, because a generalist who knows the sector can still be considered.

Niching on the problem — pricing, margin recovery, succession funding — is what makes the substitute set small. Buyers with an acute, expensive, recurring problem search by the problem and compare only people who name it. This is the axis that moves price, because it is the axis on which the buyer's alternative genuinely worsens.

Niching on the method — a named framework, a certification, a particular tool — is the fragile one. Methods age. A method-based niche is also inherited from whoever owns the method, which means your scarcity is on loan and can be diluted by the certifying body issuing more certificates. That is a substitute set someone else controls.

The combination that holds up is a problem plus a buyer: a recurring, costly problem, inside a defined group of buyers who talk to each other. The problem keeps the substitute set small; the buyer group keeps acquisition cheap. Method is a way of doing the work, not a way of being chosen. Mastery of the craft sits on the method axis, which is why on its own it so rarely translates into price.

Will you get fewer clients?

Fewer enquiries, almost certainly. Fewer clients, not necessarily — and the second is the only one that matters, because the enquiry count is a vanity number that costs money to service.

Consider what a broad positioning actually buys. Malt's study reports that 24% to 28% of freelancers' time goes on ancillary work — training, administration and canvassing — at 27% in France, 28% in Spain and 24% in Germany. It separately reports average working weeks of 37 hours in France, 42 in Spain and 41 in Germany. Joining those two figures is an assumption on my part, since the study reports them independently, but if the ancillary share is of total working time then Germany's 41 hours less 24% leaves about 31 hours; Spain's 42 less 28% leaves about 30; France's 37 less 27% leaves about 27. Roughly a quarter to a third of the week goes on finding and administering the work rather than doing it.

A generalist positioning is what makes that share large, because every enquiry has to be qualified from scratch and every proposal argues the case from first principles. And the single most-cited challenge in all three countries was negotiating with clients — 24% in France, 29% in Spain, 28% in Germany. Negotiation is difficult in direct proportion to how substitutable you are. The person with three credible competitors negotiates. The person with fifteen concedes. Being busy without making money is very often this arithmetic rather than a productivity problem.

What this looks like in practice

The useful test is not how narrow a description sounds. It is whether a buyer with your problem, writing an email at nine in the evening, could name three people to send it to — and whether you are reliably one of them. That test does real work because it can be checked rather than debated. It asks who else the buyer would consider, which is answerable from your last dozen competitive situations: who else was in the room, on what basis you won or lost, and whether price decided it. Where price decided, the substitute set was large. Where it did not, something in your description had already removed the alternatives.

It also turns the market-size worry into something countable. If you can identify roughly how many organisations have the problem you name, how often it recurs, and what it costs them when it goes unsolved, you have the two numbers that matter: whether the niche can sustain you, and whether it renews. A niche of two hundred buyers facing the problem every three years behaves quite differently from two thousand buyers facing it once.

What usually emerges is not a dramatic narrowing but a correction of axis. Most people who feel too general are in fact reasonably specific about their buyer and entirely vague about the problem — the combination that produces plenty of enquiries and weak pricing. Undercharging is frequently the symptom of that misalignment, not of nerve.

What this cannot tell you about your own case

Neither study describes your market. Malt's is platform data from 2021 fieldwork across France, Spain and Germany, in categories that may not map to what you sell. The Freelancer-Kompass covers the DACH region and reports sectors, not specialisms. Both are averages, and averages conceal the exact variation you are trying to price into.

What would settle it for you sits in evidence you already hold and probably have not assembled. Your last twenty enquiries, sorted by what the buyer said they were looking for. Your win rate when shortlisted against named alternatives versus when you were not. The share of revenue arriving through referral, and whether those referrals travelled through a buyer group or a problem description. What price resistance looked like in each case — a negotiation, or a silence. Those describe your substitute set far more accurately than any published rate table.

Nobody is normally engaged to look at them. An accountant reports what you earned, not what you could have charged. A brand consultant refines how you describe yourself without testing whether the description changes who you are compared against. A coach works on what you are willing to say. The commercial question — how many people the buyer can call instead of you, and what it would take to reduce that number — falls between all three professions and is therefore nobody's. It is also the only question whose answer moves the price.