Specialise, then refer the rest: the referral case for niching

Ask a generalist practice who its ideal client is and the answer is usually 'anyone who pays'. It sounds commercially sensible. It is commercially invisible. A firm that acts for everyone competes on the one dimension every rival shares — price — and its own contacts, asked to describe it, can rarely manage anything sharper than 'they're good, they do... accounts?'.

Vague descriptions produce vague referrals, and only occasionally. Meanwhile the specialist down the road — the accountant known for dental practices, the solicitor who lives and breathes agricultural property — gets recommended by name, unprompted, because their label maps neatly onto a specific person's specific problem. Nobody refers to a category. People refer to the firm for a situation, and generalists have quietly opted out of ever being that firm.

A niche is a repeatable client situation, not a mild preference. Working examples across the professions include GP and dental practices, farming families facing succession, law-firm partners with their distinctive tax profiles, divorcing professionals, ecommerce sellers wrestling with cross-border VAT, and first-generation owners approaching an exit. The test is threefold. Is the group findable — do they cluster in associations, publications and conferences? Is the work repeatable — does client eight benefit from everything you learned on clients one to seven? And is there money in it — will the niche support your fee levels?

A niche that passes all three turns every completed engagement into training for the next, which is why specialists get faster and more profitable at exactly the work where generalists start from scratch each time. The premium follows naturally: clients pay more for someone who has seen their precise problem many times before, because the perceived risk of a botched job falls away.

Here is the objection that stops most firms specialising: 'If I narrow down, I'll have to turn work away.' It is a fair worry, and it points straight at the real problem with niching. When you become the go-to accountant for dental practices, your clients do not stop having lives. They still ask about mortgages, pensions, protection, wills and estate planning. Your niche sharpens what you sell, but it does nothing to narrow what your clients need.

Without a plan, every one of those out-of-scope questions becomes a small moment of failure — a shrug, a 'not our area', a client left to fend for themselves. Do that often enough and the very focus that won you the client starts to feel like a limitation. Specialising creates more adjacent questions, not fewer, precisely because it deepens the relationship. The question is what you do with them.

This is where a referral network turns niching from a trade-off into a clean win. Instead of apologising for what you do not do, you answer every out-of-scope question with the most reassuring sentence in professional services: 'I know exactly who to send you to for that.' You keep the deep, defensible work your niche is built on, and you hand everything else to vetted, regulated specialists through a tracked introduction — with consent captured and disclosure paperwork generated for you.

Done this way, specialising loses you nothing. The client who needs financial advice or estate planning still gets looked after, still associates the solution with you, and still has every reason to stay. And because the introductions are tracked, they also earn: where a referral produces a fee, a member share of around 60–70% turns the questions you used to wave away into a genuine income stream. You are no longer choosing between focus and breadth — you have both.

Niching does not require sacking your client base on Monday. The lower-risk route is a twelve-month experiment: keep serving your existing clients while pointing new effort — say a fifth of your business-development time — at one chosen niche. Write a few genuinely useful pieces for that audience, show up where they gather, and tell your contacts plainly that this is the work you want more of, because they cannot send what they do not know you seek.

At the same time, put a referral route in place for everything the niche throws up that you do not handle, so no adjacent question ever goes unanswered. Review at the year mark against honest numbers — enquiries, conversion, average fee, and referral income — and feed more effort into whatever is working. Most firms that run the experiment properly never go back to 'everyone'. Everyone was never a market, and with a network behind you, it was never a necessary one either.