How to choose a niche graphic

If you’ve already accepted that narrowing is the right move and you’re stuck on what to narrow to, this piece is for you.

That’s a different problem to the one most articles solve, plenty of them will tell you why niching works, and I’ve written my own version of that argument. But knowing you should do it doesn’t help much when you’re sitting there with fifteen years of varied work behind you and no obvious answer.

What usually happens next is that you start looking outward, which industries are growing, where the money seems to be heading, what looks like a good space to be in. It feels rigorous and it’s usually how you’d approach any other business decision.

It’s also how you end up choosing a niche you’ve got no evidence you’re any good at.

Here’s how to choose a niche using what your business has already told you, plus the three shapes a niche can take and how to know when you’ve landed on the right one.

Contents

Why market research is the wrong starting point

Market research tells you where demand is, it doesn’t tell you where you’re credible.

Those are different questions, and only one of them you can act on this year. If you decide your niche is a sector you’ve worked in twice, you’ve now got to build proof, reputation and referral networks from a standing start in a market where somebody else has a decade’s head start on you.

There’s a second problem, which is that the sectors that look attractive from the outside are attractive to everyone else looking from the outside too. You’re probably not the only person who has noticed.

Growth data isn’t useless, it’s just more of a tiebreaker rather than a starting point. If you’ve got two niches you can already evidence and you need to pick one, by all means look at which market is heading somewhere… Just don’t let it make the decision on its own.

The businesses that niche well almost never pick something new, they notice what they’ve already become and then commit to it properly. It’s less exciting than a strategy offsite, and it works considerably better.

The method: rank your last twenty jobs

Pull up your last twenty or thirty jobs. Not the ones you’d put on your website. All of them, in order, no editing.

Then rank them by three things, in this order.

What you actually made. Not the invoice value, what was left once you’d accounted for the hours, the revisions, the scope that crept, the chasing. If you don’t track that precisely, your gut is fine here. as you probably know which jobs paid properly.

How smoothly they ran. Fewest arguments, clearest decisions, least rework and a job that made good margin but cost you three months of stress isn’t a job you want more of.

Which ones sent you someone else. Referrals are the market telling you where your work is memorable enough to describe to a third party, that’s rarer and more valuable than most owners realise.

The first thing you’ll notice is that your biggest job isn’t at the top (it almost never is), revenue and profitability tell you different stories and most businesses only track the one that shows up on the invoice.

Now look at your top five or six, there’ll be something they share, and it’s usually sitting in plain sight because you’ve been too close to it to name it. Same kind of buyer, same problem turning up again and again, same reason those clients came to you rather than someone else.

That’s your niche! You’ve probably been running it for years without saying so out loud.

So in reality you’re not picking a niche, you’re admitting you already have one.

The three ways to niche

This is where most people get stuck, and it’s usually because they’ve only ever pictured one version of niching: pick an industry, lose everyone else. That’s one option out of three, and it’s the one that costs you the most.

Take an electrician. Fifteen years in, good operator, does a bit of everything.

Route What he narrows Example What it costs him
By industry The sector he serves Only works on childcare centres Every client outside that sector
By customer The type of buyer Only works with builders, never homeowners Every client of the other type
By what he does The service itself Switchboard upgrades only, any building, anyone No client types at all

Same business, three routes to becoming the obvious choice, and only two of them involve turning any type of client away.

That third row is the one people miss, narrowing your service rather than your market is the least frightening version of this and often the most commercially useful, because it’s the one that makes your process repeatable and your pricing defensible.

You can also stack two, say switchboard upgrades for childcare centres is sharper than either on its own, because now there’s a specific person with a specific problem and you’re the only obvious answer.

Just don’t stack all three because narrow is good but Invisible isn’t.

Signs you already have a niche and haven’t named it

Run through these, if four or more land, the decision is largely made and you’re just avoiding saying it out loud.

  • The same industry keeps appearing in your last ten jobs without you targeting it.
  • You can predict the objections in a sales call before the client raises them.
  • You’ve built a process, template or checklist for one type of work and reuse it constantly.
  • Referrals come from the same handful of sectors.
  • There’s a type of enquiry where you quote faster and win more.
  • You’ve got extensive experience about one particular kind of problem.
  • People in one industry seem to already know who you are.
  • There’s a category of work you quietly hope doesn’t come in.

That last one is worth sitting with. The work you dread is data. It’s telling you where you’re competent but not distinctive, which is exactly the work that eats margin and generates nothing.

How narrow is too narrow

The most common objection I hear at this point is that the market’s too small.

Usually it isn’t, and the reason is a counting error. You estimate the size of your niche by tallying up the ones you personally know about, then compare that against a broad market where you’re winning maybe two per cent of what’s available. Ten per cent of a small, well-defined market beats two per cent of a big undefined one, and it’s a far easier ten per cent to win because there’s less competing for the same attention.

There’s a real limit, though, and it’s this. If you can’t name at least thirty or forty realistic buyers you could actually reach, you’ve gone too far. At that point you haven’t specialised, you’ve built a dependency on a handful of relationships.

A sensible test: could you fill a room with your ideal clients? Not a stadium. A room. If yes, the niche is viable. If you’d struggle to fill a table, widen one dimension.

What actually changes once you commit

Worth being straight about this, because the promise is usually oversold.

Your phone probably won’t ring more. Sometimes it rings less, particularly in the first few months, which is uncomfortable and is the point where most people reverse the decision.

What changes is the shape of the conversation.

You stop spending forty minutes on every call explaining what you do, how you work and why it costs what it costs. That call is half education and half audition, and it mostly disappears, because the person ringing has already worked out you’re the right fit before they picked up the phone.

The price conversation softens, not because you got better at defending your number but because they arrived expecting a specialist to cost more than a generalist. Perceived value is set before the proposal lands, which is a positioning outcome rather than a sales one.

Everything downstream gets easier as well. Your campaign work has one audience to speak to rather than four, which is why a niched business tends to get more out of the same marketing spend than the generalist sitting next to it.

And your referrals get specific, people stop saying “he’s good, give him a call” and start saying “if you’ve got this exact problem, call him.” The second one works while you’re asleep.

None of that is more lead, it’s less friction on the leads you already get, which is usually worth more.

One last thing, and it’s the part that stops people acting. You can still do the other work (nobody’s coming to check) narrowing is a decision about what you lead with in your messaging, not a rule about what you’re allowed to accept. Your website and your outbound point at one thing. Your invoices can say whatever they like.

FAQs

How do I choose a niche for my business?

Start with your own data rather than market research. Rank your last twenty to thirty jobs by profitability, how smoothly they ran, and which ones generated referrals, then look for what your top five or six have in common. The strongest niche is almost always one you’re already serving without having named it, because that’s where you already have proof, process and reputation.

What are the different ways to niche a business?

There are three: by industry (the sector you serve), by customer (the type of buyer, regardless of sector), or by service (narrowing what you do rather than who you do it for). Narrowing your service is the only one that doesn’t require turning away any type of client, which makes it the least risky place to start. You can combine two, but combining all three usually leaves the market too small to sustain you.

Will niching down mean turning away work?

No, and this is where most people get stuck. Niching is a decision about what you lead with in your marketing, not a restriction on what you accept. Your website, content and outbound point at one thing while you continue taking adjacent work as it comes. Nothing prevents you from doing a job outside your stated focus.

How narrow should my niche be?

Narrow enough to be specific, broad enough to sustain you. A workable test is whether you could name at least thirty or forty realistic buyers you could actually reach. If you could fill a room with ideal clients, the niche is viable. If you’d struggle to fill a table, widen one dimension. Most owners who think their niche is too small have made a counting error, tallying only the buyers they personally know about and comparing that to a broad market where they win a very small percentage anyway.

How long before niching starts working?

Expect months rather than weeks, and expect it to feel worse before it feels better. Enquiry volume often dips first while the quality improves, which is the point most businesses reverse the decision and lose the ground they’d gained. The compounding effect comes from holding one position long enough for the market to associate it with you.

Can I change my niche later?

Yes, though it costs you the recognition you’ve built, so it’s not a decision to revisit annually. If you’ve held a position for several years and the market has moved under you, changing makes sense. If you’re eighteen months in and getting impatient, that’s usually boredom rather than evidence, and they’re different problems.

My takeaways

Choosing a niche feels like a prediction. It’s not. It’s an observation.

Everything you need is already in your invoicing, your referrals and the jobs you quietly hope don’t come back. Most owners skip past all of it because it doesn’t feel like strategy. Strategy is meant to involve research and a big decision, not reading your own history back.

But your history is the only evidence you’ll ever have about where you’re measurably better than the alternatives. Everything else is a guess with a spreadsheet attached.

You’re not choosing a direction.

You’re admitting one you’ve been heading in for years.