Four categories get their own page here. Everything else lands in a bucket the industry calls professional services and treats as too small to bother with: urban design, surveying, town planning, executive coaching, specialist recruitment, architecture, the thing you have to explain twice at a barbecue. Search engines treated narrowness as a problem. AI does not, and that is a change worth ten minutes of your time.
For twenty years, being findable meant matching words people typed. If nobody typed your category, you did not exist, so specialists learned to describe themselves in broader terms than they wanted to. You became a design consultancy instead of a practice concerned with how a public space feels to walk through after dark.
Broad terms are also the most contested terms. So the specialist ended up competing on the generalist's ground, against firms with more budget, in language that described them less accurately. That was the trade, and everybody made it.
AI search removes the trade. People no longer type two words. They describe a situation in a sentence and ask who can help, and the model answers by finding whoever has published the most relevant substance on that exact question. There is no minimum search volume to clear. There is no bidding war for a phrase nobody else is using.
Which inverts the old disadvantage. A narrow category is one almost nobody has written about properly, so the bar to becoming the obviously citable answer is far lower than it is in a crowded one. The thing that made you hard to market is now the thing that makes you easy to name.
The catch, and it is a real one: this only works if the expertise is actually published somewhere a machine can read. An excellent practice that has never written down what it knows is invisible to this, exactly as it was invisible to the old way. The advantage is available. It is not automatic.
The narrower the specialism, the fewer people understand it well enough to refer you. So a handful of people who have actually seen your work carry the entire pipeline, and everything depends on them remembering you at the right moment.
Research across professional services firms puts numbers on how that behaves. It is American, so read it as a pattern rather than as a local statistic, but the pattern is unambiguous:
| How likely someone is to refer you | |
|---|---|
| Someone who was previously your client | 52.5% |
| Someone who worked alongside you | 43.4% |
| Someone who only knows you socially | 17.8% |
Hinge Research Institute with NACVA, Referral Marketing Study, n=354 professional services practitioners. United States sample.
The same study found that visible expertise accounts for 34.7 per cent of what drives a referral, and its absence accounts for 50.2 per cent of what kills one, more than the lack of a professional relationship and the lack of a social relationship put together. Attending networking events scored 3.5 per cent. Asking for referrals scored 2.6 per cent.
Read those two together and the conclusion is uncomfortable but useful. Referral is not really a relationship channel. It is an evidence channel that happens to travel through relationships. People refer you when they can point at something. When they cannot, the introduction quietly does not happen and you never learn it was considered.
Which is why "get out and network more" is such poor advice for a specialist. The room is small, you have probably already met everyone in it, and the thing that would actually move the number is publishing the expertise, not attending another breakfast.
Same machine as the four named verticals, built around whatever your category actually is. The work inside each stage changes. The shape does not.
Then Recommend feeds Attract, and the handful of people who could refer you becomes a body of evidence that refers you without being present.
In a small practice the growth system is a bigger share of the enterprise value than it is anywhere else, because there is less else.
Norde Homes is a Perth custom home builder specialising in passive solar design. A narrow category inside a broad industry, excellent at the work, and invisible to anyone who had not already heard of them. Which is precisely the position this page is about.
They rebranded in November 2025, the structured data went in over January and February 2026, and by April they were topping ChatGPT's recommendations for high-intent Perth queries. Google Gemini now names them a top-tier high-performance builder in Perth, alongside a firm trading since 1991. Reputation took the older firm three decades. It took Norde five months, because being citable and being long-established turn out to be different things.
Where it stands as at July 2026: roughly three qualified enquiries a week, and AI-assistant referral traffic up 140 per cent quarter on quarter.
Two things worth knowing rather than discovering later. AI answers vary between runs, so this is a position held rather than a ranking owned. And Norde builds houses, so what transfers to your field is the mechanism rather than the specifics.
The better evidence takes about a minute and is about you rather than me. Run the AI Scan. It performs a live web search and tells you whether your practice is findable and citable for the questions your buyers actually ask. Free, no pitch. If it comes back solid, you do not need me.
A narrow specialism is the easiest thing in the world to become the answer to, and the hardest thing in the world to be found for by accident.
Arnoud Gernaat, Growth RebelsRead the right-hand column properly. A wasted call costs us both the same half hour, and on this page the odds of a mismatch are higher than anywhere else on the site.
There is no way to price this without understanding what you do, who buys it and what one client is worth over the years you keep them. So the first step is a free thirty-minute call, and a good chunk of it will be me asking you to explain your work properly.
For a narrow practice the entry build is often the right answer, which is not what most agencies would tell you. If the search volume in your category is genuinely small, a full paid-demand engine is not the first thing you need. The authority work and the conversion layer usually are. We will work out which on the call.
Pricing is published on the site, because you should not have to sit through a call to find out what something costs.
Usually the opposite. A narrow category is one almost nobody has published about properly, so the bar to becoming the obviously citable answer is far lower than in a crowded one, and AI search does not require a minimum search volume the way keyword advertising does. The real limit is not category size, it is client value. If one client is worth well under fifteen thousand dollars over the relationship, the arithmetic stops working regardless of how findable you become.
Not yet, and I will not pretend otherwise on a first call. What I understand is the shape of the problem: expertise is the product, the buyer cannot assess quality in advance, the decision takes months, and growth currently depends on a small number of people remembering you. The specific knowledge has to come from you, which is why knowledge capture is part of the build rather than an add-on. If your field turns out to need regulatory knowledge I do not have, I will say so before you spend anything.
It depends on whether anyone is searching in your category yet, and for genuinely narrow specialisms sometimes they are not. Ads bring volume today where volume exists to buy. Where it does not, spending on them is burning money to prove a point. The authority work and the conversion layer are the parts that matter in a thin-volume category, and I would rather build those and leave the ads until the demand is demonstrably there.
I do. Growth Rebels is one specialist, not an agency. The person who scopes the system is the person who builds and runs it. The fair question that raises is capacity rather than size, and I answer it on the call: I take on a limited number of firms at a time, and I will tell you if you are not one of them right now.
In a small category the network is small too, and it runs out. Book a free thirty-minute call and we will work out whether being the obvious answer in your field is something worth building.
A straight conversation, not a pitch. No prepared deck.