The short answer
If only 100 to 1,000 companies could realistically buy from you, broad-reach marketing is the wrong model. Build a named list of those companies, work out who inside them influences the decision, use search to capture the demand that already exists, use paid social to create and reinforce demand among the rest, and use retargeting to stay visible across a long sales cycle. Judge it on how much of your target market knows you and how much qualified pipeline exists, not on how many leads came in this month.
Everything below is the detail behind that paragraph: how to size your market honestly, what each channel is genuinely for, how to set a budget from arithmetic rather than an agency rate card, and when the honest answer is that paid advertising is not your problem.
The one-page version
For UK specialist B2B firms, roughly 5 to 50 people, selling something technical or high-value into a market you could realistically write down on a list.
- 1Write the list. If you cannot name your buyer universe, that is the first job, before any budget is spent.
- 2Pick one offer for one buyer. A small market punishes vague, everything-for-everyone messaging harder than a big one.
- 3Own the search terms that exist, even if there are only forty a month. They are the cheapest customers you will get.
- 4Use paid social to be familiar to the people who are not searching yet. That is most of your market, most of the time.
- 5Retarget everyone who engages, for months. It is the cheapest media in the mix and the cycle is long.
- 6Report on accounts reached, accounts engaged and qualified pipeline. Lead counts will mislead you at this size.
The tiny-TAM test: how small is your market really?
gets thrown around as a fundraising word. For a specialist firm it is a practical number, and getting it right changes everything downstream. Count the companies that fit your specialism, your geography, your minimum size and your delivery capacity. Be strict. The number is usually far smaller than people expect, and far more useful.
| Buyer universe | The call | Why | Rough media |
|---|---|---|---|
| Under 100 accounts | Do not run broad campaigns | This is a sales problem with a marketing support act. Name every account, work them one to one, and use paid only to make sure your name is familiar before the call. Budget goes on research, outbound and a credible website, not on reach. | £0-£500/mo media |
| 100-500 accounts | Named-account paid, one offer | The sweet spot for precision. Upload the list, target three or four job titles inside it, run one clear offer plus retargeting. You can reach every relevant human every week for the price of one trade stand. | £750-£1,500/mo media |
| 500-2,000 accounts | Named-account paid plus search | Big enough that some buyers are actively searching. Add Google for the people already looking, keep LinkedIn for the ones who are not. Segment the list into tiers so your best-fit accounts get more weight. | £1,500-£3,000/mo media |
| 2,000-10,000 accounts | Segment and prioritise | You can no longer treat the market as one audience. Split by sector, size or trigger event, run a separate message per segment, and let performance decide where budget concentrates. Content starts to pull real weight here. | £3,000-£8,000/mo media |
| 10,000+ accounts | Conventional demand gen | At this size the usual playbook works: broad-ish targeting, category content, search coverage, efficiency metrics. You have enough volume for statistics to mean something. Most of this article stops applying. | £8,000+/mo media |
Spend bands are typical UK starting points from campaigns we run, not a formula. Contract value, competition and geography move them.
What each channel is actually for
Most channel arguments are really arguments about jobs. LinkedIn versus Google is not a fair fight, because they do different work. Here is the honest division of labour in a small market.
Only works when people search for what you sell. For genuinely niche products the monthly volume can be tiny, and that is fine: it is cheap, high intent, and you should own it. Do not judge it on volume.
The only channel where you choose the companies and the job titles. Expensive per click, cheap per relevant human reached. This is the engine for a small market.
The cheapest media you will buy. In a long cycle it is often what makes you the name they remember six months later.
Still works when it is genuinely researched and genuinely few. Works far better when the recipient has already seen your name for three months.
Unbeatable for depth, terrible for consistency. Two days of access, fifty weeks of silence. Best used as the peak of an always-on presence, not instead of one.
Slow, compounding, and increasingly about being quotable to AI assistants as well as ranking. Worth it if you are staying in the market for years.
The distinction that matters most is between and . In a niche market, the pool of people actively searching in any given month is tiny. Capture it, absolutely, because it is cheap. But if capture is your whole strategy, your growth is capped at the number of people who happened to have a problem this month.
Go deeper: how to build the full demand generation system around named accounts.
Search finds the people already looking. In a market of 500 companies, that is about four of them at any one time.
Stop counting leads. Count accounts.
This is the single biggest mental shift, and it is the one that makes small-market marketing make sense. When your universe is 500 companies, "we got 30 leads" tells you almost nothing. Thirty leads from thirty companies you do not want is a bad month. Four enquiries from four of your top fifty target accounts is an excellent one.
So change what you report on. The three numbers that matter:
- Account reach. How many of your named target companies saw you at all this quarter. If it is 40%, your problem is distribution, not creative.
- Account engagement. How many interacted: clicked, watched, visited the site, followed the page. LinkedIn reports this at company level, which is the closest thing to a scoreboard a long-cycle business gets.
- Qualified pipeline by account. Which named companies are actually in a conversation, and what they are worth.
This is what people mean by , stripped of the software and the ceremony. And yes, it works perfectly well for a 20-person firm. You do not need a platform. You need a list, a message and consistency.
Want to know how many of your target accounts have actually heard of you?
Book a 30 minute call. We will look at your market, who is really buying, and what it would take to be visible to all of them.
You are not selling to one person
In a high-value technical purchase, several people have to end up comfortable. Aiming everything at "the decision maker" is how campaigns arrive too late, after someone else has already written the requirement around a competitor's product.
Risk, reputation, whether this is a distraction. Rarely the first to notice you, usually the one who signs.
Whether it actually works. Specs, tolerances, integration, evidence. Often writes the requirement months before anyone else is involved.
Whether their day gets better or worse. Ignored by most marketing, and quietly capable of killing a deal.
Total cost, payback, what happens if it fails. Wants a number they can defend.
Process, terms, comparability. Arrives last and can only choose from the shortlist someone else wrote.
The practical version: target the whole group inside your named accounts, but do not send them the same message. The technical buyer wants evidence. Finance wants the cost of doing nothing. The MD wants to know you are a safe pair of hands. Three ads, one audience list, and you are in the conversation before procurement exists.
Once the market is sized, use the account-based marketing guide to tier the companies, map the buying group and decide how each account should progress.
Budgeting from your market size, not a rate card
"Spend £5,000 a month" is not advice, it is a default. The better question is arithmetic: what does it cost to put your name in front of every relevant human in your market, often enough to be remembered?
Worked example for a specialist supplier with 500 target companies
500 target companies x roughly 4 relevant people each = 2,000 people you need to reach.
You want each of them to see you around 4 times a month, so 8,000 impressions a month.
UK B2B LinkedIn CPMs commonly land somewhere in the £25 to £60 range depending on how narrow the audience is. Say £45.
8,000 impressions at £45 per thousand = around £360 a month in media to blanket the entire market.
Add creative testing, a retargeting layer and search coverage, and a realistic working budget is £1,000 to £1,500 a month.
Run this with your own numbers. The point is that reaching a small market properly is usually far cheaper than firms assume, and far more effective than spending three times as much on a broad audience.
Two things fall out of this. First, most niche firms are not underfunded, they are unfocused: the same budget aimed at 500 named companies rather than "UK manufacturing" performs completely differently. Second, once you have blanketed your market, extra budget does not buy much more reach. It buys better creative, more offers and faster testing. That is a different conversation from "increase spend".
Measuring when the cycle is a year long
Small-market B2B breaks standard analytics in a specific way. Someone sees you for four months, never clicks, then types your company name into Google when their supplier lets them down. Your analytics files that as direct traffic, your ads platform claims nothing, and the campaign that created the demand gets no credit at all.
- Put "how did you hear about us" on every enquiry form. One field. Cheapest and often most accurate attribution you will ever install.
- Track account engagement monthly. Which named companies are interacting, and is that list growing. On a long cycle this is your earliest honest signal that it is working.
- Capture the first touch and keep it. If you only record the last click, paid work will always look worse than it is.
- Report pipeline value, not lead count. Two conversations with the right plants can outweigh forty downloads.
- Give it two quarters before a verdict, while still expecting visible engagement and early enquiries long before that.
What this looks like in five real sectors
These are our own campaigns. Full write-ups sit on the case study pages if you want the detail.
Situation. Relied on expensive trade conferences. Cost per lead over £300, pipeline unpredictable.
What we did. Always-on engine across Google Search, LinkedIn and retargeting, aimed at plant managers and engineering buyers rather than waiting for them at events.
Result. 17 new clients worth £26k+ each. CPL from £300+ to £55. Over £1.3M revenue from under £250k of spend across five years.
Situation. Referrals and one cold caller. Two quiet months would put the year at risk.
What we did. A list of 1,800 IT decision makers at 20 to 200 staff firms within a 90 minute drive. Two angles, one simple landing page.
Result. 14 booked calls in 90 days at £71 each. 4 signed retainers worth around £94k of first-year contract value.
Situation. Needed clinical directors and funded founders. A tight, hard-to-reach group. Two leads a month.
What we did. Diagnostic quiz funnel, A/B tested LinkedIn Ads, thought leader ads timed around sector conferences.
Result. From 2 leads a month to 5 a week, with £25k+ average customer value. Now approaching capacity.
Situation. A brand new service. No existing demand, no track record, no audience.
What we did. LinkedIn to founders and finance directors for demand creation, Google on high-intent CFO terms for the demand that existed.
Result. 80+ leads in the first six weeks, several converting immediately, 7.5X ROI and growing on retainer.
Situation. Founder-led sales worked, inbound had flatlined, the founder was the bottleneck.
What we did. A built list of 2,400 qualifying firms, three creative angles tested, demo page rewritten to remove friction.
Result. 38 demo requests in Q1 at £142 each, 7 closed-won, 6.2X pipeline against ad spend.
When we tell people not to run LinkedIn Ads
We turn work away for these reasons regularly. If any of them describe you, spend the money elsewhere first.
- Your average deal is worth a few hundred pounds. The maths will not carry LinkedIn CPCs.
- Nobody can answer an enquiry within a day.
- You cannot say who you are for in one sentence.
- You need revenue this month. Paid social is not a cashflow fix.
- Your buyer is not really on LinkedIn: some trades, shop-floor roles and parts of the public sector.
- Your website would embarrass you in front of a technical buyer.
- Low deal value: Google Search only, plus a better offer.
- Slow response: fix the follow-up habit first. It is free.
- No positioning: one uncomfortable afternoon narrowing the offer.
- Need revenue now: existing clients and lapsed enquiries.
- Buyer not on LinkedIn: trade press, Google, distribution partners.
- Weak site: one strong page beats a full rebuild.
Paid distribution amplifies whatever you already have. If the offer is vague, you will buy a wider audience for a vague offer.
The decision tree
Five questions, in order. Stop at the first one you cannot answer well.
Already exhibiting at a sector show? The exhibitor list is a ready-made buyer universe. How to make an event last all year with LinkedIn.
Yes. Move on.
No. Build the list first. Best clients, trade bodies, exhibitor lists, distributor customers. Nothing else works without it.
Yes. Precision paid is affordable.
No. Google Search and referrals only. The maths will not carry paid social.
Yes. Own those terms. They are cheap and high intent, however low the volume.
No. Skip search almost entirely and put the money into demand creation.
Yes. You have a real advantage over faceless competitors.
No. It can still work, but expect to pay more for the same attention.
Yes. Start.
No. Wait until you can. A six-week test on a twelve-month cycle only produces a bad decision.
The first 90 days, if you are starting from nothing
- Weeks 1-2Size the market and write the list
Count the companies that genuinely fit. Build the named list from your best clients, trade bodies and exhibitor lists. Agree one offer for one buyer, and get tracking working before anything spends.
Output: A matched target account list and working measurement. - Weeks 3-6Become familiar inside those accounts
Two or three creatives built around one expensive problem, aimed at the whole buying group. Turn on search coverage for the terms that exist. Add retargeting on day one.
Output: Consistent presence across the market, not just the searchers. - Weeks 7-10Fix what happens after the click
One page that answers the question the ad raised, with evidence, and one obvious next step. Then check how quickly an enquiry actually gets a human reply.
Output: A credible page and a same-day response habit. - Weeks 11-13Read accounts, not lead counts
Which target companies engaged, which enquiries were real, what has entered pipeline and what it is worth. Then decide: scale the offer, change the angle, or widen the list.
Output: An evidence-based decision instead of a gut call.
Read next
The same playbook applied to named plants and technical buyers.
Beyond referrals, for accountants, consultancies and advisers.
What £1,000 to £2,000 a month realistically buys and how to split it.
Long cycles, technical buyers, and getting on the shortlist early.
Common questions
If your market is small, specific and does not know you exist
- An honest count of the companies that could realistically buy, and a named list of them.
- One offer, one buyer, aimed at the whole buying group inside those accounts.
- Search coverage for the demand that already exists, paid social for the demand that does not.
- Reporting on account reach, account engagement and qualified pipeline.
Starter £1,250/mo, Growth £1,725/mo, Scale £4,025+/mo. Ad spend is separate and stays in your own account.