Facebook & Instagram Ads for UK Small Businesses: The Honest Starter Guide (2026)

Quick answer

Meta ads (Facebook + Instagram) remain the cheapest way for a UK small business to put a message in front of a precisely chosen local or national audience, but the platform is engineered to let beginners spend badly. The honest starter numbers for 2026: expect UK CPMs of £5 to £15 (cost per thousand views), £0.30 to £1.50 per link click for decent creative, and a realistic test budget of £10 to £20/day for 4 to 6 weeks before judging anything. The three beginner-savers: never use the Boost button as a strategy, always send traffic to a page built to convert, and judge on enquiries, not reach, likes or “engagement”.

Before you spend a pound

• Have somewhere for the click to land. Ads amplify whatever they point at. A slow homepage with no clear action wastes every click; a focused landing page with one offer and one form is the difference between ads “not working” and working.
• Install the tracking first. The Meta pixel (plus Conversions API if your setup supports it) must be live before the first campaign, or you’re flying blind and the algorithm optimises on nothing. UK consent rules apply: your cookie banner needs to cover it.
• Know your numbers. If an average customer is worth £300 and you close half of enquiries, a £20 enquiry is excellent and a £100 one is a problem. Without this arithmetic, no dashboard can tell you whether ads are working.

Realistic UK benchmarks (2026)

Sector, offer and creative move these enormously. Treat them as orientation, not targets.

The setup that works for small budgets

1. One campaign, one goal. Leads or sales, not “awareness”. Small budgets can’t afford brand campaigns; leave those to brands with millions.
2. Broad-ish targeting, strong creative. 2026 Meta rewards letting the algorithm find buyers within a broad audience; the ad creative is the targeting now. Hyper-narrow interest stacking mostly raises costs.
3. Creative that looks native: real photos and phone video of actual work outperform polished stock: the same principle as organic, where consistent branded templates keep it recognisably yours. Lead with the customer’s problem in the first line, show the work, end with one clear action.
4. 3 to 5 ad variations, let Meta rotate. Refresh creative every 3 to 4 weeks; fatigue is the silent budget killer.
5. Retargeting from day one: website visitors and engagers are your cheapest conversions, often a third of cold-audience cost.

If your first ads are refused, our guide to why Facebook rejects ads helps you fix them quickly.

Meta vs Google: which first?

They answer different questions. Google captures people already searching for what you do: highest intent, costlier clicks (see Google Ads costs). Meta creates demand by interrupting people who weren’t looking. Cheaper attention, colder audience. Rule of thumb: if people search for your service (“emergency electrician near me”), Google first; if your offer is visual, impulse-adjacent or new to people (food, fitness, décor, events), Meta first. The full sequencing logic sits in SEO vs Google Ads, and either way, paid traffic belongs after the website converts, not before, per website vs social media. If running it yourself stops being viable, management costs covers what done-for-you actually buys.

Your first campaign, step by step

The concrete version, for a local service business starting from zero:

1. Business Manager and pixel first: create the Meta Business account, install the pixel on your site, and confirm it fires (the Meta Pixel Helper browser extension shows this in seconds).
2. Campaign objective: Leads or Sales, never Awareness or Engagement at small budgets.
3. One ad set, radius targeting: your service area plus 10 miles, broad age range, no interest stacking. Let the algorithm find buyers.
4. Three ads, same offer, different creative: one before/after photo, one short phone video of a real job, one customer-review graphic. First line of copy states the customer’s problem; last line says exactly what to do.
5. Budget £15/day, leave it alone for two weeks. Editing daily restarts the learning phase and burns the budget you’re trying to protect.
6. Check cost per lead weekly, not hourly: kill only the ad that’s clearly losing after 2,000+ impressions; feed the winner.
7. Answer leads within minutes, not days. Ad leads go cold at extraordinary speed; the firms that call back within five minutes convert multiples of those who reply next morning.

Six weeks of this produces the number that matters, your cost per enquiry, and a fed pixel that makes every later campaign cheaper.

Diagnosing a campaign that isn’t working

“Meta ads don’t work for my business” is almost always a specific, locatable failure that the account itself will tell you about, provided you read the metrics in the right order: impressions, then click rate, then landing page, then enquiry quality. Work down the funnel and stop at the first stage that is out of range.

The discipline that makes this work is changing one thing at a time and giving each change a fortnight. Small budgets gather data slowly, and a campaign edited three times a week never accumulates enough signal to be judged at all. If you would rather not diagnose creative failures repeatedly, note that most of them are production problems rather than strategy problems: a steady supply of on-brand, native-looking assets from proper social media design removes the most common bottleneck, which is having nothing new to test.

If you find yourself locked out of your business social accounts, there are four recovery routes.

What £500 a month actually buys

Abstract benchmarks are hard to plan against, so here is a worked model for a domestic installer running £16 a day for a month. Assume a £9 CPM, a 1.2% click-through rate and a landing page converting 6% of clicks, all within the ranges above but chosen for illustration, not measured.

• £480 spend at £9 CPM buys roughly 53,000 impressions.
• At 1.2% CTR, that is about 640 link clicks, or £0.75 each.
• At 6% landing-page conversion, that is roughly 38 enquiries, at about £12.60 each.
• At a 25% close rate (realistic for cold paid leads, well below what referrals convert at), that is 9 or 10 jobs.

If the average job is worth £400, the month returns around £3,800 on £480, which is a healthy business. If the average job is worth £120, the same performance is roughly break-even before your own time, and no amount of optimisation rescues it: the problem is the job value, not the advertising. Run this arithmetic before the first campaign rather than after the third, and rerun it with your real numbers at the end of month two. The wider question of how much a business at your stage should be committing to paid channels in total is worked through in the guide to how much a UK small business should spend on marketing, and the answer for most is a smaller paid budget spent for longer.

One structural point the model makes clear: the cheapest improvement is almost never in the ad account. Lifting landing-page conversion from 6% to 9% adds nineteen enquiries at no extra spend, which no bidding change will match, which is why the destination deserves proper attention and why what a landing page costs in the UK is the more useful number to research before increasing the daily budget.

The UK rules that apply to paid social

Paid social is advertising, and the advertising codes apply in full, including to the parts small businesses treat as informal. Four areas produce most of the trouble.

1. Substantiation. “Best in the county”, “guaranteed cheapest”, “trusted by thousands” all require evidence you could produce on request. Prices must be genuinely available, VAT treatment must be clear where consumers are the audience, and “from £X” needs real jobs at £X.
2. The pixel and consent. Tracking that stores or accesses information on a visitor’s device needs a lawful basis under the UK’s privacy and electronic communications rules, which in practice means a consent banner that actually gates the pixel rather than one that fires it regardless. A banner that reports consent it does not enforce is worse than none, because it documents the problem.
3. Special categories. Meta restricts targeting for ads relating to housing, employment, credit and certain social issues; declaring the category is mandatory and mis-declaring risks account-level action. Age-restricted products carry additional placement obligations under the advertising codes.
4. Finance offers. This one catches trades and installers regularly. Advertising “spread the cost” or interest-free payment options is a financial promotion, and communicating one in the course of business generally requires FCA authorisation or approval by an authorised firm. Mentioning that a finance provider is available and linking to them is a very different act from promoting terms yourself, and the difference is worth checking with your finance partner before the ad goes live.

None of this is onerous once set up, and it is all considerably cheaper than an upheld complaint or a disabled ad account with no route back.

Money most beginners waste, in rough order of size

Boosting posts because the button is there remains the single largest leak: it optimises for engagement rather than enquiries and produces likes from people who will never buy. Running before the destination is ready is second: the ad account is judged on the whole chain, and the chain fails wherever it is weakest. Third is stopping too early: a fortnight is not a test, and a campaign killed in week two after a slow start has usually just paid the learning cost without collecting any of the benefit.

Then there are the quieter ones. Sending every click to the homepage, where a visitor who arrived for one thing is offered nine. Ignoring retargeting for months, so the cheapest available audience (people who already visited) is never spoken to again. Judging performance on the platform’s reported conversions alone rather than counting actual enquiries in the inbox, which is where the honest method described in measuring marketing ROI without a data team earns its keep. And answering leads slowly: a form completed at 8pm and replied to at 11am the next day is competing against whoever answered at 8.05pm, and the ad spend that generated it has already been converted by somebody else.

If paid social is on the list too, our guide to Meta ads management costs sets out freelancer, agency, in-house and flat-fee prices side by side, with the ad budget kept separate.

Sources & Further Reading

• Meta Ads Guidance, Meta
• Adults' Media Use & Attitudes, Ofcom
• Marketing & Privacy Rules, Information Commissioner's Office