Meta Ads for Small Businesses: The Complete Guide

Meta Ads look simple from the outside. Pick an audience, set a budget, write something, and wait for enquiries. That impression is the reason so many small businesses have spent a few hundred pounds, seen nothing, and concluded that Facebook and Instagram advertising does not work for their type of business.

Usually what did not work was the setup. This guide covers how Meta advertising actually functions now, what it is genuinely good at, where small budgets get wasted, and how to tell within a few weeks whether it is worth continuing.

What Meta Ads are actually good at

The fundamental difference between Meta and search advertising is intent. On Google, someone has typed a problem and is looking for a solution right now. On Facebook and Instagram, nobody opened the app to buy anything.

That makes Meta strong at creating demand rather than capturing it. It works well for visual products, for offers that make sense on sight, for services with an obvious trigger, for retargeting people who already visited your site, and for building an audience before a launch. It works less well for urgent, high-consideration services someone only wants at the exact moment they need them, where search advertising will usually produce a better cost per enquiry.

Knowing which side your business falls on prevents most of the disappointment. The comparison is covered in detail in Meta Ads versus Google Ads.

How the system decides who sees your ad

Meta’s delivery system is an auction, but you are not really bidding against competitors for an audience. You are asking the system to find people likely to take the action you asked for, at the lowest cost it can manage within your budget.

Two consequences follow. First, the objective you choose matters enormously, because it determines what the system optimises toward. Choosing traffic as your objective tells the system to find people who click, and it will find people who click and do nothing else. Choosing leads or purchases tells it to find people who complete that action, which is what you actually want.

Second, the system needs data to learn from. This is the learning phase: while an ad set is new, delivery is unstable and costs are unrepresentative. Meta documents the threshold at roughly fifty optimisation events per week for an ad set to leave it. An ad set generating five leads a week never exits, which is why small budgets split across many ad sets perform so badly.

Structure: fewer, larger ad sets

The instinct with a limited budget is to split it carefully: one ad set per audience, one per interest, one per age band. That instinct made sense years ago and now actively harms performance, because it divides your data into pieces too small for the system to learn from.

A better default for a small budget is a simple structure: one campaign, one or two ad sets, several creative variations inside them, broad targeting, and enough budget concentrated in one place to accumulate events. Meta’s broad and automated targeting options have improved to the point where handing over audience selection usually beats manual interest stacking. Where the two approaches each make sense is covered in Advantage+ versus manual campaigns.

Creative is the biggest lever you control

Once targeting is largely automated, the creative becomes the main thing determining performance. The system can only deliver what you give it, and the difference between a good and bad creative is far larger than the difference between a good and bad audience setting.

What tends to work is unglamorous. Something that looks like content rather than an advertisement. A clear, specific claim rather than a general one. Video that makes its point in the first two seconds, because that is how long you have. Real footage of the product, the work or the people, rather than polished stock. And enough variations to find out what resonates, rather than one perfect ad.

Plan for creative to be the ongoing work. Audiences do not fatigue nearly as fast as creative does, and the single most common reason a campaign that worked stops working is that the same people have now seen the same ad many times.

Creative also wears out. Our guide to spotting Meta ad fatigue explains how to tell when results are falling because of the ads themselves and how often to refresh them.

Tracking, or the campaign is invisible

Without conversion tracking, you cannot tell the system what success looks like, and it will optimise toward whatever proxy you gave it.

At minimum: the Meta pixel installed and firing, the specific conversion events your business cares about defined, the Conversions API set up server-side so that browser-level restrictions do not hide a share of your results, and domain verification completed. Businesses skipping this are not running a measurable campaign, they are buying impressions and guessing.

For most advertisers, that now means running the Conversions API alongside the Pixel. We explain what it fixes in Meta Conversions API: do you need it?

Budget, and what it buys

There is no universal minimum, but there is an arithmetic floor. If your cost per lead is likely to be twenty pounds and the system wants roughly fifty events a week to stabilise, a budget of ten pounds a day will never get there. It will sit permanently in learning, with unstable costs, and the conclusion will be that Meta does not work.

A more useful approach is to work out the realistic cost per result in your market, multiply by the volume needed for the system to learn, and either commit to that or test in a shorter, more concentrated burst rather than spreading a small amount thinly over months. The arithmetic is worked through in how much to spend on Meta Ads.

Where the money actually leaks

Four patterns account for most wasted spend. Optimising for the wrong objective, usually traffic instead of leads. Too many ad sets, so nothing exits learning. Sending traffic to a homepage rather than a page built for the offer, which loses most of the people you paid for. And judging results too early, then changing everything, which resets learning and guarantees instability.

The landing page issue is the most under-recognised. Ad accounts are frequently blamed for what is a website problem, covered in why landing pages fail to convert and in why Meta Ads get clicks but no conversions.

Lead generation campaigns have their own version of this problem, covered in why Meta lead ads bring in poor-quality leads, and service businesses often recover lost interest through retargeting people who already visited.

Judging whether it is working

Look at cost per result and whether that result is real, not at reach, impressions or engagement. Reach tells you the system spent your money. Cost per qualified enquiry tells you whether it was worth spending.

Give a campaign long enough to produce a meaningful number of results before judging it, and resist editing it daily. The sensible waiting period is covered in how long to run a Meta ad before judging it. Then compare the cost per enquiry against what that enquiry is worth to you, since a thirty pound lead is excellent for some businesses and unaffordable for others.

When to bring someone in

Meta rewards accounts that are managed with some discipline: proper tracking, stable structure, a steady supply of new creative, and the patience to let the system learn. Most small businesses can run a simple campaign themselves. Where outside help pays for itself is in diagnosis, because an underperforming account can be failing for reasons that look identical in the dashboard.

Our Meta Ads management starts by establishing whether the problem is the account, the creative or the page the traffic lands on, since fixing the wrong one is how budgets disappear.

Starting without wasting the first month

If you are beginning from nothing, do three things before spending: install tracking properly, build a page that matches the offer rather than pointing at the homepage, and decide what a result is worth to you so you can judge the outcome honestly.

If you are already spending and it is not working, tell us what you are running and what you are paying per enquiry. We will look at the structure, the creative and the page, and tell you which of the three is costing you the result.