paid ads · 12 august 2026 · 7 min read

how much should a small business spend on facebook ads?

Every guide answers "how much should a small business spend on Facebook ads" with the same shrug — ten dollars a day, maybe fifty, it depends. Meta has actually published the number that matters. It isn't a budget. It's the divisor you build one from, and once you run the sum most small budgets stop being small and start being below the floor — which is a different problem with a different fix.

This post is the arithmetic. It uses Meta's own documentation for the rule, a published benchmark for the input, and honest division for the rest. You'll finish it with a figure for your business rather than someone else's rule of thumb.

Fair warning: we sell Meta ads as a monthly add-on, so we have a stake in you advertising. That cuts both ways here — a decent part of this post argues that some businesses shouldn't run lead campaigns yet. Weight our opinion accordingly.

the only budget rule meta actually publishes

Meta doesn't publish a recommended daily spend, and the "minimum budget" warning in Ads Manager is calculated per ad set rather than printed as a universal figure — their budget is too low page literally renders it as a placeholder. So there's no official "$20/day" to find, however many blog posts imply otherwise.

What Meta does publish is the delivery rule, and it's the one that governs everything. Every new ad set enters a learning phase while the system works out how to deliver it. From Meta's own about the learning phase page:

"ad sets exit the learning phase as soon as they can deliver stably. This usually occurs after about 50 results in the week after the ad set's last significant edit."

Fifty results. In a week. Per ad set.

A "result" is whatever you told Meta to optimise for — a purchase, a lead, a landing page view. Miss that bar and the ad set doesn't fail loudly; it sits in a status Meta calls learning limited, delivering unstably at a higher cost per action, indefinitely. Meta is blunt about the window itself: during the learning phase "performance is less stable, so your results aren't necessarily indicative of future performance."

That's the whole trap. A budget too small to produce 50 results a week doesn't buy you a smaller version of the campaign. It buys you a permanently unoptimised one — and then you judge Facebook ads on those numbers and conclude they don't work for your business.

the sum: your cost per result × 50 ÷ 7

Turn the rule around and it becomes a formula. If an ad set needs roughly 50 results a week to stabilise, then:

minimum viable daily budget = (your cost per result × 50) ÷ 7

You need one input: what a result costs you. If you've run ads before, use your own number — it's sitting in Ads Manager and it beats every benchmark on the internet. If you haven't, you need a placeholder to start with.

The most usable published set comes from LocaliQ's Facebook advertising benchmarks (last updated October 2025). Across industries they report an average cost per lead of $27.66 for the leads objective, up from $22.87 the year before, and an average cost per click of $0.70 for traffic campaigns and $1.92 for leads campaigns. Read that as a vendor's aggregate of mostly US advertisers, not a law of physics — the spread is enormous, from $3.16 per lead in restaurants and food to $76.71 for dentists.

Now do the division. At the $27.66 average, 50 leads a week costs $1,383, which is roughly $198 a day. At the dentist end it's about $548 a day. At the restaurant end, $23.

Those are our sums from LocaliQ's figures, not numbers they publish — but they explain something that no amount of creative testing will. A business running a $15/day lead campaign in a $27 CPL industry is buying about four leads a week. Meta needs fifty. The campaign was never going to leave the learning phase, and no hook, headline or audience fixes an arithmetic problem.

Where p.a. fits: we build the ad creative, the copy variants and the A/B and budget plan — and we'll tell you before you spend anything whether your numbers clear the floor above. If they don't, the honest recommendation is organic content until they do.
See what the Meta ads add-on includes →

when you can't afford the answer, change the event — not the budget

Most small businesses run this sum and don't like the result. The instinct is to shave the budget and hope. The fix is to change what you're counting.

The formula has two variables and you can only move one of them. If $198/day is out of reach, optimise for a cheaper result:

• Optimise for link clicks or landing page views instead of leads. At LocaliQ's $1.92 average cost per click for lead campaigns, 50 results a week is about $96 — roughly $14 a day. Same rule, same fifty, one-fourteenth of the money, because you picked an event that happens far more often.
• Run one ad set, not five. Splitting $30/day across four audiences gives each of them a quarter of the results and none of them fifty. Meta says this outright in its own best practices — "avoid high ad volumes," because the system "learns less about each ad and ad set" — and recommends combining similar ad sets so you "combine learnings" too.
• Stop editing. Significant edits reset the learning phase, and Meta's first listed best practice is to wait until an ad set is out of it before touching anything. Budget anxiety is the most common cause of a permanent learning phase: change the budget on day three, restart the clock, repeat forever.

The trade is real and worth naming. Optimising for clicks means Meta finds you people who click, not people who buy — you've bought a cheaper, weaker signal. That's an acceptable stage-one position while you gather data and build an audience. It is not a substitute for conversion optimisation once you can afford it.

And if neither version fits your budget, that's information, not failure. Paid ads are one channel among several, and the case for starting elsewhere is made properly in our no-waste guide to Meta ads for small business.

your daily budget is not a daily cap

One more thing worth knowing before you set a number, because it causes a specific kind of panic on day two.

A daily budget is a weekly target, not a hard daily ceiling. Per Meta's about daily budgets page, you're "setting a daily spending target over the course of a calendar week" running Sunday to Saturday, and on days with better opportunities Meta "may spend up to 75% over your daily budget." The guardrails: spending won't exceed 175% of your daily budget on any day, or seven times your daily budget across the week. With ad set budget sharing turned on, a single day can reach 210%.

So a $20/day budget is really a commitment to $140 a week, in which any given day might bill $35. Check the week, not the day. People who watch the daily number see $35, panic, cut the budget — and reset the learning phase they were paying to get through.

the floor rises every year

The last input is time. Meta's advertising costs aren't static, and the direction is consistent.

In its second quarter 2026 results (published 29 July 2026), Meta reported revenue of $60.8 billion, with ad impressions across its Family of Apps up 14% year over year and the average price per ad up 12% year over year.

More ads sold, each one dearer. That 12% compounds against your floor: a budget that cleared 50 results a week last year may not clear it this year, without anything about your business changing. It's the quiet reason campaigns "stop working" — and the reason to re-run this sum annually rather than setting a budget once and leaving it. Tracking that properly is its own discipline, which we cover in how to know if your marketing is working.

so how much should a small business spend on facebook ads?

Enough to buy about fifty results a week of whatever you've asked Meta to optimise for — and if that's out of reach, enough to buy fifty of something cheaper. That's the answer, and it's a different number for a dentist than for a café, which is exactly why nobody can hand you a figure.

Work it out honestly. Take your real cost per result, multiply by 50, divide by seven. If the number is affordable, commit to it for a full month without edits and judge it on week four. If it isn't, optimise for a cheaper event or spend the money on content until your economics change. What doesn't work is funding a conversion campaign at a quarter of its floor and calling the result a verdict on Facebook ads.

The budget question was never about how much you can afford to spend. It's about whether what you can spend clears the bar Meta already told you about.