Not every inquiry becomes a customer, whether it arrives as a form, a message or a call. The sums therefore start from your own sales record rather than from anything Meta reports.
A budget from what a customer is worth#
- What you would spend to win a customer Start from what a typical job or order leaves you after your costs, which is not the price, and decide how much of it you will give up to win a new customer. A plumber who keeps $300 from a typical job might spend $100 of it to get the job.
- The most a lead may cost Count every inquiry, including the people who never reply, and see how many out of ten become customers. Multiply what you would spend by that share.
- The budget Decide how many new customers you want a month, work back to the leads you need, multiply by the most a lead may cost, and divide the month into days.
A worked sum
| The most a lead may cost: $100 × 2 ÷ 10 | $20 |
| Leads needed: 5 × 10 ÷ 2 | 25 a month |
| Monthly budget: 25 × $20 | $500 |
| Daily budget: $500 ÷ 30 | about $17 a day |
This is the most the plumber should spend to win those five jobs. If leads cost less than $20, you need less budget. If they cost more, each job costs more than the $100 set aside, and a bigger budget only buys more of those jobs.
The cost per lead calculator#
Type in your own figures, in any currency. The answer is a ceiling, and above it each new customer costs more than you set aside to win one.
Work out your own numbers
$17a day
That is $500 a month, for 25 leads at up to $20 each.
At the ceiling, everything you type in the first box is spent on winning the customer. Type the whole of what a job leaves you and the ceiling is where the ads break even; type less and you keep the rest. The plumber typed $100 of the $300 a job leaves, keeping $200 of every job won at $20 a lead; the whole $300 would have put the break-even at $60 a lead.
| If you spend this share of a customer’s worth | What it means |
|---|---|
| All of it | You break even. Worth it only if customers come back or refer others |
| Half | You keep half of what the customer is worth |
| A third | A cautious plan with room for leads that cost more than expected |
A customer who comes back is worth more than one job. If one in three of the plumber’s customers calls again within the year for a similar job, a new customer leaves about $400 that year instead of $300. A third of that is about $133, and at two inquiries in ten a lead may then cost up to about $27.
Whether ads can pay at your figures#
In Verizon’s 2025 survey of 600 US small and midsize business owners and decision makers, 82% said they use Facebook to promote products and connect with customers. It does not say how many pay for ads or whether the ads paid, and no survey Madder found of what Facebook ads return for small businesses gave a method that could be checked.
Published 2026 figures put a lead at about $12 to $62 by industry in the US, with $27.39 typical, and at about £8 to £40 in the UK. With two inquiries in ten becoming customers, this is how far a customer’s worth goes against them:
| A customer worth, and the most a lead may cost | Against the published $12 to $62, or £8 to £40 |
|---|---|
| $25 or £25: $5 or £5 a lead | Not worth it from leads alone. You need repeat orders |
| $75 or £75: $15 or £15 a lead | Tight. It breaks even only if your leads come in at the cheap end |
| $150 or £150: $30 or £30 a lead | Break-even at about the US typical figure. Profit only if your leads cost less |
| $500 or £500: $100 or £100 a lead | Worth testing. There is room for mistakes |
| $2,000 or £2,000: $400 or £400 a lead | Worth it even at several times the average lead cost |
A store runs the same sums with an order in place of an inquiry; what an order costs in ads, by kind of store, is in Facebook and Instagram ads for a small online store.
When Facebook ads are not worth it#
When a one-time customer is worth under about $60, or £40, the sum does not break even, even at the cheapest published lead costs. Ads also lose money when nobody can answer an inquiry the same day, because you pay for every lead, including the ones that have gone elsewhere by the time you call. They cannot rescue a page or an offer that does not sell either. Ads send more people to the same page, and if visitors from elsewhere do not buy, there is no sign that paid visitors will.
Ads are a poor fit, too, if you cannot leave the money in for a month, because a small budget takes weeks to show a pattern. If you need customers this week, call past customers first. A business that sells to other businesses in a narrow trade may find its buyers faster elsewhere.
What guides suggest spending#
These ranges were published in 2026, and they disagree, which is the reason to do your own sums first.
| Source and date | What it suggests |
|---|---|
| 99 Dollar Social, US, updated February 2026 | $10 to $20 a day, about $300 to $600 a month, to begin testing |
| Adwave, US, July 2026 | $1,000 to $1,500 a month as a common starting point, and $20 to $50 a day per campaign |
| Whito, UK, updated September 2026 | £300 a month (£10 a day) as the minimum for meaningful data, £500 a month as right for most small businesses, and £100 a month to test one audience with one ad and nothing more |
| Startups.co.uk, UK, updated February 2026 | £176 to £528 a month as typical for a business aiming at 10 to 30 leads a month |
In WebFX’s 2026 survey of more than 250 US marketers, reported on its pricing page, 63% said their business spends $500 a month or less on Facebook ads. Meta itself publishes no single minimum. It says minimums vary by country and by the result you ask for, and Ads Manager warns you when a budget is below one.
What a fair test costs#
Four weeks at a steady budget make a fair test, which at the guides’ figures means about $300 to $600, or £300 in the UK. At published US lead costs, $300 buys roughly 5 to 25 inquiries, and at UK costs £300 buys roughly 7 to 37. That shows whether your ceiling is in reach, though a month is still a small sample. Whito’s UK guide also suggests £140 for a two-week test at £10 a day, without stating the data behind it. A UK test also pays Meta’s location fee and VAT.
Ten dollars a day is enough to start. At $20 a lead it brings about 15 leads a month, so one good or bad week changes the picture, and a small budget takes longer to settle.
Set the test up so it can answer the question:
- One product or one service The one with the best margin.
- One page to land on Checked on a phone.
- A way to count inquiries or orders Clicks alone tell you little.
- A fixed daily budget for four weeks With no edits in the first week.
- A number written down in advance The most a lead may cost.
At the end, divide what you spent by the customers you won. If that is below what you set aside to win a customer, raise the budget slowly. If it is far above, stop and fix the offer before you try again.
Madder and your budget#
Madder writes the daily budget you choose into the plan you approve, before anything exists on Meta. In most currencies Meta stops a campaign Madder creates once it has spent two weeks of its daily budget (about $100 at the least), so a four-week test needs that limit raised in Ads Manager before the second week ends. Madder does not bid to a cost per lead or predict one. It shows what your leads cost, counted from the actions you chose, to set against your ceiling. See Daily budgets and spending limits.
Questions owners ask#
What is the minimum budget for Facebook ads?
Meta publishes no single figure; it says minimums vary by country and by the result you ask for. Guides published in 2026 put a workable start at $10 to $20 a day in the US (99 Dollar Social) and about £10 a day in the UK (Whito). Less can run, but it produces too few results to learn from.
Are Facebook ads worth it for a small business?
They are when a customer is worth more than the leads it takes to win one. With published lead costs of about $12 to $62 in the US and £8 to £40 in the UK, and two inquiries in ten becoming customers, a customer needs to be worth roughly $60 to $310, or £40 to £200, for the ads to break even.
How do I calculate the most I can pay for a lead?
Multiply what you would spend to win a new customer by the share of leads that become customers. Spending up to $100 a customer, with two inquiries in ten won, a lead may cost $20 at most; above that, each customer costs more than you meant to spend.
Should I count repeat business in what a customer is worth?
Yes, if customers do come back, and only for a period you can count on, such as a year.
How much do I need to spend to know if Facebook ads work?
About $300 to $600, or £300 in the UK, over four weeks. A smaller sum gives too few results to judge.