- Why the "10% of revenue" rule is nonsense for most small businesses
- Work it out backwards from what a customer is worth to you
- What happened with my own client's budget
- A realistic monthly range by business size
- Don't forget the budget you're not putting into spend
- When to increase your budget, and when to pull it back
- A simple monthly checklist before you set your figure
- Frequently asked questions
- Where to check the details
The short version: most small businesses need a minimum of £1,000 to £1,500 a month before Facebook's algorithm has enough data to work, and if you can't stretch to that, you're better off putting the money into organic content until you can. Below that figure you're not really testing anything, you're just donating small amounts to Mark Zuckerberg while he learns your audience on your dime. The right number after that isn't a percentage of revenue, it's a number you work out backwards from what a customer is worth to you.
Why the "10% of revenue" rule is nonsense for most small businesses
You'll see this advice everywhere. Spend 5 to 10% of revenue on marketing, split half into paid ads. It sounds sensible until you try to apply it to a real business.
I worked with a client last year, a small accountancy practice in the Midlands turning over about £180,000 a year. Ten percent of revenue on marketing is £18,000, half of that on ads is £9,000 a year, or £750 a month. Fine on paper. Except that business had 40% margins and was already stretched on staff costs. £750 a month felt reckless to them, and pushing it on them because "the rule says so" would have been bad advice.
Compare that to a client selling a £2,000 online course. Same revenue-based logic gives a similar monthly figure, but that business can afford to spend £150 to acquire a customer worth £2,000 and still make good money. The accountancy firm's average client is worth £600 a year. Same budget rule, completely different reality underneath it.
The percentage-of-revenue approach ignores margin, customer value, and sales cycle. It's a rule of thumb that got repeated so often it started sounding like law. It isn't.
Work it out backwards from what a customer is worth to you
This is the calculation that matters, and it takes about ten minutes:
- What is one customer worth to you, on average, in the first year? (Not lifetime value, that's a different conversation. First year, real number.)
- What can you afford to pay to acquire that customer and still be profitable? Be honest, not optimistic.
- What's your current conversion rate from lead to customer? If you don't know, guess conservatively, 10 to 20% is typical for a warm lead in most service businesses.
- Divide your acceptable cost-per-acquisition by your conversion rate to get your acceptable cost-per-lead.
- Multiply that cost-per-lead by how many leads you need each month to hit your growth target.
Real example. A client running a small coaching business had a £1,500 average client value, could afford to pay £150 to acquire a customer, and converted about 1 in 5 leads into paying clients. That's a maximum cost-per-lead of £30. She wanted 15 new leads a month. That's £450 a month in ad spend, not a penny more, or she'd be paying more to acquire customers than they were worth to her in year one.
That's the whole method. No percentage of revenue involved. Just what a customer is worth, what you can pay for one, and how many you need.
What happened with my own client's budget
I'll give you the messier version because the tidy version above skips the part everyone wants to know: what happens in month one.
A client of mine, a small home renovation business, started at £1,200 a month across two campaigns. First two weeks were rough. Cost per lead came in at £68, way above the £30 target we'd set based on their numbers. I got a panicked call asking if we should pull the plug.
We didn't. Facebook was still in what's called the learning phase, where the algorithm is testing audiences and placements before it settles into delivering efficiently. That phase typically needs around 50 conversion events before performance stabilises. At £68 a lead with a target of maybe 20 leads a month, we weren't going to hit 50 events until we'd spent close to £1,400, which was more than the monthly budget on its own.
We held the budget steady rather than cutting it, because cutting spend mid-learning-phase resets the clock and you end up paying the "learning tax" twice. By week five, cost per lead had dropped to £34. By week eight it was sitting at £27, under target. That business went on to close four kitchen renovation jobs worth roughly £8,000 each from that campaign in its first quarter.
The uncomfortable bit nobody likes admitting is this: if that client had only been able to afford £400 a month, we'd never have gotten past the expensive early weeks. The budget would have run out before the algorithm found its feet, and on paper it would have "proven" Facebook ads don't work for that business. They'd have been wrong, but they'd have believed it. This is the real reason so many small businesses tell you Facebook ads didn't work for them, they under-budgeted for the learning phase and quit right before it would have paid off.
A realistic monthly range by business size
With all the caveats above about doing your own maths, here's roughly what I see working in practice across UK small businesses in 2026:
- Solo or very early-stage business: £500 to £800 a month, one campaign, one clear offer. Enough to get some data, not enough to run multiple tests at once.
- Established small business, £100k to £500k turnover: £1,000 to £3,000 a month. This is where most of my clients sit, and where you can realistically run two or three ad sets and let the weaker ones die.
- Growth-stage business actively scaling: £3,000 to £10,000 a month. At this level you're not just testing, you're funding proven winners and can afford genuine A/B testing across creative and audiences.
- E-commerce with tight margins: often needs more than service businesses, sometimes £5,000-plus, because average order values are lower and you need volume to make the maths work.
None of these figures include the cost of someone managing the account, or making the ads themselves. If you're paying an agency or freelancer to run campaigns, expect a management fee on top, typically 10 to 20% of ad spend or a flat monthly rate, and my full breakdown on how much to charge for marketing services covers what's reasonable to pay there.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Don't forget the budget you're not putting into spend
A mistake I see constantly: a business sets aside £1,000 a month for ads and spends none of the time or money on the actual creative. The ad itself, the image, the video, the words on the page, does more to affect your cost per lead than almost any budget decision. A weak ad at £2,000 a month will still underperform a strong one at £500.
Before you touch your ad account, get clear on what good Facebook ads look like, and spend proper time on creatives that convert rather than assuming the algorithm will save a weak offer. If you're completely new to the platform, it's worth reading a straightforward beginner's guide to Facebook ads before you set a budget at all, because half of overspending comes from not understanding what you're paying for in the first place.
When to increase your budget, and when to pull it back
Once you're past the initial learning phase, here's what I watch for:
- Increase spend when your cost per lead has been stable or falling for two to three weeks straight and you're converting those leads into paying customers at your normal rate. Increase by 20% at a time, not double, otherwise you trigger a fresh mini learning phase.
- Hold steady when performance is inconsistent week to week. This usually means your audience or creative needs work, not your budget.
- Pull back only after giving a campaign a genuine chance, at least two to three weeks and enough spend to cover 50 or so conversion events. Pulling back sooner almost always means you've paid for the expensive learning phase and left before getting the payoff.
Where you're running these ads matters too. Facebook's own placements (feed, Stories, Reels, Marketplace, Audience Network) all behave differently on cost, and I've written a full breakdown of where to post Facebook ads for the best results if you want to get more specific than "let Facebook auto-optimise it."
A simple monthly checklist before you set your figure
- Work out what one customer is worth to you in year one, honestly.
- Decide the maximum you can pay to acquire that customer and stay profitable.
- Divide by your realistic lead-to-customer conversion rate to get your target cost per lead.
- Multiply by how many new leads you need this month.
- Add at least 20% on top for the learning phase in month one.
- Don't touch the budget for at least two to three weeks once it's live.
Get those numbers down on paper before you open Ads Manager. It takes ten minutes and it will save you from either wildly overspending or quitting three weeks before the campaign was about to work.
Related reading: facebook ads price.
If you write on this topic, see Write for Us Advertising: Advertising Blog Guest Post Guidelines.
Frequently asked questions
What's the minimum I should spend on Facebook ads per month?
Realistically, £500 is the floor for getting any usable data, and £1,000 to £1,500 is where most small businesses start seeing the algorithm settle into consistent, efficient delivery. Below £500 a month, you're usually spread too thin to learn much before the budget runs out.
Should Facebook ad budget be a percentage of my revenue?
Not really. A percentage of revenue ignores your margins and what a customer is worth to you. Two businesses with identical revenue can have completely different budgets that make sense, depending on average order value and profit margin. Work out your acceptable cost per lead first, then set the budget from there.
Why did my Facebook ad budget seem to underperform in the first couple of weeks?
This is almost always the learning phase, where Facebook's algorithm needs roughly 50 conversion events before it optimises delivery. Costs are typically higher in this window and drop once the account has enough data. Cutting the budget during this stage often resets the process rather than fixing it.
How much extra should I budget if I'm hiring someone to manage my Facebook ads?
Expect a management fee on top of ad spend, usually 10 to 20% of spend or a flat monthly rate depending on the size of the account. On a £1,000 a month ad budget, that's typically an additional £100 to £250 a month for someone experienced enough to improve performance rather than just log in and check numbers.
