£500 or $600 a month is about £16 or $20 a day. That is enough to find out whether Facebook and Instagram can bring you enquiries. It is not enough for five campaigns, ten audiences and a retargeting funnel. Here is how to set it up so the money teaches you something.
What the budget buys
Divide the monthly figure by 30.4, the average number of days in a month: £500 is about £16.45 a day and $600 about $19.75. Meta treats a daily budget as an average across the week, so some days it spends more and some days less.
Nobody can tell you in advance what a lead will cost. It depends on your trade, your area, your offer and your ads, and it varies a lot between businesses. What you can do is the arithmetic, then decide which row you could live with.
| Cost per lead | Leads from £500 | Cost per lead | Leads from $600 |
|---|---|---|---|
| £10 | 50 | $12 | 50 |
| £25 | 20 | $30 | 20 |
| £50 | 10 | $60 | 10 |
| £100 | 5 | $120 | 5 |
Arithmetic, not a forecast.
Then work backwards from a job. If a typical job is worth £2,000 and you win one enquiry in five, ten leads a month at £50 each is two jobs from £500. Twenty leads that nobody answers is none. The numbers only matter once you know what happens after the lead arrives.
Keep the structure simple
Meta’s delivery system learns who responds to your ads. An ad set typically needs around 50 optimisation events in the seven days after its last significant edit to finish learning. At about £115 or $140 a week, that would mean a lead for £2.30 or less. Most service businesses will not get there, and the ad set will show Learning limited.
That label is not a penalty. It tells you the system has little data to work with. The answer on a small budget is not to fight it, but to avoid making it worse by splitting the money. A sensible starting shape:
- One campaign with the Leads objective (or Sales if you sell online).
- One ad set, so every result feeds the same learning.
- Three to five ads in that ad set, in more than one format.
Then leave it alone for a week at a time. Changes to targeting, the optimisation event, budget and adding ads can all count as significant edits that restart learning. Batch your changes and make them together, rather than a tweak every other day.
Audiences: Advantage+ or detailed targeting
With Advantage+ audience, Meta finds the people most likely to respond. You can give it suggestions (age, interests, a customer list) and it starts there before looking wider. You can also set hard limits it will not go past, such as location, minimum age and language.
For a local service business, location is the control that matters most. Set it to the area you actually serve, and check whether it is set to people who live there or also people passing through. Beyond that:
- Start broad. Advantage+ audience with your location limits and, if you like, a suggestion or two. On a small budget, narrow interest audiences typically cost more per person reached.
- Use your customer list. Upload past customers as a suggestion, or exclude them if you only want new ones.
- Test narrow targeting later. If lead quality is poor after a month, try a second ad set with detailed targeting then, not on day one.
Creative does the targeting now
With a broad audience, the ad itself decides who stops scrolling. Say who it is for and where in the first line and the image: “Boiler servicing in your town, booked this week” does more work than a clever headline.
- Keep three to five ads live. Fewer and the system has little to choose from. More and a small budget spreads too thin.
- Add two or three new ads a month and retire the ones that have stopped working. Add them in one batch.
- Mix formats. A photo of real work, a short phone-filmed video, a before and after, a customer’s words (with permission).
- Watch frequency. In a small area, the same people see your ads again and again. If frequency keeps rising while results fall, the ads are tired.
Lead forms or your website
Meta’s instant forms open inside Facebook or Instagram with the person’s details filled in. Your website asks them to leave the app. Each has a place.
| Instant form | Your website | |
|---|---|---|
| Lead volume | Typically higher | Typically lower |
| Lead quality | Mixed; some people forget they filled it in | Usually stronger intent |
| Needs | A fast way to get leads out and call them | A quick mobile page and working tracking |
| Good for | Slow or untracked websites | Sites that already convert visitors |
If you use a form, pick the Higher intent type, which adds a review step before submitting, and add one or two qualifying questions such as postcode (or ZIP code) and timescale. Then call quickly. A form lead cools within hours, and a lead nobody answers costs the same as one that books.
If you send people to your website, make sure the Meta pixel records the enquiry as a result. Without it, Meta is optimising for something else, usually clicks.
What to expect
- Weeks one and two. Noisy. Costs jump around and a single good or bad day means little. Resist changing things.
- Weeks three and four. Your first honest read: cost per lead, and whether the leads are real people who want what you sell.
- Months two and three. Judge on jobs booked, not leads. Keep the ads that bring customers, replace the rest, and only then consider a second ad set or more budget.
Some businesses find Meta works well at this level. Others find Google, where people are already searching, is the better use of a small budget. Three months of clean data will tell you which you are.
Keeping it on track without living in Ads Manager
The hard part of a small budget is discipline: not tinkering, keeping spend on pace, and looking at results against the goal rather than clicks. That is what AdsEye is built for. You tell Max, the AI ads specialist, your goal, and he works every campaign against it each morning, checks spend hourly and reports on the 1st in plain English. Bigger decisions wait for your yes, with the reasons and numbers. It works from about £500 a month of ad spend per platform. Results vary, and no one can guarantee them.
