Ask ten people how much to spend on Facebook and Instagram ads and you'll hear ten numbers. A friend spends $5 a day. A marketing agency quoted you a monthly minimum. A video online swears by one magic figure. None of them know what a customer is worth to your business, and that's the only number that really matters here.
A small business should set its first Facebook and Instagram ad budget from what a new customer is worth, not from a rule of thumb. Work out how much you can afford to pay for one lead, fund a short test big enough to produce real leads, then judge it by cost per lead and actual customers. Likes and reach don't pay the bills.
I won't give you a benchmark cost per click or a "typical" monthly spend. Those numbers swing wildly by industry, offer, season and creative, and a made-up average does more harm than good. What I'll give you is the method I'd use if it were my money.
Step 1: Work out what a customer is worth
Start with profit, not revenue. If a typical job brings in $2,000 but $1,400 goes to materials and labor, the job is worth about $600 to you. If customers tend to come back, a house cleaning client or a patient who returns for years, count some of that repeat value too. Be honest and a little conservative.
Then think about how many leads it takes to win one customer. If you usually close about one in three quote requests, every lead is worth about a third of a customer.
Here's a simple example with made-up numbers, just to show the math:
- A new customer is worth about $600 in profit.
- You close about 1 in 3 leads.
- So one lead is worth about $200 to you before you'd lose money.
- You'd like to keep most of that, so you set a target of $60 or less per lead.
That target is your guardrail. Nobody can promise what ads will cost, but this line tells you whether a campaign is working.
Step 2: Fund a test that can actually teach you something
A test budget needs to be big enough to produce a handful of real leads. If your target is $60 per lead and you want about 20 leads to judge fairly, that's roughly $1,200 over the test. If that's too much, lower the lead goal and run the test longer, but know that fewer results means a fuzzier answer.
There's also a reason the first week or so looks messy. Meta calls it the learning phase. In its help article About the learning phase (checked in 2026), Meta says ad sets usually exit learning after about 50 results in the week after the last significant edit, and that during learning, results are less stable and usually cost more. If your ad set can't get enough results, Ads Manager shows "Learning limited."
For most local businesses, 50 leads a week isn't realistic, and that's okay. It just means two things:
- Don't judge the first few days. Early numbers bounce around.
- Don't keep tinkering. Meta says editing during learning resets it. Change one thing, then wait.
Meta also warns against budgets that are very small or inflated, and against lots of little ad sets. One focused campaign with one or two ad sets learns faster than six tiny ones.
Step 3: Know how Meta spends a daily budget
This one surprises people. A daily budget on Meta is an average, not a hard cap. In About budgets (2026), Meta explains that on days with better opportunities it may spend up to 75% over your daily budget and less on other days, but it won't spend more than 7 times your daily budget in a week.
So a $40 daily budget can show $65 one day and $20 the next. That's normal. If you need a firm ceiling for the whole test, Meta says a lifetime budget is a hard cap.
Step 4: Judge by cost per lead, not likes
When the test ends, look at these numbers in this order:
- Leads. Calls, forms, messages and bookings you can actually count.
- Cost per lead. Total spend divided by leads. Compare it to your guardrail.
- Customers. How many leads turned into paying work.
- Clicks and reach. Useful for spotting problems, like lots of clicks but no forms.
Likes, shares and follower counts come last, if at all. I spent years in lead generation in one of the most expensive ad markets online, and the lesson was always the same. The ad with the most likes was rarely the one that paid for itself. If you can't tell which leads came from ads yet, fix that first. Our sister studio HEX Web Design handles the website side, like forms and tracking.
Your budget checklist
- I know roughly what a new customer is worth in profit.
- I know about how many leads it takes to win one customer.
- I've written down my target cost per lead.
- My test budget is big enough to produce a handful of leads.
- I'll run it for at least a few weeks without constant edits.
- I can tell which leads came from the ads.
- I'll decide based on leads and customers, not likes.
When to spend more
Scale up only after the test shows a cost per lead under your guardrail and some of those leads turned into customers. Then raise the budget in steps rather than doubling it overnight, since Meta notes that frequent budget changes can send an ad set back into learning.
Timing matters too. If your work is seasonal, like AC repair before a Houston summer or catering before the holidays, run your test a few weeks ahead of the rush. That way the learning is behind you when demand shows up.
If the test missed, don't just pour more money in. Change the offer, the creative or the audience, one at a time. That's how we approach paid social campaigns, with plain-language monthly reporting that tells you what each lead cost and what we're changing next.
Want a second set of eyes on yours?
If you're already spending and can't tell whether it's working, we'll look for free. Our Social Check reviews your profiles, your posting rhythm, whether your content looks like you, how comments and reviews are handled, whether your ads are aimed at the right people and whether social visits show up as calls and forms. You get a short written summary. If the answer is "not much," we'll say that. Request your free Social Check.
