The question I get asked most about Meta ads isnât âdo they work.â Itâs âhow much do I actually need to spend.â And most people asking have already picked a number in their head, usually something small, and theyâre really just hoping Iâll confirm itâs enough.
It usually isnât. Not because ads are expensive. Because the number that matters isnât the daily budget, itâs whether youâve given the campaign enough runway to actually learn something.
Hereâs what I tell people to budget for, month by month, and why the figure looks different at each stage.
Why there isnât one number
Every trade and service business asking this question has a different job value, a different service area, and a different amount of competition bidding for the same audience. An electrician in a regional town and a landscaper doing full backyard builds in a capital city are not working with the same economics, even if their ad spend looks identical on paper.
What is consistent, regardless of trade, is the shape of the first three months. The spend should move. If itâs flat from day one to day ninety, somethingâs wrong, either youâre overspending too early or youâre not adjusting based on what the data is telling you.
Month one: budget for tuition, not results
In the first month, youâre not really buying leads. Youâre buying information about which audience responds, which message lands, and which offer gets someone to actually stop scrolling.
For most trade businesses, that means somewhere in the range of $1,000 to $1,500 for the month, spread across a small number of ad variations so you can see whatâs actually working rather than guessing. Spend less than that and Metaâs algorithm doesnât get enough signal to optimise properly, you end up paying for a system that never really learns who to show your ad to.
Expect the cost per enquiry to be the highest it will ever be this month. Thatâs normal. Youâre paying to rule things out.
Month two: budget for narrowing, not scaling
By month two you should know roughly which audience and which message is pulling its weight. This is where the budget doesnât necessarily go up, it gets pointed more precisely. Cut whatâs clearly underperforming, put more weight behind whatâs converting.
A rough guide is to hold spend steady or increase it modestly, maybe 10 to 20 percent over month one, while you consolidate around fewer, stronger ad sets. The goal isnât more traffic yet. Itâs a lower, more consistent cost per enquiry.
This is usually the month business owners get impatient and want to either double the budget or kill the campaign. Neither is usually the right call yet.
Month three: budget for a decision
By the end of month three, you should have enough data to know your rough cost per booked job, not just cost per enquiry, but what it actually costs to turn a click into paid work. That number is the one that matters. Everything before it is groundwork.
If the number works for your margins, month three is when it makes sense to genuinely increase spend, because now youâre scaling something proven rather than hoping. If it doesnât work, month three is when you stop or rework the offer, not extend the experiment indefinitely on the theory that itâll turn a corner.
As a rough total, most trade and service businesses I work with are budgeting somewhere between $3,000 and $5,000 across the first three months to get a genuine read on whether the channel works for them. Thatâs not a hard rule, a business with a low job value and thin margins might need to test smaller and slower. But businesses that try to find this out on $300 a month rarely get a real answer either way.
The mistake that wastes the whole budget
The most common way I see this go wrong isnât the total spend, itâs the pattern. Someone spends $500 in week one, sees a couple of enquiries that donât convert, panics, and pulls the budget to $50 a week. At that level the algorithm never properly learns, so the campaign limps along indefinitely without ever getting a fair test.
Cutting spend to zero to âwait and seeâ is a legitimate decision. Cutting it to a level too small to function is just a slower way of wasting the same money.
The other version of this mistake is the opposite: spending the whole three-month budget in the first two weeks because early results looked promising, then having nothing left to consolidate on once you know whatâs actually working. Pace matters as much as total spend.
What has to be true before any of this makes sense
None of these numbers mean anything if the landing page the ads point to isnât built to convert, or if enquiries sit unanswered for two days once they come in. A well-paced budget pointed at a weak destination just shows you, more efficiently, that the destination is the problem.
Budget for the ads. But budget the same attention toward what happens the moment someone clicks.
The honest takeaway
If youâre asking what to budget for the first three months, the real answer is: enough to let the algorithm learn, enough to narrow in on what works, and enough runway to make an actual decision at the end rather than a guess. For most trade and service businesses, that lands somewhere in the low thousands total, not the couple of hundred dollars people often start with.
Spend less than that and youâre not really testing Meta ads. Youâre testing whether $50 a week can outsmart an auction system built to need more data than that. It canât.