Most small businesses need roughly $1,500 to $3,000 a month in Meta ad spend before the numbers mean anything. Below that, you are buying noise. The right figure is not a benchmark you copy — it is your cost per lead multiplied by the number of leads you can actually handle, plus enough volume for Meta’s system to learn.
How much do small businesses actually spend on Facebook ads?
Across US lead-generation campaigns, the average cost per lead on Facebook is $27.39. That figure comes from WordStream’s 2026 Facebook Ads benchmarks, built from 452 unique advertiser accounts running between 1 April 2025 and 30 June 2026. It is US data, not Canadian, so treat it as a sanity check rather than a forecast for Mississauga.
The spread by industry matters far more than the average. In the same dataset, home and home improvement leads campaigns averaged $42.95 per lead at a 5.32% conversion rate, while real estate leads campaigns averaged $13.74 per lead at a 9.95% conversion rate. Same platform, same quarter, a three-fold difference in what a lead costs.
So "what should I spend" has no industry-free answer. A home services company and a realtor with identical $1,500 budgets are buying 35 leads and 109 leads respectively.
Why does Meta need a minimum budget to work at all?
Because Meta’s delivery system is a prediction engine, and prediction needs data. Meta’s own documentation sets the learning phase threshold at roughly 50 optimisation events per ad set per week — 50 leads, 50 purchases, 50 whatever you told it to optimise for. Until an ad set clears that, delivery stays unstable and your cost per result bounces around week to week.
Run the arithmetic on the US home services average and the problem is obvious. Fifty leads a week at $42.95 is about $2,150 a week, or roughly $9,300 a month — for a single ad set. Almost no owner-operated business has that, which is why so many small accounts sit permanently in "Learning limited".
The fix is not to spend more. It is to optimise for an event that happens more often. If leads are too scarce to feed the algorithm, optimise for landing page views or link clicks while volume builds, then switch to lead or purchase optimisation once the account is producing enough events to sustain it. That decision — which event, at which spend level — is most of what good Meta ads management actually consists of.
What does each budget level realistically buy?
Treat the table below as a planning tool, not a promise. It applies the $42.95 US home services cost per lead from WordStream’s 2026 benchmarks, which is deliberately on the expensive end.
| Monthly Meta budget | Leads at $42.95 each | What it is honestly good for |
|---|---|---|
| $500 | ~11 | Proving an offer exists. One audience, one or two creatives. Do not expect stable cost per lead. |
| $1,000 | ~23 | A single campaign with a real creative test running. Enough to tell a good hook from a bad one. |
| $2,500 | ~58 | The first level where weekly numbers are worth acting on. Room for creative refresh. |
| $5,000 | ~116 | Two or three campaigns, retargeting layer, and enough events to approach the learning threshold. |
One more thing to price in: Meta is getting more expensive. In Meta’s own Q2 2026 results, released 29 July 2026, average price per ad rose 12% year over year while ad impressions delivered rose 14%. A budget that worked in 2025 buys measurably less attention now.
How do you set your own number instead of copying a benchmark?
Work backwards from the job, not forwards from the budget. Three figures decide it.
First, your average sale value and your close rate on inbound leads. If a job is worth $600 and you close one lead in four, each lead is worth $150 to you, and a $43 lead is comfortable.
Second, your capacity. There is no point buying 60 leads a month if you can only quote 20 — you will burn budget generating leads you never call, and Meta will learn from your worst follow-up, not your best.
Third, your test runway. Plan on at least six to eight weeks at your chosen spend before judging the account. Changing budget weekly resets the learning phase and guarantees you never see a stable number.
Small budgets can work when they are aimed narrowly. One lawn care company in Durham Region booked 28 calls in its first 30 days on a $772 total season budget — a tight service radius and a single clear offer, not a big number. You can see the figures in our case studies.
What to do this week
- Calculate your true lead value: average sale value multiplied by your inbound close rate. Write it down. Any cost per lead well under that number is a working campaign, whatever your gut says.
- Open Ads Manager and check how many of your ad sets say "Learning limited". If most of them do, you are spreading your budget across too many ad sets — consolidate before you add a dollar.
- Count the leads you can realistically quote in a month, and cap your budget at that number multiplied by your expected cost per lead. Capacity, not ambition, sets the ceiling.
- Commit to a fixed daily budget for six weeks and stop touching it. Then judge the account.
Frequently asked questions
Is $500 a month enough for Facebook ads?
It is enough to test whether an offer gets any response, and not enough to produce a reliable cost per lead. At the US home services average of $42.95 per lead reported by WordStream for 2026, $500 buys about 11 leads a month — too few for Meta’s delivery system to optimise against, and too few for you to tell a seasonal dip from a broken campaign. If $500 is your ceiling, run one campaign, one audience, two creatives, and judge it on whether the phone rings at all rather than on cost per lead.
Why did my cost per lead go up when I raised my budget?
Two common reasons. Raising a budget sharply pushes an ad set back into the learning phase, so costs destabilise for several days before settling. And a larger budget forces Meta to reach further into your audience, past the cheapest, most responsive people it found first. A gradual increase — roughly 20% at a time, with a few days between changes — usually holds costs steadier than a sudden jump.
Should I split my budget across Facebook and Instagram separately?
Usually not. Separating placements splits your conversion events across two ad sets, and halving your event volume is the fastest way into "Learning limited". Leave Advantage+ placements on and let Meta put spend where results are cheapest. The exception is when your creative is genuinely format-specific — a vertical Reels video versus a square feed image — and you have enough spend to support both properly.
How long before Facebook ads start producing leads?
Expect the first leads within days and a trustworthy cost per lead in six to eight weeks. The early period is the system testing audiences and creatives against each other, so week-one numbers are not predictive in either direction. Keep the budget flat, keep the offer the same, and let it run. Most accounts that never work were never left alone long enough to find out.
Not sure whether your budget is the problem or your campaign structure is? Book a free 48-hour ad account audit and we will tell you which one it is.
Sources: WordStream, Facebook Ads Benchmarks 2026 (US data, April 2025–June 2026); Meta Reports Second Quarter 2026 Results, 29 July 2026; Meta Business Help Center, about the learning phase.