Most small business owners size their Facebook ads budget the wrong way: they pick a number that feels affordable — $10/day, $300/month — instead of working backward from what a customer is actually worth. That's backwards, and it's why so many campaigns get killed after two weeks with a shrug of \"Facebook didn't work for us.\" It usually wasn't Facebook. It was the math.
Start with job value, not a percentage of revenue
Forget the \"spend 5-10% of revenue on marketing\" rule. It tells you nothing about whether an individual ad dollar makes sense for your business. What matters is: how much is one closed job worth, and how much can you afford to pay to generate the lead that becomes that job?
Here's the actual math. Take your average job value, multiply by your lead-to-close rate, and that tells you your maximum sane cost per lead (CPL).
- Roofer, average job $11,000, closes 1 in 5 leads: break-even CPL is roughly $2,200. Real-world roofing CPL on Facebook runs $25-60, so there's enormous room to profit — see facebook ads for roofers for the full breakdown.
- Electrician, average ticket $650, closes 1 in 4 leads: break-even CPL is about $162. Actual electrician CPL runs $18-40, again with plenty of margin — details at facebook ads for electricians.
- Cleaning business, average recurring client worth $1,800/year, closes 1 in 6 leads: break-even CPL is around $300 on lifetime value, but most owners price against the first job (~$150), which pushes break-even CPL down to $25 — tight but workable, covered in facebook ads for cleaning business.
Once you know your break-even CPL, you compare it to real Facebook CPL ranges for your trade (see how much do Facebook ads cost for a full table) and you know instantly whether this channel can work before you spend a dollar.
The Meta learning phase, explained in dollars
Meta's ad delivery system needs data to optimize. The commonly cited benchmark is about 50 conversion events (leads, calls, form fills) per ad set in a rolling 7-day window before delivery stabilizes and cost per lead settles into a predictable range. Below that, Meta is still guessing which users to show your ad to, and your CPL will swing 2-3x from day to day.
Translate that into dollars per trade, using real CPL ranges:
| Trade | Typical CPL | Weekly spend for 50 conversions | Realistic weekly spend (15-20 conversions) |
|---|---|---|---|
| HVAC | $25-45 | $1,250-2,250 | $375-900 |
| Plumbing | $18-35 | $900-1,750 | $270-700 |
| Electrician | $18-40 | $900-2,000 | $270-800 |
| Roofing | $25-60 | $1,250-3,000 | $375-1,200 |
| Cleaning | $12-25 | $600-1,250 | $180-500 |
| Gym / fitness | $8-20 | $400-1,000 | $120-400 |
| Dentist | $25-50 | $1,250-2,500 | $375-1,000 |
Notice the gap: hitting Meta's textbook 50-conversion benchmark costs $900-3,000/week for most trades — money almost no local business is going to put behind a single ad set. That's fine. In practice, 15-20 conversions a week is enough to see a real pattern and make a keep/kill decision, and it costs a fraction of that. The 50-conversion number is Meta's ideal for the algorithm, not a requirement for your business to get value.
Launch your first Facebook & Instagram ad in 2 minutes
Describe your business: the AI writes the copy, designs the visual, sets the targeting, and publishes your ad. Leads arrive with their phone number.
Try Leadria free7-day free trial — no credit card — cancel anytime
Why $5/day fails for most trades — and where it doesn't
$5/day is $150/month. At a $30 CPL — normal for HVAC, roofing, or dental — that's 5 leads a month. Five leads isn't a sample size, it's a coin flip. One bad lead skews the whole month's numbers, and you can't tell if the ad, the offer, the targeting, or plain bad luck caused it.
$5/day fails specifically because:
- It can't clear the initial learning phase cleanly — Meta needs volume to learn, and $5/day rarely generates more than 1 lead every 3-6 days for higher-CPL trades.
- It gives you no statistically meaningful read on cost per lead within a month, so you end up making decisions off noise.
- For high-ticket trades (roofing, remodeling, HVAC replacement), the math simply doesn't support it — you'd need 20+ months of $5/day spend to match what a single week of proper budget would test.
Where $5/day can actually work: businesses with very low CPL and low job value, where the math is already forgiving. Restaurants running local awareness or a specific promo often see $3-10 CPL (see facebook ads for restaurants), so $5/day can generate a usable trickle of engagement. It also works as a pure creative test — not to generate leads, but to see which headline or image gets cheaper clicks before you commit real budget to it.
A concrete example: electrician in Tampa, FL
Say a licensed electrician in Tampa runs residential service calls averaging $260, with panel upgrades averaging $2,100, and about 1 in 4 leads books a job (mixed between the two job types, blending to roughly $650 average value per booked job).
Facebook CPL for electricians in the Tampa market typically runs $20-38. At $35/day ($1,050/month), that's 27-52 leads a month depending on where CPL lands. At a 25% close rate, that's 7-13 booked jobs. At $650 average value, that's $4,550-$8,450 in booked revenue against $1,050 in ad spend — a 4x to 8x return before accounting for materials and labor cost.
Drop that same electrician to $10/day ($300/month): 8-15 leads, 2-4 booked jobs, $1,300-$2,600 in revenue. Still profitable, but the sample is small enough that a single slow week can make the whole month look like a failure when it was really just variance. The lesson: the math works at both budgets, but $35/day gives you a result you can actually trust and optimize from.
When this does NOT work
Be honest about the cases where no budget size fixes the underlying problem:
- Job value under $75 with no repeat/referral engine. A $20 car wash or a one-off $40 handyman task can't absorb a $15-25 CPL, no matter how much you spend — the unit economics are broken before the ad runs.
- Long, considered sales cycles. Custom home builds, commercial construction bids, and other $50,000+ purchases rarely close from a cold Facebook lead. Buyers in that category are searching, not scrolling — Google Ads or referrals fit better. See Facebook ads vs. Google ads for small business for when to switch tools.
- No fast follow-up. A lead that sits for 6 hours before a callback converts far worse than one called within 5 minutes. If your team can't answer or call back same-day, extra budget just generates more leads that go cold — see why are my Facebook ads not working.
- You're already at capacity. A landscaping crew booked out 8 weeks doesn't need more leads; it needs to raise prices or hire. Spending on ads here is wasted money regardless of CPL.
- Rural or thin service areas. A 15-mile radius around a town of 8,000 people exhausts its audience fast — CPMs stay flat but the same people see your ad repeatedly, and CPL climbs weekly even with budget increases.
How to set your actual number this week
Take your average job value and your rough close rate, calculate break-even CPL, then compare it against the real CPL range for your trade — the tables in facebook ads for small business and how to get leads from Facebook ads cover most trades. Pick a daily budget that gets you to at least 15-20 leads a week, not 50 — that's the realistic minimum for a trustworthy read. If the math doesn't work at that volume, the problem is your offer or close rate, not your budget.