Start with the formula that measures fact
The actual cost per lead is simple: ad spend divided by the number of leads received (messages, calls, form submissions, whatever the campaign's goal was). That number already shows up right inside the ads platform (Meta labels it "cost per result"), no separate calculation needed.
The real question isn't "what does a lead cost right now," it's "what can a lead cost before the ad stops paying for itself." Those are two different numbers, and the fact should be compared against the second one, not against someone else's example from the internet.
Step 1: know the margin on one service
Not revenue, margin, what's left after real costs (materials, the specialist's time, rent allocated per visit). Example: a haircut costs 40 GEL, direct cost (materials plus the specialist's time) is 15 GEL, margin is 25 GEL.
Step 2: estimate how many leads turn into paid visits
Out of ten DMs or form submissions, not all become paid visits, some drop off mid-conversation, some don't show up. Without a history of your own leads yet, a cautious starting estimate is 20-30% (a starting assumption, not an industry standard, and it needs checking against fact after the first few weeks).
Step 3: calculate the acceptable cost per lead
Acceptable cost per lead = margin on one service × the share of leads that turn into paid visits. At a 25 GEL margin and 25% conversion: 25 × 0.25 = 6.25 GEL, that's the ceiling above which the ad stops paying for itself on a one-off service. If clients typically return, the formula should use expected lifetime value across several visits instead of one-time margin, which justifiably raises that ceiling.
Step 4: compare against the platform's actual number
- The actual cost is below the calculated ceiling: the ad is working, budget can rise gradually (no more than 20-30% at a time, so the algorithm's learning doesn't reset).
- The actual cost is above the ceiling, but CTR (click-through rate) is fine: the problem is almost always what happens after the click, a slow DM reply, an awkward booking form, a confusing page.
- Both a high cost per lead and a low CTR: the problem is the ad itself or the audience, fix the copy and the image/video first, only then adjust geography.
A separate guide in this series, "ads aren't bringing leads," walks through step 4 in more detail for each case.
Common mistakes, in short
- Anchoring to a "normal" cost per lead from foreign benchmarks instead of your own margin, markets and ad competition differ sharply by country.
- Calculating margin as revenue, without subtracting direct costs.
- Not distinguishing a "lead" from a "paid visit," the gap between them can be several times over.
- Judging by one or two days of delivery instead of 3-5 days of a stable result.
Checklist
- [ ] The margin on one service is known (not revenue)
- [ ] There's an estimate of the share of leads that turn into paid visits
- [ ] The acceptable cost-per-lead ceiling is calculated
- [ ] The platform's actual number is compared against that ceiling, not someone else's figures
- [ ] The decision is based on 3-5 days of a stable result, not a single day
What you can hand off to us
Working out the real cost of a lead for your niche is part of the audit and 90-day plan, from $400. You can also discuss the numbers for your own ads.