Chapter 06 · 6 min read
The Economics of Verification
What verification actually costs, who pays, and why the unit economics matter at scale. Plus the conversion math nobody talks about.
Verification looks like a cost center until you do the math. Then it looks like the highest-leverage spend in your customer acquisition stack.
Unit costs
| Check | Wholesale | Retail (vendor) |
|---|---|---|
| KYC (document + liveness) | $0.30–0.80 | $1.50–5.00 |
| KYC + AML screening | $0.50–1.20 | $2.00–8.00 |
| KYB (basic) | $2.00–5.00 | $5.00–25.00 |
| KYB (with UBO mapping) | $8.00–20.00 | $25.00–80.00 |
| Background check (US, standard) | $8.00–15.00 | $25.00–60.00 |
| AML monitoring (per user per year) | $0.05–0.20 | $0.50–3.00 |
The wholesale-to-retail spread is the verification industry's gross margin. It's wide because integration cost dominates: it takes a customer 4–12 weeks to onboard a verification vendor. Once they're integrated, switching has friction. Vendors price for that.
The conversion math
The most important number in verification is the completion rate: of users who start the flow, what percentage finish?
Industry medians:
- Web flow: 50–65% completion
- Mobile native flow: 70–85% completion
- Mobile web flow: 35–55% completion (the worst)
If your customer has $100 LTV and you charge $5/check, you'd assume your "cost per acquired user" is $5/0.6 = $8.33 (because 60% complete). Easy.
What actually happens: the 40% who drop off would have included a disproportionate share of high-value users (because high-value users have more friction to lose by switching to a competitor's flow). Your real cost per acquired-and-retained high-value user is closer to $20.
Which is why a 10-percentage-point completion-rate improvement is worth more than a 50% price reduction.
Where verifications fail (and how to fix it)
| Failure mode | % of drop-off | Fix |
|---|---|---|
| Camera permissions denied | 15–25% | Pre-prompt explainer, mobile deep links |
| Document blur / low light | 20–30% | Real-time quality feedback during capture |
| Liveness fails on first attempt | 10–20% | Clear instructions, retry without restarting |
| "I'll come back later" | 25–40% | Magic link to resume, time-bounded session |
| Document not supported | 5–10% | Wider document type coverage upfront |
The "I'll come back later" bucket is the largest and most fixable. Resumable sessions with email/SMS reminders typically claw back 30–50% of those.
Volume discounts and the sales-call tax
Most verification vendors price publicly for the first tier, then require sales calls for volume discounts. This adds two costs:
- Direct cost: 4–8 weeks of procurement.
- Indirect cost: The vendor knows your willingness to pay before quoting, so the discount is exactly enough to keep you, not what they could afford to give.
The right pricing structure (which we believe in deeply): publish every rate. Publish every volume tier. No sales call, ever. Customers self-select up the curve as they grow. Vendors lose some price discrimination revenue but gain enormous trust + faster sales cycles.
What you actually pay for
The per-check price is a small part of your real cost. The total cost of verification includes:
- The check itself
- Engineering integration (one-time)
- Ongoing maintenance as the vendor's API evolves
- Manual review staff (for the % that requires human judgment)
- Compliance documentation overhead
- Customer support for users with verification problems
For a typical mid-market company doing 10k verifications/month, the per-check price is <30% of total cost. Choose your vendor optimizing for the rest.