Original research Piece 08 · ~11 min read

The cost-per-acquisition benchmarks I track across verticals.

Honest benchmarks for healthcare, e-commerce, professional services and home services — drawn from anonymised aggregate data across the four agencies.

By Sanam Munshi · All pieces

Most published CAC benchmarks are useless. They're either pulled from an industry survey of agencies grading their own homework, or they're aggregated across so many sub-verticals that the averages don't match anything that actually exists in any specific business. The healthcare benchmark that says "average cost per lead $48" is averaging cosmetic injectables (with a lifetime value in the thousands) against bulk-billed GP enquiries (with a lifetime value in the dozens). It's true at the average and useless for any individual clinic.

The benchmarks below are drawn from anonymised aggregate data across the four agencies — Pracxcel for healthcare, Conquerra for US/UAE multi-industry, Skyward for Australian non-healthcare, Trade Pulse for US home services. Where the underlying client engagements are too few to publish responsibly, I've left the cell empty rather than fill it with a number that wouldn't survive scrutiny. The numbers below should be treated as orientation, not precision.

Healthcare — Australian clinics (Pracxcel)

Healthcare CAC varies dramatically by specialty, by service line within the specialty, and by the clinic's geographic catchment. The numbers below are for cost per booked appointment — the metric that actually matters, rather than cost per lead — averaged over the trailing twelve months across active engagements.

The CAC numbers above are blended across paid search, paid social, organic and local — they're what the clinic actually pays per booked appointment when total marketing spend is divided by total bookings, not the CPA shown inside Google Ads or Meta.

US home services (Trade Pulse)

Home services CAC is measured per verified lead rather than per booked job, because Trade Pulse operates on a pay-for-leads model and lead quality is the contractual unit. Lead-to-job conversion at the contractor end typically runs 30%-65% depending on the trade and the contractor's sales process.

US/UAE multi-industry (Conquerra)

For Conquerra, the most useful unit is cost per pipeline-qualified lead (B2B SaaS and professional services) and cost per acquisition (e-commerce). Numbers below are blended across paid channels and organic-attributable.

Australian businesses outside healthcare (Skyward)

Skyward's industry mix is broader and the published numbers below are necessarily wider in range as a result. They should be treated as orientation by industry rather than precision.

How to use these benchmarks

Two ways. First, as a rough check on whether your current CAC is in a reasonable band for your industry. If you're paying three times the upper bound, something is structurally wrong with the campaigns. If you're paying half the lower bound, either you've found a genuine efficiency edge or — more likely — you're not measuring CAC properly and there's untracked spend somewhere.

Second, as the foundation for an LTV-to-CAC ratio that tells you whether to scale or to fix. The ratio that matters is LTV-to-CAC of 3:1 or higher. Below 2:1, the unit economics don't work and no amount of ad budget will fix it. Between 2:1 and 3:1, you're operating viable but tight, and the work to do is on retention and average order value rather than acquisition. At 3:1 or above, the work is to scale acquisition until the ratio compresses to your sustainable target.

Benchmarks are useful for orientation. They're useless as a substitute for measuring your own numbers properly.

The numbers above will be updated quarterly. Where the data set is large enough to publish narrower bands — breaking dental into cosmetic vs. routine, splitting cosmetic injectables by service type — the next iteration of this benchmark will do that. Where it isn't, I'll keep the bands wide rather than fabricate precision.