Paid Media Economics
CPA vs CAC: measure the event you bought — and the customer you actually acquired.
CPA and CAC are often used as if they mean the same thing. They can be close, but they answer different questions. This guide shows how to separate platform conversion cost from true customer acquisition cost, reconcile the two, and avoid setting paid-media targets from the wrong denominator.
CPA and CAC are related, but they are not automatically interchangeable
CPA usually means cost per acquisition or cost per action. The “acquisition” may be a lead, registration, purchase, booking or another conversion event configured in an ad platform or analytics tool.
CAC means customer acquisition cost. The denominator is an actual new customer, not simply a tracked conversion event. That distinction matters whenever leads fail to become customers, repeat buyers are counted as conversions, offline sales are missing, or platform attribution credits the same customer more than once.
CPA is normally an optimization metric. CAC is a business economics metric.
CPA and CAC formulas
The numerator can also differ. A media-only CAC may use ad spend, while a fully loaded CAC can include agency fees, sales costs, software, creative production or other acquisition expenses. Decide which definition you are using before comparing periods or teams.
Why platform CPA can look better than customer acquisition cost
Google Ads, Meta Ads and analytics platforms report conversions under their own attribution rules. A campaign can therefore show a £30 CPA while the CRM shows that only half of those conversions became qualified customers.
That does not make platform CPA useless. It makes it a different view of the funnel. Platform CPA is still valuable for bidding, campaign comparison and detecting changes in efficiency, provided the conversion event is meaningful and stable.
- Keep the optimization event close to the business outcome when possible.
- Do not compare platform CPA with CRM CAC as though attribution rules are identical.
- Watch for duplicate leads, existing customers and low-quality form submissions.
- Use backend customer data as the source of truth for actual customer economics.
Worked example: the campaign CPA is £40, but CAC is £100
Both numbers are correct. They simply describe different stages. If a media buyer is optimizing lead generation, £40 CPA is useful. If the business is deciding whether acquisition is profitable, £100 CAC is the more relevant number.
Set CPA targets from downstream economics, not from a benchmark
A target CPA should connect to the value and conversion rate of the event being bought. If a qualified lead converts to a customer 25% of the time and the business can afford a £120 CAC, the implied break-even lead CPA is roughly £30.
This is more defensible than copying a target from another advertiser because it starts with your own economics. The same logic works backwards from registration, demo or trial events as long as the downstream conversion rate is measured reliably.
For lead generation, quality is often the missing bridge
Lead-gen accounts frequently optimize toward the cheapest measurable action while the business cares about qualified opportunities or closed customers. When lead quality varies by campaign, audience or keyword, overall CPA can improve while actual acquisition economics deteriorate.
Track at least one downstream quality stage if the sales process allows it. Even a simple qualified/unqualified status can reveal whether a low CPA campaign is producing useful demand or just inexpensive form submissions.
For ecommerce, separate new-customer CAC from purchase CPA
A purchase campaign may contain a mixture of new and returning buyers. Purchase CPA answers how much it cost to generate an attributed order; new-customer CAC answers how much it cost to add a new customer. They diverge when repeat purchase volume is high or prospecting and retention traffic are mixed together.
If the business values repeat purchases, connect CAC with contribution LTV rather than judging acquisition only on first-order revenue. TheLTV:CAC Calculator can help model that relationship, but the lifetime assumptions still need to come from real customer data.
A practical CPA-to-CAC reconciliation workflow
- Choose one reporting period. Use the same dates for spend, platform conversions and CRM outcomes.
- Define the platform event. Lead, purchase, registration and qualified lead are not interchangeable.
- Deduplicate where possible. Separate repeat conversions, spam and existing customers.
- Count new customers from the business system. CRM, ecommerce backend or finance should win over platform attribution for customer count.
- Calculate both metrics. Keep CPA for optimization and CAC for economics.
- Investigate the gap. A widening gap can point to lead quality, attribution, sales follow-up or customer-mix changes.
Common mistakes
A form submission or checkout event is not automatically a new customer.
Margin, refunds, fulfilment and repeat value determine what acquisition can actually cost.
Platform CPA and CRM CAC can be compared, but only after acknowledging they count outcomes differently.
Lead-to-sale cycles can make recent CAC look temporarily worse because customers have not matured yet.