CPA

Metricuno
May 17, 2026
4 min read
CPA — CPA explained: the cost-per-acquisition formula, how it differs from CAC, paid-channel benchmarks, and how to use it to defend ROAS without hiding leaks.
Quick answer

CPA is paid spend divided by conversions on a single channel — the per-platform cousin of CAC. Here's the formula, typical ranges, and how to read it without fooling yourself.

Definition
Acquisition metrics

CPA (Cost Per Acquisition)

CPA is paid spend on a channel divided by the conversions that channel produced — a per-platform view of acquisition cost.

Cost Per Acquisition (sometimes Cost Per Action) measures how much you paid a specific channel — Meta, Google, TikTok, a particular campaign — for each conversion attributed to it. It's the number that lives inside ad platforms and gets optimised against daily.

CPA is often confused with CAC (Customer Acquisition Cost), but they answer different questions. CPA is channel-attributed and narrow: 'how efficient is this campaign?'. CAC is blended and total: 'what does it cost the business to win a customer, all-in?'. A healthy paid program watches both, because a falling CPA with a rising CAC usually means attribution drift, not real improvement.

Also known as
Cost Per Action
Cost Per Conversion

The conversion in the denominator isn't always a purchase. Depending on how a campaign is set up, a CPA can be measured against email signups, add-to-carts, lead-form submissions, app installs, or completed checkouts. Always check what 'acquisition' means in the report you're reading before comparing numbers between platforms.

This is why CPA from Meta Ads Manager and CPA from Google Ads aren't directly comparable out of the box. Meta might be counting view-through conversions on a 7-day window; Google might be click-only on a 30-day window. Same metric name, different denominators — and a chunk of disputes between performance and finance teams trace back to this one detail.

Formula

CPA = Ad spend / Conversions

Variables

Ad spend

Ad spend

Total media cost for the channel, campaign, or ad set in the period — usually excludes agency fees and creative production.

Conversions

Conversions

Count of the defined conversion event attributed to that spend under the platform's attribution model.

Worked example

A Shopify apparel brand runs Meta paid social for a swimwear launch over 30 days.

Ad spend (Meta, 30d): €18,000

Attributed purchases: 360

€50.00 CPA

Each purchase from Meta cost €50 in media. If average order value is €85 and contribution margin is 40% (€34), the channel is losing €16 per order on first purchase — only acceptable if repeat behaviour pays it back inside the payback window.

Read CPA against contribution margin, not revenue. A €50 CPA on a €120 order with 50% margin clears €10 per order; the same €50 CPA on a €60 accessory loses money on acquisition and only works if repeat purchase is reliable. Margin-blind CPA targets are how brands scale themselves into a cash-flow problem.

Benchmark

Typical CPA ranges by channel and store vertical (first-purchase conversions)

VerticalMeta AdsGoogle SearchTikTok Ads
Apparel & accessories€25-€55€18-€40€20-€45
Beauty & skincare€20-€45€15-€35€18-€40
Home & furniture€60-€140€45-€110€55-€130
Consumer electronics€40-€90€30-€75€35-€80
Food & beverage (DTC)€18-€40€14-€30€16-€35

These are first-purchase ranges for stores in the €1M-€15M revenue band. Higher-AOV verticals tolerate higher CPAs because contribution per order is larger; lower-AOV categories only work if repeat rate carries the LTV. If your reported CPA sits well below these ranges, suspect attribution overlap before celebrating — the same conversion may be counted in two platforms.

Frequently asked

CPA: frequently asked questions

CPA is per-channel and uses that platform's attribution: it's the number Meta or Google reports. CAC is blended — total acquisition spend (paid, agency, salaries, tools) divided by total new customers across all channels. CPA optimises a channel; CAC tells you whether the business is winning customers profitably.

No. CPC (Cost Per Click) measures the cost of a click on an ad. CPA measures the cost of a conversion — a purchase, signup, or other action. CPA is downstream: it depends on CPC plus landing-page conversion rate.

Work backwards from contribution margin and payback window. Take average order value, subtract COGS, shipping, and payment fees to get contribution per order, then decide how many months of repeat margin you'll spend to acquire. Target CPA = contribution per order + planned repeat contribution within payback window.

Meta uses its own attribution window (often 7-day click, 1-day view) and counts conversions it 'saw' the user interact with. Shopify counts orders by last-click or by source tag. The two will rarely match — Meta typically over-reports, GA4 and Shopify under-report. Use a consistent source for board-level reporting.

By default, no — most platforms report media-only CPA. For a true cost view, build a 'loaded CPA' that adds agency retainers, creative production, and ad-ops tooling, then divide by the same conversion count. Loaded CPA is usually 15-30% higher than media CPA.

There's no universal number — it depends on AOV and margin. A useful test: CPA should be less than contribution margin on the first order plus expected contribution from repeat purchases within your payback window (often 3-6 months). Most healthy apparel and beauty stores keep first-purchase CPA between €20 and €55.

CPL (Cost Per Lead) is CPA where the conversion event is a lead — an email or form submission, not a purchase. CPL is common in higher-consideration categories (furniture, financial products) where the buy doesn't happen in-session. A CPL strategy needs a measured lead-to-customer rate to translate back to economics.

They're two views of the same trade-off. CPA targets work well when AOV is stable; ROAS targets work better when AOV varies (bundles, sales). Most teams set a CPA ceiling per channel and a ROAS floor across the account, and review whichever is the binding constraint that week.

Since iOS 14.5, platform-reported CPA has become less reliable — Meta and TikTok rely on modelled conversions for opted-out users. Reported CPA can drift 10-25% from server-side truth. Use platform CPA for in-channel optimisation, but reconcile against server-side data (Shopify, GA4, a CDP) for finance reporting.

CPA is one of the core ecommerce metrics teams watch alongside ROAS, CAC, AOV, conversion rate, and LTV. It tells you per-channel efficiency; the others tell you whether that efficiency translates into a profitable business. Reading CPA in isolation is how channels look great while the P&L gets worse.

Track CAC, channels, and funnel conversion in one place

Metricuno connects ad spend, funnel events, and revenue so you can see CAC by channel, cohort, and campaign — without stitching together five tools.