How to use CAC Measurement
A practical guide to measuring customer acquisition cost in a post-iOS14, multi-touch world — covering the formula, new-customer definitions, and the CAC variants that matter to finance, marketing, and the board.
CAC Measurement
CAC measurement is the practice of computing customer acquisition cost consistently across channels, time windows, and stakeholders.
CAC measurement is how you turn marketing spend and new-customer counts into a single number that finance, growth, and the board can all act on. The math itself is simple — total acquisition cost divided by new customers — but every term in that fraction hides a decision. What counts as acquisition spend? Does a returning subscriber count as 'new'? Do you measure paid CAC, blended CAC, or fully-loaded CAC?
In a post-iOS14 world where Meta's reported conversions overstate paid performance and GA4's attribution windows changed, getting these definitions right is the difference between a profitable scale-up and a six-month detour into unprofitable spend.
Most online stores in the €1M-€15M revenue band report a CAC number that's wrong by 20-40%. The math isn't the problem — the inputs are. Marketing teams pull paid spend from ad platforms; finance pulls new-customer counts from Shopify; nobody reconciles them against the same time window or the same definition of 'customer'.
The result: the CAC you defend to your board doesn't match the CAC your media buyer is optimising against, and neither matches the CAC your finance lead uses to model payback. This guide walks through the four decisions that fix that — the numerator, the denominator, the CAC variant, and the reporting cadence.
What belongs in the CAC numerator
The strictest definition of CAC includes every euro you spent to acquire customers in the period — paid media, agency fees, marketing salaries, attributed software costs, creative production, and affiliate payouts. That's fully-loaded CAC, and it's what the board wants to see.
The looser definition — paid CAC — includes only working media: Meta, Google, TikTok, programmatic. It's what your media buyer optimises against because it's the only number they can move week-to-week. Both are valid; they just answer different questions.
The debate between Blended CAC vs Paid CAC matters because the gap between them tells you how much of your growth is paid versus organic. A Shopify apparel store with €80 paid CAC and €55 blended CAC is getting meaningful lift from email, SEO, and word-of-mouth. The same store with €80 paid and €78 blended is essentially paying for every customer.
Don't double-count discounts
If you give first-time buyers a 15% welcome code, that discount is already reflected in your revenue line — don't also subtract it from gross margin and then add it back as 'acquisition cost'. Pick one place to recognise the cost and stick to it across reports.
Defining a 'new customer'
The denominator looks trivial — count new customers — until you realise Shopify's 'new customer' flag, GA4's 'new user' metric, and Meta's 'purchases' all measure different things. Shopify flags any email that hasn't ordered before. GA4 flags any device that hasn't visited before. Meta counts a conversion against whichever ad it last-touched within the window.
For CAC, the only definition that matters is the one tied to the customer record in your store. A beauty brand running heavy retention campaigns will see Meta report 4,200 'purchases' in a month while Shopify shows 2,800 genuinely first-time buyers. Using the Meta number deflates CAC by a third — and you'll spend yourself into a hole acting on it. New Customer CAC is the variant that fixes this by anchoring the count to the order-level flag, not the ad platform's claim.
Reported 'new customers' across platforms — same Shopify store, same month
The gap isn't a bug — each platform is measuring what it can see. But when you compute CAC, only the Shopify number reflects actual new customers paying actual money. Everything else either double-counts repeat buyers, counts cross-device visits as separate people, or claims credit for purchases it didn't drive.
Which CAC variant to report
Most teams need three CAC variants running in parallel. Paid CAC for the media team's weekly optimisation calls. Blended CAC (or its inverse, MER — Marketing Efficiency Ratio) for the monthly growth review. Fully-loaded CAC for quarterly board reporting and unit-economics modelling.
MER deserves a special mention because it sidesteps the attribution problem entirely. Instead of trying to assign each customer to a channel, you divide total revenue by total marketing spend. It's blunt but honest — and in a world where Meta's reported ROAS routinely overstates true incremental lift by 30-50%, blunt-and-honest beats precise-and-wrong.
Typical CAC ranges by vertical and CAC variant (DTC, €1M-€15M revenue band)
| Vertical | Paid CAC | Blended CAC | Fully-loaded CAC |
|---|---|---|---|
| Apparel & accessories | €35-€55 | €22-€38 | €42-€68 |
| Beauty & skincare | €28-€48 | €18-€32 | €35-€58 |
| Home & lifestyle | €55-€90 | €38-€65 | €68-€110 |
| Food & beverage (subscription) | €42-€75 | €30-€52 | €55-€95 |
| Consumer electronics | €70-€120 | €48-€85 | €85-€145 |
The right CAC also varies by channel, which is why CAC by Channel reporting is the operational layer underneath the headline number. Your blended €38 might be hiding €22 from email, €65 from Meta prospecting, and €110 from a TikTok creator deal that hasn't paid back yet. The headline number tells you if you're profitable; the channel breakdown tells you where to cut and where to lean in.
Reporting CAC to different stakeholders
A media buyer needs CAC by channel, by campaign, refreshed daily, with platform-reported conversions as a directional signal. Finance needs fully-loaded CAC by cohort, refreshed monthly, reconciled to the P&L. The board needs blended CAC trended over six quarters alongside payback period and LTV:CAC ratio.
If you report the same CAC number to all three, two of them will make worse decisions. The media buyer will optimise too slowly against a monthly fully-loaded number. The board will panic at the weekly volatility of a daily paid CAC. Match the variant to the decision.
The 30-day reconciliation rule
Once a month, take the previous month's closed numbers and reconcile paid CAC (from ad platforms) against new-customer CAC (from Shopify). The delta is your attribution drift — track it as its own metric. If it grows beyond ±15%, your platform-reported numbers are no longer reliable for optimisation, and you need to lean harder on blended CAC or MER.
Frequently asked questions
CAC = total acquisition cost ÷ number of new customers in the same period. The 'total acquisition cost' can be paid media only (paid CAC), all marketing spend (blended CAC), or marketing plus salaries, tools, and overhead (fully-loaded CAC). The denominator should always be new customers, not total orders.
Paid CAC divides paid media spend by customers attributed to paid channels. Blended CAC divides total marketing spend by all new customers, regardless of channel. Blended CAC is lower when organic, email, and referral are pulling weight; the gap between the two is one of the cleanest signals of marketing efficiency.
For board and investor reporting, yes — fully-loaded CAC including salaries, agency fees, and tooling is the honest number. For weekly media optimisation, no — your buyer can't act on a number that includes their own salary. Run both variants in parallel.
Shopify and most e-commerce platforms key on email, so a returning buyer with a new email gets flagged as new — inflating your new-customer count and deflating CAC. For a more accurate denominator, dedupe on a combination of email, phone, and shipping address before computing CAC.
Match the window to the typical consideration cycle for your category. Impulse-buy categories like beauty and apparel can use same-month spend and same-month customers. Considered purchases like furniture or electronics need a 30-60 day lag between spend and the customers it produced.
Platform-reported conversions (especially Meta) became less reliable after iOS14's ATT prompt, typically overstating performance by 20-40%. Most teams now treat platform CAC as directional and use blended CAC or MER as the source of truth, reconciling the two monthly to track attribution drift.
MER (Marketing Efficiency Ratio) is total revenue divided by total marketing spend. It avoids attribution entirely, which makes it the most honest signal in a post-iOS14 world. Use it alongside CAC for monthly and quarterly reporting; it's especially useful when channel attribution is unreliable.
There's no perfect answer — every attribution model (last-click, linear, time-decay, data-driven) makes a different trade-off. Pick one, document it, and stick with it for at least two quarters so trends are comparable. Use incrementality tests (geo holdouts, spend-pulse tests) to validate the picture quarterly.
For DTC with strong repeat rates, 3-6 months is healthy and 6-12 months is workable if LTV is strong. Beyond 12 months you're funding growth from working capital, which works only as long as cash holds out. Payback under 3 months usually means you're under-investing in acquisition.
Exclude VAT from both numerator (ad spend ex-VAT) and denominator-adjacent revenue figures. Including VAT in one side and not the other is the most common source of a 20%+ error in CAC reports, especially for stores selling across EU markets with different VAT rates.
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.