What's a normal CAC for a US D2C brand — and what can you actually afford to pay?

A normal CAC for a D2C brand in 2026 sits between $20 and $73 depending on category and price point — the median across 4,000+ Shopify brands is roughly $34 in apparel, $44 in beauty, and $73 in health & wellness. But that number is not what most founders think it is, and comparing against it without adjusting will overstate your efficiency by half.

That adjustment — and what your brand can actually afford to pay, which is a different question from what's normal — is what this piece works through, with the calculations shown.

What's a normal CAC for a US D2C brand right now?#

The most reliable published medians in 2026: $34.03 in apparel, $44.29 in beauty, $49.39 in food & beverage, and $72.66 in health & wellness, measured across 4,000+ Shopify brands by Polar Analytics. On Meta specifically, the median cost per acquisition across ~35,000 brands was $38.19 in 2025.

Here is the current picture, from the two largest panels that publish their data:

Price point moves the number more than category does. By AOV band, Polar's July 2026 medians run: under-$50 products acquire at $20.56, $50–$150 at $34.46, $150–$1,000 at $64.07, and over-$1,000 at $168.63. A $70 skincare brand and a $900 mattress brand live on different planets, which is why a single "average ecommerce CAC" is a number for slide decks, not decisions.

Two honesty notes that no benchmark page prints. First, there is no US-scoped D2C CAC benchmark anywhere — Polar, Triple Whale, and Shopify all publish without declaring a country. These panels are US-heavy, so Profitbox360 reads them as the best available proxy for the US market, but that is an inference, stated as one. Second, some widely-quoted numbers fail basic tracing: the "$377 electronics CAC" circulating since 2021 comes from a Shopify measurement of stores with fewer than four employees (today's measured electronics median is $43.38), and the "$68–$84 average ecommerce CAC" appears on none of the pages it's attributed to. If a benchmark can't name its measurer, it isn't a benchmark.

Sources: Polar Analytics ecommerce benchmarks (4,000+ brands, medians, updated weekly) · Triple Whale Meta benchmarks (~35,000 brands, full-year 2025)

Am I even calculating CAC the way the benchmarks do?#

Almost certainly not — and the gap is about 60%. The benchmark "CAC" figures above are actually cost per order, counting repeat buyers, with ad spend as the only cost. Your real cost per new customer, on the same data, is roughly 1.6× the published number. Check the definition before comparing anything.

This is the trap, and it's checkable. Triple Whale's own methodology defines its metric as ad spend divided by purchases — all purchases, not new customers. Their July 2026 panel data shows what share of orders actually come from new customers: at a $1M–$10M apparel brand, 63.8%. The rest are returning buyers who cost nothing to re-acquire but still dilute the average.

So the conversion is one line of arithmetic:

True new-customer ad CAC = published CPA ÷ new-customer order share Apparel: $36.76 ÷ 0.638 = $57.62

That's the Meta ad cost of a genuinely new apparel customer — 57% above the number the benchmark page shows. (This is Profitbox360's calculation from Triple Whale's published inputs; no one publishes it as a single figure.)

And that's still only ad spend. A fully-loaded CAC adds agency and freelancer fees, creative production, the tools stack, and first-order discounts — the inclusion list Polar Analytics and Shopify both publish, tracing back to Brian Balfour's original "fully loaded CAC" essay. There is no survey measuring what share of brands count these (two independent sources confirm no consensus exists), which means every founder comparing their spreadsheet against a benchmark is very likely comparing two different metrics. The practical rule: benchmark against per-order CPA with your own per-order ad cost, and run your affordability math — the next section — on fully-loaded, new-customer numbers only.

Sources: Triple Whale benchmark methodology · Polar Analytics on CAC · Balfour, "How to actually calculate CAC"

What CAC can my brand actually afford?#

The CAC you can afford has nothing to do with the benchmark — it equals your contribution margin per order before marketing. Sell at an $82 AOV with $31 of pre-marketing margin, and $31 is your break-even CAC on the first order. Above it you're financing customers; below it you're printing them.

The affordability line, built for a typical apparel brand — every input sourced, every assumption named:

So the first-order break-even CAC for this brand is about $31 — against a true new-customer Meta CAC of ~$58 from the last section. On first orders alone, this brand loses roughly $27 per customer acquired, and the published benchmarks would have told it everything was normal.

That's not a doom verdict — it's the setup for the only question that decides whether high CAC is survivable, which is where the repeat rate walks in.

Source: NYU Stern margin data · Contribution margin structure per Common Thread Collective (healthy CM after ad spend: 20–35% of net sales)

When is a CAC above that line still profitable?#

When repeat purchases close the gap fast enough for your bank account to survive the wait. The full equation: affordable CAC = contribution margin per order × orders per customer — but only orders that arrive within your cash window count. LTV says yes; cash flow decides.

Work the same apparel brand both ways:

At 63.8% new-order share, the average customer places 1 ÷ 0.638 ≈ 1.57 orders. Lifetime affordable CAC = $31 × 1.57 ≈ $49 — still below the $58 true CAC. This brand loses money on Meta even counting repeats, and no benchmark comparison would ever show it.

Now a brand whose customers order 2.2 times: $31 × 2.2 ≈ $68 — the same $58 CAC is now profitable. Same ads, same product economics, opposite verdicts. The repeat rate is the decision variable.

The repeat reality for D2C: Gorgias's 12,000-merchant data shows repeat customers are just 21% of customers but 44% of revenue. Triple Whale's July 2026 cohort data shows how strongly this scales with size — at under-$1M apparel brands, 89.7% of orders are new (there is no repeat cushion yet); at $1M–$10M, it's 63.8%. A small brand quoting the industry's repeat-purchase folklore at itself is borrowing another brand's cushion. (The most-circulated figure, "28.2% average repeat rate," traces to a 2019 study of 65 hand-picked retention-focused brands whose source page is now offline — don't plan on it.)

Then the cash check: payback months = CAC ÷ (margin per order × orders per month per customer). No measured payback benchmark for D2C exists — every "3 to 6 months is healthy" figure in circulation, including Polar's, is a recommendation, not a measurement. Use the recommendation as a conversation floor, but run your own number: if payback exceeds the months of inventory and payroll you can float, the CAC is unaffordable for you, whatever LTV promises.

Sources: Gorgias repeat customer data · Triple Whale benchmarks · Polar Analytics on payback

How do I bring CAC down — and when is CAC not the real problem?#

Work the levers in order of evidence: creative volume first, site speed and checkout second, automation with a measurement plan third. And before any of it — reread the affordability math, because if margin is thin and customers buy once, no ad optimization can save the equation. Fix AOV, margin, or retention first.

What the published evidence honestly supports, ranked:

Creative, treated as a numbers game. Motion's 2026 study of 578,750 creatives found only 4–8% become winners, and the median account ships 6–7 new creatives a week to find them. That's a base rate, not a CPA promise — but it's the clearest published picture of how winning accounts actually operate: they out-iterate, not out-tweak. (The famous "creative drives 47% of results" figure is a US CPG TV study; the "Meta says 32% lower CPA from creative diversity" stat traces to no Meta page at all.)

Site speed and checkout. Deloitte and Google measured +8.4% conversions and +9.2% AOV for retail from a 0.1-second mobile speed improvement — correlational, but across 30 million sessions. Baymard's famous "35% conversion uplift from checkout fixes" is a modelled ceiling from lab testing of large sites, not an average result. And Microsoft's experiment platform found only one third of A/B tests improve their target metric — so CRO pays, but as a testing discipline, not a guaranteed lift.

Campaign automation, measured properly. Meta's own study says Advantage+ cuts cost per purchase 12% (15 tests, their product). Haus's independent 640-experiment incrementality dataset found 58% of brands did better on manual campaigns, with Advantage+ 12% worse on incremental ROAS. The lever that matters is not the toggle — it's measuring incrementally before believing either side's number.

Scaling and marginal CAC. No published saturation curve for ecommerce ad spend exists. The panels' cross-sections (CAC roughly flat from $1M to $100M+ brands) compare different brands, not one brand scaling. Treat the affordability line from section three as your scaling governor: scale until the marginal CAC — the cost of the next customer, not the average — crosses it, then stop and work the levers.

And the reframe. Every lever above attacks the left side of CAC ≤ margin × orders per customer. The right side is usually cheaper to move: raise AOV, rebuild margin, earn the second order. The brand in our worked example doesn't have a CAC problem at $58 — it has a $31-margin, 1.57-order problem. That's the diagnosis the benchmark tables can't give you, and it's the one that decides where the next month of effort goes.

Sources: Motion Creative Benchmarks 2026 · Deloitte/Google, Milliseconds Make Millions · Baymard checkout research · Kohavi & Longbotham on A/B testing · Meta on Advantage+ · Haus Meta Report

---

FAQ#

Is a $50 CAC good or bad? Meaningless without your margin and repeat rate. $50 is unaffordable for an $80-AOV brand whose customers buy once ($31 break-even in our worked example), and comfortable for the same brand at 2+ orders per customer ($68 affordable line). Judge CAC against your own equation, not a table.

What's the difference between CPA and CAC? CPA is ad spend ÷ all purchases, repeat buyers included. CAC is total acquisition cost ÷ new customers only. The big public benchmarks publish CPA under a CAC label — on current panel data the gap is about 1.6× for a mid-size apparel brand.

What LTV:CAC ratio should I aim for? The circulating 3:1 rule has no published measurement behind it for D2C. A 3:1 ratio with a two-year payback can still strangle cash flow, while a leaner ratio that pays back in six weeks is healthy. Payback speed against your cash window is the harder, better test.

How do I calculate my CAC payback period? Payback months = CAC ÷ (contribution margin per order × orders per customer per month). A $58 CAC recovered at $31 margin with a customer ordering every two months pays back in about 3.7 months — survivable if your cash cycle can float it, dangerous if it can't.

---

Profitbox360 builds and runs growth systems for D2C and ecommerce brands — reading the numbers first, so the spend goes where the equation says it should.