Cost per acquisition calculator

CPA vs. CAC: stop conflating these (the CFO is watching)

CPA (cost per acquisition) is per-channel, per-campaign. CAC (customer acquisition cost) is blended across everything. Marketers use the terms interchangeably and accidentally mislead their finance teams constantly. Your Meta CPA might be $58. Your Google CPA might be $31. Your blended CAC — total S&M spend divided by new customers acquired — is almost certainly $95+ once you include agency fees, creative production, tooling (Klaviyo, Triple Whale, Northbeam, your attribution platform), and salary-loaded marketing payroll. This calculator computes CPA; it also teaches the gap between CPA and true CAC, because that gap is where most DTC brands get surprised.

The rule of thumb I use: if your internal team is reporting Meta CPA, multiply by 1.4–1.7 to approximate true CAC including overhead. For agency-managed accounts, multiply by 1.5–1.9. The blended number is what actually shows up on your P&L — CPA just helps you decide which channels to push.

CPA benchmarks by channel and industry (2026)

How to set a target CPA (the math most teams skip)

Target CPA = (LTV × desired payback ratio) − contribution margin adjustments. The fast version: LTV ÷ 3 is the LTV:CAC 3:1 target many startups anchor on, but that's a benchmark, not a rule. Here's a more honest framework:

1. Start with first-purchase contribution margin. For a $120 AOV Shopify store at 45% gross margin, you net $54 per first purchase before marketing cost.
2. Layer in repeat-purchase LTV. If 35% of first-time buyers place a second order averaging $95 in year 1, that's an additional $15 expected LTV per acquired customer.
3. Decide your payback horizon. For self-funded DTC: 30–60 days. For VC-backed SaaS: 12–18 months. This is a financing decision, not a marketing decision.
4. Target CPA = LTV accumulated within payback horizon × buffer. If 60-day LTV is $62 and you want 30% profit margin on new customer acquisition, target CPA = $62 × 0.70 = $43.

Why CPA is rising for everyone (and what to do about it)

Industry-wide CPA has climbed 25–60% since 2020 across DTC. The drivers are well-documented:

• iOS 14.5 ATT (2021) broke Meta's attribution. Reported CPA became less accurate; true CPA was always higher than reported.
• Creative fatigue is faster. Algorithmic-driven placements burn through creative in 2–3 weeks vs. 6–8 weeks pre-2022. More creative needed per dollar spent.
• CPM inflation. Q4 Meta CPMs rose 80%+ from 2019 to 2024 as more brands bid for the same impressions. Use the Meta CPM Trend tool to model this.
• Increased competition. DTC brand count roughly tripled from 2019 to 2024. Same eyeballs, more bidders.

The playbook that works: improve LTV (retention, cross-sell, price increases), not chase lower CPA with desperate creative. If you can lift 12-month LTV by 25%, your CPA tolerance rises proportionally, which lets you outbid competitors and scale profitably.

Reducing CPA: the real levers

1. Landing page conversion rate. Doubling CVR from 2% to 4% cuts CPA in half instantly. Usually the cheapest lever. Use the LP CVR Lift tool to model this.
2. Offer structure. Bundle, free shipping threshold, free trial length — offer changes often produce 20–40% CPA reductions.
3. Audience segmentation. Tighter audiences have higher CPM but dramatically lower CPA.
4. Creative velocity. More fresh creative per week, especially UGC/creator-led, consistently lowers CPA.
5. Funnel fix. Broken form, slow page, missing trust signals. Run hotjar/session replay before assuming ad-level problems.

CPA measurement errors to avoid

• Counting only new customers. Returning customers in a channel's reported "conversions" inflate CPA and hide real performance. Filter for new vs. repeat in Shopify/your CRM.
• Ignoring view-through attribution. Especially for brand-heavy campaigns, 30–50% of conversions arrive without a click. Model with view-through holdout tests.
• Different attribution windows per platform. Meta's default 7-day-click-1-day-view vs Google's 30-day last-click. Normalize before comparing.
• Forgetting organic cannibalization. Branded search ads have CPAs of $5, but 60–80% of those conversions would have happened organically at zero cost.

Three CPA archetypes — DTC, SaaS, and B2B walked through end-to-end

Target CPA only makes sense against concrete unit economics. These three archetypes capture roughly 80% of the models I run into, with April 2026 benchmarks and full channel-mix math. Copy the template that matches you and plug in your numbers.

Archetype 1: DTC CPG (skincare, supplements, pantry)

$68 AOV at 62% contribution margin after COGS, pick-pack, and payment fees. Blended CAC target of $38, payback in 8 months driven by 2.4 repeat orders in year one and a 34% 12-month retention cohort. Channel mix: Meta Advantage+ Shopping at $42 CPA (55% of new customers), Google Shopping at $29 CPA (22%), TikTok Spark Ads at $46 CPA (12%), Klaviyo-driven email/SMS flows at roughly $4 CPA (11%). LTV lands at $186 against $38 CAC — a clean 4.9:1 LTV:CAC with 8-month payback that matches a self-funded cash cycle. The stack costs are real: Klaviyo at a 25k-profile list runs about $600/month, Triple Whale on a Shopify Plus plan lands near $400/month, Shopify Plus itself is $2,300/month, and Meta creative production averages $4,500/month for UGC + editing. Those overheads push blended CAC to roughly $48 once you load them into the S&M denominator — which is the number that actually hits the P&L.

Archetype 2: Mid-market SaaS ($3,600 ACV, self-serve + sales-assist)

Target CAC of $420 with a 14-month payback. Gross margin on SaaS is 78%, so contribution LTV on a 22-month median tenure is $6,180. Channel mix: Google Search brand protection + category capture at $380 CPA for free-trial signups (38% of paid pipeline), LinkedIn Sponsored Content + Lead Gen Forms at $240 MQL / $620 SQL (28%), G2 and Capterra paid placements at roughly $310 per SQL (16%), and content-led organic driving the remaining 18% at blended $85 CAC. Free-trial-to-paid sits at 18%, so the $75 paid trial CPA translates to a $416 paid CAC. Tool stack taxes CAC: HubSpot Pro at $800/month for 5 seats, a paid Salesloft seat at $125/user/month, Clearbit at $12k/year, and a two-person SDR team loaded at $160k/year. Loaded CAC ends up at $520 — still healthy at LTV:CAC of 11.9:1, but payback stretches to 16 months once overhead is real.

Archetype 3: B2B enterprise services (average contract $48k, 18-month term)

Target CAC of $2,400 with a 22-month payback window. Deal count is low — maybe 45 new logos per year — so per-lead math drives the model more than channel benchmarks. Channel mix: outbound SDR + LinkedIn Sales Navigator at $1,900 per opportunity (50% of pipeline), paid webinars + conference sponsorships at $3,200 blended per opportunity (22%), partner-sourced referrals at $900 per opportunity (18%), and inbound content at $420 per opportunity (10%). Close rate is 22% on qualified opportunities, so a $1,900 opportunity cost rolls up to $8,600 CAC pre-overhead on outbound — the only way that pencils is that contribution LTV is $62k on an 18-month deal with 30% expansion. The number that matters here is sales-cycle length: if it stretches past 5 months, stop counting new CAC and instead model pipeline velocity, because CAC gets distorted by the 6-9 month revenue lag.

Channel-by-channel April 2026 paid-media rate card

Measurement stack that makes CPA trustworthy

Your CPA number is only as good as the plumbing underneath it. The 2026 reference stack for a DTC brand spending $100k-$500k/month on paid: Shopify or Shopify Plus as the source of truth for orders, Google Analytics 4 for session-level funnel, Meta CAPI + Google Enhanced Conversions + TikTok Events API all firing server-side via a Google Tag Manager server container (Stape at $120/month or a self-hosted Cloud Run container at $30/month), Triple Whale or Northbeam at $300-$600/month for blended MER reporting, and a weekly manual MER reconciliation in a shared sheet. The sheet is the part most teams skip and it is the one that catches everything: take total S&M spend divided by new customers from Shopify and compare to the sum of platform-reported new customers. The gap is your attribution decay — typically 18-32%. If that gap is growing, your platform CPA numbers are drifting.

For B2B SaaS, the equivalent stack is HubSpot or Salesforce as the CRM source of truth, Segment or RudderStack routing events ($120-$500/month at SMB volume), Dreamdata or HockeyStack for multi-touch attribution ($800-$2,500/month), and a monthly finance reconciliation joining closed-won ARR from the CRM to ad spend in the finance system. Without that last step, your CPA and CAC will diverge by 40%+ within 6 months and you won't know it until the board asks.

Kill-criteria for a paid channel

A channel earns budget by proving CPA within target inside its learning phase. When it stops, cut. My personal kill rule: 2 weeks of optimized delivery past the learning phase + CPA at 3x target + CVR sub-target = stop spend, not "let me try one more creative." Before you cut, verify three things: (1) the pixel is firing correctly with server-side events via Meta CAPI or Google Enhanced Conversions — 30% of "dead channels" I audit are just broken tracking, (2) you're past platform learning exit (50 conversions per ad set in 7 days on Meta), and (3) the CPA lift isn't seasonal. If all three check out and CPA still sits at 3x target, reallocate to the channel with headroom, document the reason for the future, and move on. Teams that run the kill rule cleanly recycle 15-25% of annual spend into winners; teams that don't bleed 10-15% of budget into dying channels every quarter.

The inverse scale rule is equally mechanical. Target CPA hit for 14 consecutive days at current spend, learning phase exit maintained, and no degradation past day 7 of a 20% budget step means you can step up another 20% and repeat the watch. Scaling faster than 20%/week on Meta or Google re-triggers the learning phase and craters short-term CPA; TikTok tolerates 30-40%/week scaling; the Klaviyo-email lane scales instantly because no algorithm needs retraining.