How to Calculate Customer Acquisition Cost (CAC Formula, DTC Benchmarks, and the Marginal CAC Most Brands Miss)

Customer acquisition cost is what you spent to land one new customer. The textbook formula is total sales and marketing spend divided by new customers acquired. That formula is what most pages on this topic give you, and it's a fine starting point, but it isn't how operators actually run the business. The version you should be using is blended CAC: variable marketing spend in a month divided by new customers in the same month. The version that drives decisions is maximum allowable CAC: contribution margin two (CM2) per customer multiplied by your target payback period. Here is the full breakdown, the four CAC numbers that all sound similar but mean different things (blended, paid, NCAC, fully loaded), benchmarks from 35 DTC brands across $650M+ in revenue, the marginal-CAC framework most agencies skip, and a link to our free max-CAC calculator.

What is customer acquisition cost?

Customer acquisition cost is the dollar amount a business spends to acquire one new customer over a defined period. The cleanest operator definition is variable marketing costs (paid ads, influencer commissions, creative production for paid channels) divided by new customers in the same period.

That's blended CAC, what we'll spend the majority of this page on, because it's the number that holds up across attribution noise, channel mixing, and the iOS 14 / Android privacy environment. There are three related CAC numbers you'll hear (paid CAC, NCAC, fully loaded CAC), and we'll define each one before getting into the formulas, but the headline definition you should commit to is: blended CAC = variable marketing spend in a period, divided by new customers in that period.

For the full breakdown of blended versus paid CAC and when each one matters, see blended CAC vs paid CAC and the blended-vs-paid CAC gap by vertical.

The customer acquisition cost formula

Blended CAC, the version you should run the business on, has a single clean formula:

Blended CAC = (Variable Marketing Spend in the Period) ÷ (New Customers in the Period)

Variable marketing spend means paid ads, influencer commissions, creative production tied to those ads, and anything else that scales with how hard you're trying to acquire customers. It does not include the fixed costs that show up regardless of acquisition volume, agency retainers, marketing-team salaries, SaaS tools, software. There's a specific reason fixed costs don't belong in this formula and we'll come back to it.

Worked example. A Shopify apparel brand spends $40,000 on Meta ads and $5,000 on influencer commissions in a month. They get 250 new customers in Shopify Analytics. Blended CAC = $45,000 ÷ 250 = $180.

The trap most brands fall into is the denominator. If you include repeat purchasers, your CAC looks lower than reality. Use Shopify's "First-Time vs Returning Customers" report (or the equivalent in your CRM) and pull NEW only.

Why fixed costs are excluded from blended CAC. If you bundle salaries and tools into the CAC denominator, you can sub-optimize the wrong lever. The number moves when you fire a marketer or cancel software, neither of which actually changed your acquisition economics. Keeping the formula to variable costs only makes the metric responsive to acquisition decisions: when blended CAC moves, you can attribute it to creative, audience, or channel changes, not to HR or procurement noise. Fully loaded CAC (defined below) is still useful for board reporting, but it's the wrong number to scale spend against.

Blended CAC, paid CAC, NCAC, and fully loaded CAC

There are four numbers that all get called "CAC" in conversations and each one means something different. Get the definitions straight and the rest of this page makes sense.

Blended CAC. Variable marketing spend divided by new customers, both pulled for the same period. This is the steering metric. The denominator is your actual new-customer count from Shopify or your data warehouse, no attribution model needed, no platform-reported number trusted, no double counting. It's the only CAC that survives the iOS 14 measurement break because it doesn't depend on platforms telling you who they delivered.

Paid CAC. Paid acquisition spend divided by new customers attributed to paid. Essentially the same construct as blended CAC narrowed to paid channels. Use it as a channel diagnostic, the trend line on paid CAC tells you which channel is getting expensive faster than the others, but not as a steering metric for the whole business.

NCAC (New Customer Acquisition Cost). Attributed CAC to new customers using marketing attribution. This is what a tool like Triple Whale, Northbeam, or Rockerbox produces: a model-based estimate of which channel deserves credit for each new customer. It's a more granular version of paid CAC and useful for channel allocation when you're spending across 3+ paid channels. Blended CAC and paid CAC are essentially NCAC at a coarser grain, they're the same family of metric.

Fully loaded CAC. Variable marketing spend plus all fixed sales and marketing costs (agency retainers, software, team salaries, overhead), divided by new customers. It's the P&L number, what your CFO sees in management reporting. We don't use it much for operating decisions for the reason explained in the previous section: fixed costs aren't responsive to acquisition activity in the short term, so including them makes the metric noisy. Worth tracking quarterly for board context. Not worth scaling spend against.

Side-by-side for the same brand-month: blended CAC $58 (variable spend / new customers), paid CAC $45 (per Meta's number, narrowed to paid channels), NCAC $52 (per Triple Whale's attribution model), fully loaded CAC $73 (with team + tools + agency). All four are arithmetically correct. Which one do you scale spend against? Blended. Which one shows you Meta is getting expensive? Paid (or NCAC if you have a model). Which one goes in the board deck? Fully loaded.

What is a good CAC for ecommerce?

Industry averages are mostly useless because a "good CAC" depends on your contribution margin and your payback target, not on what the apparel industry averages. That said, here are the benchmarks from our 35-brand DTC portfolio so you have a reference point.

Read those numbers as orientation, not as your target. The brand running $34 apparel CAC isn't doing something magic; they have a strong organic engine (TikTok content, returning customers, referral) which drops the denominator-side mix. The brand at $112 is paid-heavy on Meta with mediocre creative. Most brands sit closer to the median.

A "good" CAC for YOUR brand is whatever number is below your max allowable CAC, which we get to in a couple of sections. If your max is $90 and you're at $60, you have room to scale. If your max is $50 and you're at $58, you're spending into a loss and need to either improve LTV or cut spend at the marginal cliff.

If you want full benchmarks by vertical with year-over-year movement, see our average CAC by ecommerce vertical post for the underlying dataset, and average CAC by revenue stage for how the number shifts as you scale.

Marginal CAC: the cost of your next customer

This is the section that matters most. Most pages on this topic stop at the average CAC formula. The reason most ecom brands are under-spending or mis-allocating spend isn't that they're reckless, it's that they're making decisions against the average instead of against the margin.

Marginal CAC (also called incremental CAC, next-dollar CAC, or "what's it cost to get the next customer") is the cost of the NEXT customer at your current spend level, not the average across all your customers. The formula:

Marginal CAC = ΔSpend ÷ ΔNew Customers

Why it diverges from average CAC: paid ads have diminishing returns. The first dollar finds the easiest customer to convert. The hundredth dollar is fighting harder for someone slightly less ready to buy. By the thousandth dollar, you're paying for impressions on people who never had intent to buy from you.

Worked example. A brand spends $20,000 on Meta and gets 200 customers, average CAC $100. They scale to $25,000 and get 230 customers; the extra $5,000 produced 30 customers, so marginal CAC at that step was $167. They scale to $30,000 and get 248 customers; the next $5,000 produced 18 customers, marginal CAC $278. The AVERAGE CAC across $30K is still $121, which looks tolerable. The MARGINAL CAC is $278, which is signaling stop.

If the brand's max allowable CAC is $90, they should have throttled back at the first step. They didn't notice because the average looked healthy. The mirror situation also happens, and more often: marginal CAC at the current spend level is $60, max allowable CAC is $120, and the brand is spending $20K when they could profitably spend $40K. They under-spent because they were watching the average and assuming it would scale up linearly. It usually does, until it doesn't.

How to measure marginal CAC weekly without a model. Pick one campaign. Increase spend 10–20% for 7 days. Compare incremental new customers to incremental spend over the same period. That ratio is your marginal CAC for that campaign at that scale. If it's above your max allowable CAC, throttle back. If it's below, scale more.

How to measure with a model. Lifesight, Rockerbox, and Recast all build response curves that estimate marginal CAC at every spend level. For brands at $5M+ in ad spend a media-mix model is worth the cost. Below that, the simple weekly test is fine and beats not doing it.

Maximum allowable CAC

Maximum allowable CAC is the ceiling above which acquiring a customer loses you money. This is the most important number on this page and the one that drives every spend decision.

The formula uses contribution margin two (CM2) per customer, not just contribution margin or gross margin. CM2 is revenue minus product cost minus fulfillment cost minus payment fees and returns reserve, the actual dollars a sale contributes after all variable order economics. If you don't have a CM2 number for your business, start there. See the contribution margin pillar for the full CM1/CM2/CM3 ladder, or the what is CM2? glossary deep-dive.

Max Allowable CAC = CM2 per customer × Target Payback Period (in CM2 units)

What target payback period to use. Two rules of thumb depending on your business model:

• Low-repeat ecom (most apparel, accessories, one-purchase categories): aim to break even on the first purchase. Max allowable CAC = CM2 of the first order. If you can't acquire profitably on order one, you're betting on a repeat rate you probably don't have.
• Subscription / high-repeat ecom: aim to break even by month 3 to month 6. Max allowable CAC = cumulative CM2 across the first 3 to 6 months of the cohort curve. Subscription brands can tolerate longer payback because the lifetime is more predictable, but only if you've verified the cohort retention, not just assumed it.

Worked example for a DTC apparel brand. AOV $89, CM2 per first order $31. Low-repeat business model. Max allowable CAC = $31. If actual blended CAC is $25, you have room to scale spend hard. If it's $40, you're acquiring at a loss every time and the only way to break even is the repeat purchase, which has to be real.

Worked example for a subscription brand. AOV $42, CM2 per order $18, average customer takes 3 months to hit churn cliff. Cumulative 3-month CM2 (assuming month-1 cohort returns: month 1: 100% of cohort × $18 = $18, month 2: ~75% × $18 = $13.50, month 3: ~70% × $18 = $12.60) = $44. Max allowable CAC ≈ $44 at a 3-month payback. At a 6-month payback (if cohort retention holds past the cliff) you can stretch higher.

Use the free maximum CAC calculator to run your own numbers. Inputs: CM2 per order, repeat rate by month, target payback. Outputs: max allowable CAC plus the implied LTV:CAC ratio.

CAC payback period

CAC payback is how many months it takes for the cumulative contribution margin from a customer cohort to recover the CAC. Most pages on this topic give you a flat formula:

CAC Payback = CAC ÷ Monthly CM2 per Customer

That flat formula assumes every customer comes back every month, which they don't. The honest version requires a cohort curve.

Here's how the cohort curve actually behaves. In month 1, you acquire a fresh cohort of 100% of your new customers. Their first-order CM2 covers part of the CAC. In month 2, somewhere between 10% and 50% of that cohort comes back and places a second order, this is the biggest cliff in the customer lifecycle and the single highest-leverage retention number to track. The CM2 from those returning customers covers more of the CAC. In month 3, somewhere between 50% and 90% of the month-2 returners come back again (much higher retention from month 2 onward), and their CM2 chips away further. By month X, cumulative CM2 across the cohort crosses the CAC line, and that's your payback period.

The implication: CAC payback is sensitive to the month-1-to-month-2 retention rate more than to anything else. A brand that gets 30% of customers to come back in month 2 has a fundamentally different payback profile than a brand that gets 10%, even at the same CAC and same CM2-per-order. Get them back to month 2 and the rest of the curve cooperates. The best DTC brands we work with see 95%+ retention from month 2 onward, but that whole curve is dictated by getting them past the first cliff.

Operator rules for ecom payback against the cohort curve:

• Low-repeat ecom: break even on the first purchase. If payback requires month 2+, you're betting on retention you can't budget for.
• Subscription / high-repeat: 3 to 6 months is workable if you have verified month-1-to-month-2 retention above 30%. Under 30% and your effective payback is much longer than the model predicts.
• Anyone with payback past 12 months on the cohort curve: you don't have a business, you have a fundraising treadmill.

For payback benchmarked across public DTC brands see CAC payback in public DTC 2026, the glossary definition in what is CAC payback period, and payback split by business model in CAC payback by business model.

For the SaaS-origin framing of payback see OpenView's CAC Payback Basics, the math is right, but apply it on a cohort curve, not a flat assumption.

CAC by channel: which channel is breaking first

Channel-level CAC is a diagnostic, not a steering metric. The blended CAC is what tells you whether the business works; the per-channel CAC is what tells you WHERE it's failing first.

Approximate 2026 ranges for DTC ecom brands, from our portfolio and confirmed by our agency partners:

• Meta paid: $35–$95 blended in most verticals. Wide variance driven mostly by creative quality and audience fit. Advantage+ is hiding the true cost of specific creative hooks; the only way to know is split testing.
• Google Shopping / Performance Max: $25–$80 in most verticals. Narrower spread than Meta because the intent signal is higher. PMax is a black box where brand search gets mixed with new acquisition, artificially lowering reported CAC. Strip out branded search before you call it acquisition.
• TikTok paid: $40–$110, high variance, very creative-dependent. CAC drops significantly when paired with a high-volume organic content engine; purely paid TikTok is volatile.
• Influencer (gifted + paid): blended $30–$150 depending on tier. Micro-influencers with strong creative briefs can hit the low end; sponsored celebrity posts the high end.
• Klaviyo / email: marginal CAC near zero once the list exists. The CAC is in building the list, which lives in your other channels.
• SMS (Attentive, Postscript): $5–$15 for SMS-attributed conversions. List build cost varies wildly.

The post-iOS14 attribution gap means platform-reported Meta CAC is 30 to 50 percent lower than the truth. Your real Meta CAC is closer to blended CAC times the percentage of new customers Meta drove (which you can verify with a post-purchase survey: "How did you hear about us?"). For more on the attribution side, see average CAC by channel for the cross-vertical numbers.

CAC for Shopify brands

Shopify-specific things distort the CAC number in ways generic CAC guides don't cover. Shop App attribution conflates organic Shop App traffic with paid (often inflates organic, deflates paid). Shopify Audiences influences your Meta CAC by sending Meta first-party signal that lowers the platform-reported CAC even if blended hasn't moved. Post-purchase upsell apps (One-Click Upsell, Rebuy) shift AOV, which changes max allowable CAC without changing CAC itself. Subscription apps (Recharge, Bold) move the denominator question: is a subscription customer a "new customer" once, or every renewal?

The cleanest Shopify CAC measurement is to ignore the platform's built-in attribution entirely. Pull total ad spend from your ad platforms, pull NEW customers from Shopify's customer report (filtered to first-time buyers), and divide. Triple Whale, Northbeam, and Rockerbox add post-purchase-survey attribution which is more useful for channel-level diagnostics but doesn't change the blended CAC truth.

CAC for Amazon sellers

Amazon is genuinely hard because you can't see exactly who your individual customers are, Amazon owns the relationship, not you. Calculating a clean blended CAC the way you can on Shopify is essentially impossible. Most Amazon sellers and operators are guided instead by two Amazon-native metrics: TACoS and ACoS.

ACoS (Advertising Cost of Sales) is Amazon ad spend divided by attributed sales from those ads. It's the campaign-level efficiency metric.

TACoS (Total Advertising Cost of Sales) is total Amazon ad spend divided by total Amazon revenue (ad-attributed + organic). It's the all-in number you steer the Amazon business on, and it's the closest proxy you'll get to a blended CAC equivalent on Amazon.

Trend lines on both: if TACoS is rising while revenue is flat, ad efficiency is degrading. If TACoS is falling while revenue is growing, organic discovery is doing the work and ads are catalyzing, exactly what you want. Anchor your Amazon CAC analysis on TACoS and ACoS rather than trying to back into a per-customer number from a denominator you can't see.

Subscribe & Save complicates the math the same way Shopify subscriptions do. Account for whether you're counting the first order or the cohort lifetime.

CAC for subscription DTC

Subscription DTC tolerates longer payback periods because the customer's lifetime value is more predictable than one-purchase ecom. But the entire model depends on one number: the percentage of your month-1 cohort that returns in month 2.

The biggest churn cliff in subscription is from month 1 to month 2. Whatever percentage of customers makes it past that cliff sets the slope of your entire LTV curve. From month 2 onward, retention typically holds up dramatically, we've seen subscription brands run 95%+ retention month-over-month once a customer is past the first cliff. The whole model becomes about getting them past month 1.

What that means for CAC budgeting:

• Track month-1-to-month-2 retention as your single most important number. It's a more reliable signal of LTV than any modeled lifetime estimate.
• Invest in the second-month experience. Onboarding, expectations setting, second-order incentive, anything that lifts month-1-to-2 retention is the highest-ROI lever in your business.
• Build CAC against the verified cohort curve, not the modeled one. If your data says 35% of customers come back in month 2 historically, build max CAC against that 35%, not against a "we hope 60%" assumption.

Free-trial and first-month-discount offers distort the CAC math because the discount cohort behaves differently from the full-price cohort. Build a separate CAC budget for the discount cohort, and don't apply the full-price retention curve to discount customers without verification.

Reactivation customers are not new customers. If a churned subscriber comes back, that's a retention win, not an acquisition win. CAC accounting should keep them separate.

CAC vs CPA, CPL, CPM, and MER

The acronyms all sound similar and the difference between them is the unit you're paying for.

CAC is the right metric for whether the business works. CPA is the right metric for campaign optimization, it's a per-order number, not a per-customer one, which is why a single new customer placing two orders gives you two CPAs but one CAC. CPM is for brand-awareness work and platform-inventory comparisons. MER has become more popular than CAC in 2025+ because it sidesteps the attribution mess entirely; total revenue and total ad spend are both verifiable numbers, no platform-reported math required. The downside of MER is it conflates new and repeat revenue, which is fine for some businesses (subscription, high repeat) and misleading for others (one-purchase categories).

For standalone definitions see what is MER and what is ROAS, or the head-to-head in ROAS vs MER vs blended CAC.

How to reduce CAC

The counterintuitive answer: most of the time, don't.

The most important thing you can do here is not think about CAC in isolation. CAC is a single output number. The actual machine that produces it is your full conversion funnel, end-to-end, across every source. The right operating move is to break the funnel down into its stages and let the math tell you where to work.

The full conversion funnel, stage by stage

For every traffic source (Meta, Google, TikTok, email, SMS, organic, influencer), the funnel looks roughly like this:

• Marketing impressions: how many eyeballs the spend bought you.
• Click-through rate: what percentage of those impressions clicked through.
• Site sessions: how many of those clicks landed on the site without bouncing.
• Add-to-cart rate: what percentage of site sessions added an item.
• Reach-checkout rate: what percentage of add-to-carts started checkout.
• Complete-checkout rate: what percentage finished and paid.

Multiply those percentages together with the source's CPM and you get effective CAC for that source. The reason "reduce CAC" is the wrong frame is that CAC is a derived number, it's the result of every stage in that funnel. You don't reduce CAC directly; you improve one of the stages and CAC falls out.

What this looks like in practice

Break each of your top 3 to 5 sources into the funnel above. Calculate the percentage at each stage. Compare across sources. The math will show you exactly where the leverage is: maybe Meta is bringing the cheapest sessions but the worst add-to-cart rate. Maybe Google has the best checkout completion but the smallest impressions volume. Maybe your $40K of TikTok spend is buying impressions that never click, and the fix is creative, not budget.

When you have the funnel broken down by source, you can use the math to figure out which dollar spent gives the highest return back. That's the actual answer to "how do I lower CAC", find the funnel stage with the most slope and the most addressable problem, fix that, and CAC moves on its own.

What we explicitly don't recommend

Cutting paid spend lowers your reported CAC and lowers your growth at the same time. That's not reducing CAC, that's reducing your business. The trap of "reduce CAC" thinking is that it pushes founders toward defensive spend cuts when the actual unlock is usually somewhere on the funnel they aren't looking at.

If you don't have the data infrastructure to break your funnel down by source and stage, that's the first thing to build. Triple Whale, Northbeam, Rockerbox, and Shopify's own analytics will get you 80% of the way there. The 20% remaining is post-purchase survey for source attribution, which is cheap and worth doing.

Sometimes the most important way to reduce CAC is to not reduce it at all, to look at the funnel, find the leverage, and improve a non-CAC number that pulls CAC down as a side effect.

Free CAC calculator

We already have an interactive maximum CAC calculator live. Inputs: CM2 per order, repeat-rate assumption by month, target payback period. Outputs: maximum allowable CAC, implied LTV:CAC ratio, health indicator. Use it to set the ceiling. Once you know your max, compare to your blended CAC and marginal CAC weekly and you have the full operating picture.

For benchmark inputs to plug into the model, see our ecommerce CAC calculator benchmarks.

We also have a Google Sheets version of the max CAC model with the cohort curve built in. Want it? Book a 30-minute call and we'll send it along plus run your numbers against the 35-brand DTC portfolio.

Conclusion

Blended CAC is what you run the business on: variable marketing spend divided by new customers, in the same period, no fixed costs muddying the signal. Maximum allowable CAC is what you cap spend at: CM2 per customer times your target payback period, break-even on order one for low-repeat ecom, 3 to 6 months on the cohort curve for subscription. Marginal CAC is what you scale or throttle on this week. The CAC payback question is a cohort-curve question, and the single most important number in the curve is your month-1-to-month-2 retention rate. Don't try to "reduce CAC" in the abstract. Break your funnel down by source and stage and let the math show you which lever moves the most.

Want help benchmarking your CAC against 35 DTC brands and finding the funnel stage that's silently leaking your money? Talk to a CFO.

Frequently Asked Questions

how do i calculate cac if meta and google attribution don't match?

Trust the blended number, not the platform-reported number. Calculate blended CAC as variable marketing spend (paid ads + influencer commissions + creative production) divided by new customers across all channels. Meta and Google each take credit for the same customers, which is why their numbers look low and add up to more conversions than you actually had. Blended CAC has full visibility because the denominator is just your actual new-customer count from Shopify. Use channel-level CAC as a diagnostic, but run the business on blended.

what's the difference between blended cac and paid cac and which one matters?

Blended CAC is variable marketing spend across all channels divided by all new customers. Paid CAC is paid acquisition spend divided by new customers attributed to paid. Blended is essentially the same family of metric as NCAC (new customer acquisition cost), just at a coarser grain, paid CAC and NCAC are paid CAC at finer granularity. Run the business on blended. Use paid CAC and NCAC for channel diagnostics, which channel is getting more expensive faster than the others.

what's a good cac for an ecommerce brand?

Depends on your CM2 per customer and your business model, not on an industry average. From our 35-brand portfolio: apparel DTC blended CAC median is $58 (top decile $34, bottom decile $112). Beauty median $42 (top $26, bottom $89). A good CAC is one below your maximum allowable CAC: CM2 per customer times your target payback period. For low-repeat ecom that means break-even on the first order. For subscription it means break-even by month 3 to 6 on the cohort curve.

what is marginal cac and why does it matter for my paid ads decisions?

Marginal CAC is the cost of the NEXT customer at your current spend level. Average CAC is what your existing customers already cost. They diverge because paid ads have diminishing returns: the more you spend on the same audience, the more each additional customer costs. Marginal CAC tells you whether to scale a channel up or down this week. If your marginal CAC is well below your max allowable CAC, you can spend more. If it's above, throttle back. Average CAC alone is too slow a signal to drive spend decisions.

how do i figure out my max allowable cac?

Multiply your CM2 (contribution margin two, revenue minus product cost, fulfillment, payment fees, and returns reserve) per customer by your target payback period. For low-repeat ecom brands, target break-even on the first purchase, max allowable CAC equals CM2 of the first order. For subscription or high-repeat brands, target break-even by month 3 to 6 on the cohort curve. Use the free max CAC calculator to run your own numbers.

how do i calculate cac in shopify or shopify plus?

Pull variable marketing spend over a period (paid ads, influencer commissions, creative production) from your platforms and invoices. Pull NEW customers (not repeat) from Shopify Analytics over the same period. Divide. The Shopify dashboard's built-in CAC uses Shopify's own attribution which conflates Shop App traffic with paid; it's directional but not the number to make decisions on. Triple Whale, Northbeam, and Rockerbox produce attribution-modeled NCAC at the channel level. Blended CAC from raw spend and raw new-customer count is always the cleanest steering number.

what's a good cac payback period for a dtc brand?

Payback is a cohort-curve question, not a flat number. For low-repeat ecom, target break-even on the first purchase, if your model requires waiting until month 3 to break even, you're betting on retention you can't budget for. For subscription or high-repeat, 3 to 6 months on the cohort curve is workable, IF your verified month-1-to-month-2 retention is above 30%. Past 12 months on the cohort curve means you don't have a business, you have a fundraising treadmill.

is an ltv:cac ratio of 3:1 actually realistic for ecom?

Yes, 3:1 over a one-year horizon is the best target for most ecom brands. Five-to-one is ideal, that's the level where the business compounds without paid acquisition pressure. Build your max CAC against the 3:1 over one year first, then aim to push the ratio toward 5:1 by improving CM2, raising AOV, or improving month-1-to-month-2 retention. The textbook 3:1 over 5 years is the SaaS framing; ecom should hold to that ratio across a tighter one-year window because acquisition is more expensive and lifetimes are shorter.

how do i reduce cac without slowing growth?

Counterintuitively, don't think about CAC in isolation. Break your full conversion funnel down by source: impressions to click-through to site sessions to add-to-cart to reach-checkout to complete-checkout. Compare percentages across sources. The math shows you which funnel stage has the most leverage and which dollar gives the highest return. Cutting paid spend lowers reported CAC and lowers growth at the same time, that's not reducing CAC, it's reducing the business. The actual answer to "lower CAC" is usually a non-CAC funnel improvement that pulls CAC down as a side effect.

how does ios14 affect my cac measurement?

Meta's reported CAC under-counts conversions because iOS users opt out of cross-app tracking; a customer who saw an ad on Instagram and bought a week later doesn't get credited to Meta. Your real Meta CAC is higher than Meta says. Blended CAC (variable spend over total new customers) doesn't depend on platform-reported attribution and is the only number with full visibility. Post-purchase surveys give you channel attribution. MMM tools like Recast give you a model-based version. Either way, never make a budget decision off platform-reported CAC alone.