How much can you afford to spend to acquire a customer? Up to your profit per order before ad cost. Most stores never work this out, so they scale themselves into a hole. That ceiling has a name: your maximum allowable cost per acquisition (CPA). The formula: (average order value × gross margin %) − shipping − payment fees − packaging. Whatever is left is the most you can pay for one customer and still profit. Spend below it and you win. Spend above it and every sale loses money, no matter how busy the dashboard looks. Here is the exact maths, a worked Rand example, and how lifetime value raises that ceiling, from V8 Media, the team behind R2+ billion in client sales since 2018.
The mistake that quietly kills profitable-looking stores
This is eCommerce mistake number eight, and it is the most expensive one on the list.
Most founders never work out what a customer is actually worth to them. So they never know what they can afford to pay to win one.
They just spend. They boost a post, watch a few sales roll in, and call it a win.
But "a few sales" is not the same as profit. Not even close.
Here is the thing. Anyone can launch a store. Building a store that actually makes money is a different sport.
And the founders who win all do the same boring thing first. They figure out the most they can pay to get a new customer, before they spend a single rand on ads.
Get that one number wrong and you can scale yourself straight into a hole. Faster than Eskom burns diesel.
What "what you can afford to spend" actually means
It means your maximum allowable CPA. The most you can pay to land one customer and still come out ahead.
Marketers have a name for it. Dave Chaffey's digital marketing glossary calls it the allowable cost per acquisition: the maximum amount you are prepared to pay to acquire a customer within your margins.
The idea is dead simple. Every sale has a profit ceiling. Your ad cost has to sit below it. If it does not, every order bleeds money. Simple as that.
So the real question is never "how much should I spend on ads?" It is "how much is one customer worth to me?"
Answer that, and the ad spend question answers itself.
The formula for your max allowable CPA
Here is the one to write on a sticky note and stick to your monitor.
Max allowable CPA = (AOV × gross margin %) − shipping − payment fees − packaging.
Start with what an order is worth. Subtract the product cost. Then strip out shipping, fees, packaging. Whatever is left, that is the ceiling. Every rand you pay Meta or Google has to fit under it.
This is the same maths the big eCommerce playbooks teach. Common Thread Collective and Shopify both frame it the same way: your acquisition cost has to sit under your gross profit per order, or the first sale loses money.
Let's put real numbers on it.
A worked Rand example
Say you run a store with these numbers.
- Average order value (AOV): R1,000.
- Product cost (COGS): R400 per order. That is a 60% gross margin.
- Shipping: R80 per order.
- Payment fees: roughly 3% of R1,000 = R30 (PayFast, Yoco and Stripe sit around there).
- Packaging and a small returns allowance: R90.
Gross profit is R1,000 − R400 = R600.
Now strip the rest: R600 − R80 − R30 − R90 = R400.
That R400 is your max allowable CPA. The ceiling on what you can pay for one customer on the first order.
Pay R250 to win that customer and you profit R150. Pay R400 and you break even. Pay R550 and you lose R150 on every single sale.
Now picture the founder who never did this. Their ads are pulling customers at R550 a head. The store is "growing". Orders are coming in.
And they are losing R150 every time the till rings. The more they "scale", the faster the money burns.
That is not a traffic problem. It is a numbers problem wearing a traffic problem's costume.
Your margin sets the number, not a benchmark
People always ask for "the right CPA". There isn't one.
Your margin decides it. A fat margin can afford a big CPA. A thin margin can barely afford anything.
Here is what the ceiling looks like across common margins, on a R1,000 order, with about R200 of shipping, fees and packaging stripped out.
| Gross margin | Gross profit per order | Max allowable CPA |
|---|---|---|
| 20% | R200 | R0 (you cannot afford ads at all) |
| 30% | R300 | R100 |
| 40% | R400 | R200 |
| 50% | R500 | R300 |
| 60% | R600 | R400 |
| 70% | R700 | R500 |
Illustrative: max allowable CPA = gross profit per order − ~R200 variable costs (shipping, ~3% payment fees, packaging) on a R1,000 AOV.
Look at the 20% row. After costs, there is nothing left to pay for ads. That store cannot run paid traffic profitably on the first sale. Full stop.
This is why two stores can run the identical ad and one prints money while the other goes broke. Same ad. Different margin. Different reality.
So before you ever ask what to spend, fix your margin. Not sure where yours sits? Start with our guide to eCommerce profit margins and benchmarks.
CPA vs CAC: same idea, different scope
You will hear two terms thrown around. CPA and CAC. People use them like they mean the same thing. They almost do.
The difference is what you count.
| Term | What it counts | Best used for |
|---|---|---|
| CPA (cost per acquisition) | Ad spend ÷ number of sales. Usually just the platform cost. | Judging a single campaign or channel |
| CAC (customer acquisition cost) | All sales and marketing cost ÷ new customers. Fully loaded. | Judging the whole business |
CPA is the ad-level number. CAC is the real, fully loaded number. NetSuite and Triple Whale both stress that a true CAC should include more than ad spend: software, agency or staff costs, the lot.
For day-to-day ad decisions, CPA is your fast feedback loop. For "is this whole business healthy?", CAC is the honest mirror.
Watch both. But never let a pretty CPA hide an ugly CAC.
The benchmark trap
People love a benchmark. So here is one, then here is why you should not lean on it.
Across categories, the average cost to acquire an eCommerce customer sits roughly between $68 and $90, and it swings hard by vertical, per 2026 benchmark data from Eightx.
| Category | Typical CAC (global, USD) |
|---|---|
| Food & beverage | $53–$100 |
| Apparel / fashion | $90–$120 |
| Beauty | ~$110 |
| Electronics | $100 and well up |
Global directional benchmarks, Eightx "Average CAC by eCommerce Vertical 2026". Use as a sense check, not a target.
Useful as a gut check. Useless as a target.
Why? Because a R1,600 CAC is a triumph for a store selling R8,000 furniture sets and a death sentence for a store selling R300 phone cases.
The benchmark does not know your margin. You do. Your max allowable CPA beats any industry average, every time.
Lifetime value: how to raise your ceiling
So far we have judged a customer on one order. That is the cautious way to play it. It is also leaving money on the table.
Because most customers do not buy once. They come back.
Lifetime value (LTV) is the total profit one customer brings over the whole time they buy from you. And it changes everything.
Back to the example. One order banks R400 of profit. So your first-order ceiling is R400.
But say the average customer buys three times before they drift off. Now that customer is worth about R1,200 in profit, not R400.
Suddenly you can afford to spend far more than R400 to win them. You might "lose" R100 on the first sale and still walk away well ahead by the third.
This is how big brands outbid everyone else. They are not reckless. They already know the customer will come back. So they can afford to spend more upfront. You are bidding for one order. They are bidding for three.
The healthy yardstick most use is an LTV to CAC ratio of about 3 to 1. Shopify and Common Thread Collective both point to roughly 3:1 as the sign of a sustainable store. Earn three rand in lifetime profit for every one rand you spend acquiring the customer.
Want to go deeper on this number? Read our full guide to eCommerce customer lifetime value. It is the single biggest lever on what you can afford to spend.
Target CPA vs max allowable CPA
Quick but important distinction. Your max allowable CPA is your break-even ceiling. It is not your goal.
Run your ads right at the ceiling and you make zero profit. You are just trading rands with Meta for fun.
So set a target CPA below the ceiling. That gap is your profit per customer.
In the example, the ceiling is R400. Set a target CPA of R250 and you bank R150 of profit on every first order. Then LTV stacks more on top as they reorder.
The rule is simple. Ceiling tells you where you start losing. Target tells you where you want to live. Always live below the ceiling.
How to actually lower your CPA
Knowing your number is step one. Getting your real CPA under it is step two. Here is where the wins hide.
- Raise your average order value. Bundles, upsells and a free-shipping threshold ("free delivery over R750") spread the same ad cost across a bigger sale. Bigger order, more room for CPA.
- Fix your conversion rate before your ads. If your site turns 1% of visitors into buyers and a rival turns 3%, they can pay triple your CPA and still win. Fix the funnel first.
- Sharpen the offer. A stronger offer converts cheaper traffic. The ad is rarely the problem. The offer usually is.
- Mine repeat customers. Email and WhatsApp flows bring buyers back at near-zero cost, which lifts LTV and lets you afford a higher CPA on the first sale.
- Cut wasted ad spend. Kill the campaigns above your ceiling. Pour budget only into the ones below it. Boring. Brutally effective.
This is the lens we run every client's Meta Ads and Google Ads through. Every campaign gets judged against the number, not the vibe.
It also pairs with the wider eCommerce marketing math every store should know, and the ROAS vs POAS debate that decides whether you scale on revenue or actual profit.
How V8 Media uses this number
Most stores that come to us think they have an ad problem. They almost never do.
They have a "we never worked out what a customer is worth" problem. The ads were doomed before they launched.
So before we touch an ad account, we work out the real margin, the max allowable CPA, and the LTV. Then we only scale what banks money.
No guessing. No "let's just push more budget and hope". Just the number, and campaigns that beat it.
It is the same thinking behind everything we run, since 2018 and R2+ billion in client sales. The goal is never the biggest revenue screenshot. It is the most profit in your account.
Frequently asked questions
How much can I afford to spend to acquire a customer?
Up to your profit per order before ad cost. Take your average order value, multiply by your gross margin, then subtract shipping, payment fees and packaging. What is left is your maximum allowable CPA. Spend less than that to win a customer and you profit; spend more and you lose.
How do I calculate my maximum allowable CPA?
Max allowable CPA = (AOV × gross margin %) − shipping − payment fees − packaging. On a R1,000 order at a 60% margin, with R80 shipping, R30 fees and R90 packaging, your ceiling is R600 − R200 = R400 per customer.
What is the difference between CPA and CAC?
CPA (cost per acquisition) is usually your ad spend divided by the number of sales, so it is the platform-level cost. CAC (customer acquisition cost) is fully loaded: all sales and marketing costs, including software and staff, divided by new customers. Use CPA for campaign decisions and CAC to judge whether the whole business is healthy.
What is a good customer acquisition cost for eCommerce?
There is no universal number. Global benchmarks put the average eCommerce CAC around $68 to $90, but a "good" CAC is any figure below your max allowable CPA. Your margin and lifetime value decide it, not an industry average.
How does lifetime value change what I can spend?
If a customer buys more than once, their lifetime value is higher than a single order, so you can afford a higher CPA. A customer worth R400 on the first order but R1,200 over three orders lets you spend more upfront and still profit. Many stores target an LTV to CAC ratio of about 3:1.
Why is my store making sales but no profit?
Most likely your real cost to acquire each customer is higher than your profit per order. Revenue is not profit. Once product cost, shipping, payment fees and VAT come out, the sale keeps far less than it looks, and if your CPA sits above that, every order loses money.
Should I include VAT and salaries in the calculation?
Include VAT (15% in South Africa) in your margin if you are registered, because it comes out of every sale. For your fully loaded CAC, include staff, agency and software costs too. For a quick max allowable CPA on a single order, the core variable costs (COGS, shipping, fees, packaging) are enough to start.
What is the difference between target CPA and max allowable CPA?
Max allowable CPA is your break-even ceiling, the point where profit hits zero. Target CPA is the lower number you actually aim for, and the gap between the two is your profit per customer. Always run your ads below the ceiling, not at it.
Key takeaways
- What you can afford to spend to acquire a customer is your profit per order before ad cost, your max allowable CPA.
- Max allowable CPA = (AOV × gross margin %) − shipping − payment fees − packaging.
- Your margin sets the number. A 20% margin store often cannot run paid ads profitably on the first sale at all.
- CPA is the ad-level cost; CAC is the fully loaded business cost. Watch both.
- Lifetime value raises your ceiling. A repeat customer is worth far more than one order, so you can spend more to win them.
- Set a target CPA below the ceiling. The gap is your profit. Ignore generic benchmarks; your margin wins.
Not sure what you can actually afford to spend on ads?
Most stores bleed out one "winning" campaign at a time because nobody did this maths. We do it first. Then we only scale what actually makes money. Not what looks good in a report. We have driven R2+ billion in client sales since 2018. See how we grow eCommerce stores profitably, or get a free audit of your Meta Ads and Google Ads.
Claim Your Free Audit
