Deividas here. CPA gets used constantly in performance marketing conversations and defined precisely almost never. People throw the number around in Slack, in agency calls, in Ads Manager, and half the time they're comparing it to a benchmark that doesn't apply to their business model at all. Here's what it actually means and, more usefully, how to know whether yours is any good.
CPA stands for cost per acquisition. It's the total amount you spent to acquire one customer. The formula is simple: total ad spend divided by the number of new customers acquired in that period. Spend $10,000 on Meta, get 200 new customers, your CPA is $50.
The "acquisition" part matters. CPA should refer to actual customers, not leads, not add-to-carts, not clicks. Some teams track cost per lead and call it CPA. That's a separate metric and a useful one, but confusing the two leads to decisions based on a number that doesn't measure what you think it does. If you're running an e-commerce brand, CPA is the cost to acquire a paying customer. Full stop.
These three metrics measure the same ad spend from different angles. ROAS tells you how much revenue you generated relative to spend. MER tells you how the whole marketing budget is performing against total revenue. CPA tells you what you paid to get one customer.
All three are useful. None of them on its own tells the full story. A low CPA looks good until you realise the customers it's buying never come back and have a lifetime value of $30. A high CPA looks bad until you account for the fact that those customers spend $800 over two years. CPA is an acquisition cost metric. Whether that acquisition cost is sustainable depends entirely on what happens after the customer buys.
The number that actually tells you whether your CPA is good is the ratio of customer lifetime value to cost per acquisition. LTV:CPA is the frame that puts your acquisition cost in context of the business model it's serving.
At Triple Scale, we use 3:1 as the floor for a healthy ratio. If a customer is worth $150 in lifetime value, you can afford to spend up to $50 to acquire them and still have a viable economics. A 5:1 ratio — spending $30 to acquire a customer worth $150 — is excellent and gives you room to invest more aggressively in acquisition without running into margin problems. Below 3:1, you're usually either spending too much to acquire or not converting customers into repeat buyers at the rate the business needs.
This is why comparing CPAs across businesses or even across verticals is largely meaningless. A $200 CPA might be a disaster for a $40 product with one purchase per customer. The same $200 CPA might be conservative underinvestment for a subscription product with a $1,200 LTV.
Your overall account CPA includes conversions from retargeting, from past customers who came back through a Meta ad, and from true new customer acquisition. These are very different numbers with very different implications for growth.
Blended CPA is what most platforms report by default. It looks good because it includes warm audiences who were going to buy anyway, and it's not the number that tells you whether your acquisition engine is working. New customer CPA is the number that matters for growth. If new customer CPA is rising while blended CPA stays flat, your prospecting is getting less efficient and your retargeting is masking it.
On Meta specifically, we target a new customer ROAS of 2x or better, which you can translate to a CPA target by working back from your average order value. At a $100 AOV, a 2x new customer ROAS means your maximum new customer CPA is $50 — that's the ceiling before the channel stops paying for itself at a basic level, before lifetime value is factored in.
CPA has three main inputs: how much you pay for attention (CPM), how many people who see the ad click it (CTR), and how many people who land on your site buy (conversion rate). Improving any one of these lowers CPA without changing your budget.
Creative quality drives CPM because Meta's algorithm charges less per impression for ads that earn strong engagement signals. It also drives CTR because a better hook and better copy pull more clicks from the same impressions. Landing page and offer quality drive conversion rate. Most CPA problems that operators treat as targeting or budget problems are actually creative or landing page problems in disguise.
There is no universal benchmark. The honest answer is: a good CPA is one that sits below one-third of your average customer's lifetime value, produces a new customer ROAS of 2x or better at the channel level, and holds stable or improves as you scale spend. That's it.
If someone tells you a "good" Meta CPA is $30 or $80 or $150 without knowing your AOV, your margin, and your LTV, they're guessing. Your CPA target should be derived from your unit economics, not borrowed from a benchmark that was calculated for a different business model on a different day.
For the measurement framework that puts CPA in context alongside MER and channel ROAS, Marketing Efficiency Ratio (MER): What It Is and Why It Matters More Than ROAS covers the full picture. For how conversion rate feeds the CPA equation, Facebook Ads Conversion Rate: What's Good in 2026 breaks down what healthy looks like by funnel stage. And for the scaling context around CPA thresholds, How to Scale Facebook Ads Without Killing Your ROAS addresses when the numbers justify pushing spend. The Triple Scale media buying course covers CPA-based account management as part of the full operating model.
