How to set a target CPA before you launch a paid campaign

Set your target CPA before launch by working backwards from what a customer is worth, not forward from what you feel you can afford. Take the lifetime value of a customer, adjust it for gross margin, then divide by the return you need. Most teams want three dollars back for every dollar spent, so a customer worth $1,000 in margin gives you a target CPA near $333. That number is the ceiling. You set it on paper before the algorithm ever touches your budget.

Skip this step and you launch blind. You watch the cost per conversion climb, you sense it's too high, but you have no line that tells you when to pause or when to push. A target CPA is that line. It's the most you can pay to acquire a customer and still come out ahead, and it turns every later decision from a gut call into arithmetic.

Start with what a customer is actually worth

Lifetime value is the input everything else hangs on. The simple version: average revenue per account, times gross margin, times how long a customer stays. Lifespan is roughly one divided by your monthly churn rate. Say you sell a $50 a month plan, your gross margin is 80%, and 4% of customers leave each month. That gives an average lifespan of 25 months and a margin-adjusted LTV of $1,000 ($50 x 0.80 x 25).

Use the margin-adjusted number, never raw revenue. Revenue LTV counts money that goes straight back out in hosting, support, and payment fees, so it flatters the math and pushes you to overspend. The $1,000 is what a customer is genuinely worth to you. That's the figure you get to divide up.

Turn lifetime value into a target CPA

Now split that value between you and acquisition. The common benchmark is an LTV to CAC ratio of 3 to 1: for every dollar you spend winning a customer, you want three dollars of margin back. On our $1,000 example, that puts the target CPA at about $333.

The ratio is a dial, not a law. Slip below 2 to 1 and the spend stops paying for itself once you add overhead. Climb above 5 to 1 and you're usually leaving growth on the table, sitting on margin you could reinvest to win more customers. Three to one is the working default for most SaaS; pick your own number with eyes open, then hold it.

Convert the target CPA into a bid you can hold

A target CPA is a per-customer number, and the ad auction charges you per click. Bridge the two with your conversion rate. Max CPC equals target CPA times the rate at which a click becomes a customer. If your landing page turns 2% of clicks into customers, your ceiling bid is $333 x 0.02, or about $6.66 a click.

Most funnels have more than one step, so multiply the rates. If 10% of clicks become leads and 20% of leads become customers, your real click-to-customer rate is 2%, the same $6.66. Getting this wrong is how teams bid $15 a click on a page that can only carry $6, then wonder why the channel bleeds. Fix the funnel math before you touch the bid.

Give the system enough conversions to work

If you hand your target CPA to an automated bid strategy, it needs data to hit it. Google's Target CPA bidding asks for at least 15 conversions in the last 30 days to switch on, recommends 30, and does its best work above 50. Under that, the algorithm is guessing, and it will happily blow past your target while it learns.

Small campaigns often can't feed it. If you're below 30 conversions a month, run manual or maximize-clicks bidding, cap your CPC at the number you calculated, and treat the target CPA as a ceiling you enforce by hand. Volume also decides how much you should spend to learn anything at all, which is a separate calculation worth doing first. We covered it in how much budget you need to test a paid channel.

Revisit the number, not every day

A target CPA set at launch is a starting estimate built on assumed churn and an assumed conversion rate. Both move. Recalculate when your margin changes, when churn settles into a real figure after a few months, or when you find the early conversion rate was optimistic. Do it on a schedule, not in reaction to one noisy week.

The other reality check is time. A healthy target CPA still costs you cash up front and pays it back over months, so the ceiling only makes sense next to how long that payback takes. We walked through the timing in when paid acquisition actually pays back. Set the number, hold the bid, and let the arithmetic decide what to scale.

If you want help turning your unit economics into paid targets you can actually run a campaign against, that's the kind of work we do.