Prop Firm CAC and ROAS: What Good Actually Looks Like

Why blended ROAS is the wrong steering metric

Blended return on ad spend divides all revenue attributed to paid by all paid spend. It is useful as a board-level summary and close to useless as an operating metric, for three reasons.

It averages across tiers. A firm selling evaluations from 10K to 200K has an enormous spread in order value. A blended figure that improves might reflect nothing more than the mix shifting toward larger accounts, with acquisition efficiency flat or worse.

It averages across channels. In our Q1 data the blended figure was 4.79 times. Inside that, Google reached 8.63 in its best month and Meta 6.35 in its. The blended number told us the quarter was fine. The channel split told us where the next euro should go.

It ignores what happens after the purchase. A challenge sale is not the end of the customer relationship. It is the beginning of an obligation. A trader who passes and gets funded generates a payout liability. A trader who fails and retries generates a second sale. Neither shows up in the number you are steering by.

Aggregated from active prop firm accounts, anonymised under our confidentiality agreements. Past performance does not guarantee future results.

That spread is the point. A single quarter contained a month at 6.07 and a month at 3.66. If you only look at 4.79 you never ask what happened in March, and the answer to that question is usually where the next improvement lives.

The metrics that actually govern the economics

Cost per acquisition, split by tier

Not blended. A 10K evaluation and a 100K evaluation have different prices, different buyer sophistication and different pass rates. If your CPA is reported as one number you cannot tell whether the campaign that looks efficient is efficient or simply selling the cheapest product.

Track CPA per tier and compare it against the fee for that tier. The ratio is the number that tells you whether a campaign is contributing.

Pass rate by acquisition source

This is the metric almost nobody tracks and the one with the largest effect on real profitability. Traders acquired through different creative angles pass at meaningfully different rates.

An earnings-led ad attracts people responding to the earnings promise. They tend to be less experienced, fail faster, and either churn or retry at a lower rate. A rules-led or platform-led ad attracts traders evaluating a product. They pass more often, which increases your payout liability, and they also retry and refer more.

Neither is automatically better. What matters is that they are different, and if you do not segment pass rate by source you are blind to a variable that changes the entire calculation.

Retry rate

Most traders fail their first evaluation. A significant share buy another. Retry revenue arrives with no additional acquisition cost, which means it is the highest margin revenue in the business, and it is invisible in a first-purchase ROAS figure.

A firm with a 40 percent retry rate and one with 15 percent can support entirely different acquisition costs while reporting identical blended ROAS on first purchase.

Payout ratio

The share of revenue leaving as trader payouts. This is the closest thing prop firms have to cost of goods sold, and it is the reason challenge revenue cannot be treated as margin. A campaign that produces excellent ROAS by acquiring traders who pass at a high rate can be less profitable than one producing lower ROAS at a lower pass rate. Uncomfortable, and true.

Contribution per acquired trader

The number that ties it together. First purchase revenue, plus expected retry revenue, minus expected payout liability, minus acquisition cost. Modelled at the tier and source level, this is the only figure that tells you whether to scale a campaign.

Where the money actually leaks

Attribution double counting

Meta claims a conversion. Google claims the same conversion. Both report to their own dashboards. Sum the platform figures and your reported revenue exceeds actual revenue, often by twenty to forty percent in a multi-channel account. Every downstream decision made on those numbers is wrong in the same direction.

Brand cannibalisation

Traffic that would have arrived anyway, bought at auction and reported as incremental. Performance Max is the usual culprit, and it is covered in more depth in Google Ads for prop firms.

Tier mix drift

Algorithms optimise toward the conversion that is cheapest to produce, which is normally your smallest evaluation. Left alone, smart bidding will quietly shift your mix downward. Revenue stays flat, volume rises, margin falls, and the dashboard reports improvement.

Creative decay treated as a bidding problem

Costs rise, the response is a bid or budget adjustment, performance degrades further. The underlying cause was fatigue. Diagnosis is in creative fatigue in prop firm ads.

Downtime

The largest single line item, and the one that never appears in a report. Two weeks of a suspended account is not two weeks of paused spend. It is lost revenue plus a rebuild period at elevated acquisition cost. See why prop firm ad accounts get banned.

Setting a target you can hold campaigns to

Work backwards from contribution rather than forwards from a ROAS number someone quoted on a podcast.

1. Establish average revenue per acquired trader by tier, including expected retry revenue over a twelve-month window.
2. Subtract expected payout liability using your actual pass rate and profit split for that tier.
3. Subtract processing, platform and support cost per trader.
4. What remains is your acquisition budget per trader. Divide by tier price to get a target ROAS that means something.
5. Set the threshold below that for scaling campaigns, above it for pausing, and hold the range for a full evaluation cycle before judging.

Firms that do this generally discover their real target is lower than they assumed, which means they had been throttling campaigns that were profitable. Occasionally it is higher, which means they had been scaling campaigns that were not. Both findings are worth more than another month of optimisation against a number nobody derived.

What to look at, and how often

Daily: spend pacing and delivery status. You are checking that nothing has broken, not making decisions.

Weekly: CPA by tier and channel, creative performance, search term leakage. This is the operating rhythm.

Monthly: contribution per acquired trader, tier mix, channel allocation. This is where budget moves.

Quarterly: pass rate and retry rate by acquisition source, payout ratio trend. This is where the strategy changes.

Making budget decisions on daily data is the single most expensive habit in prop firm media buying. Conversion windows in this category run long enough that a bad Tuesday is noise, and reacting to it destroys the algorithmic learning that produces good Thursdays.