← All insights

The unit economics every performance marketer should know

Break-even CPC, ROAS, margin of safety and why a profitable test can still be a bad business. A practical glossary with worked examples.

Most campaigns that fail do not fail because of bad creative. They fail because the numbers never worked in the first place. Before we spend a rupee, dollar or pound on traffic, we model the economics. Here are the measures we use, with worked examples.

Break-even cost per click

The single most useful number. It tells you the most you can pay for a click before losing money.

Break-even CPC = payout × conversion rate

Example: a lead pays $7 and your landing page converts 10% of visitors. Break-even CPC is $0.70. If your traffic costs $0.45 a click, you have room to make money. If it costs $0.90, you need a better conversion rate, a higher payout or cheaper traffic before scaling.

ROAS and ROI

ROAS (return on ad spend) is revenue ÷ ad spend. A ROAS of 1.5 means every $1 spent returns $1.50. ROI also accounts for all other costs (tools, tracking, creative, time). A campaign can show a positive ROAS and still lose money once software subscriptions and creative production are included. We track both.

Margin of safety

Payouts change, conversions get reversed for quality, and ad costs rise in competitive seasons. We aim for a cushion of at least 25–30% between our actual cost per conversion and the payout before we consider a campaign scalable. Without that cushion, one bad week can wipe out a month.

A quick reference

MetricFormulaWhy it matters
CTRclicks ÷ impressionsDoes the ad earn attention?
CPCspend ÷ clicksWhat attention costs
CVRconversions ÷ clicksDoes the page deliver?
CPAspend ÷ conversionsCost of one outcome
EPCrevenue ÷ clicksValue of one click
ROASrevenue ÷ spendEfficiency of media
Profitrevenue − all costsThe only one that pays salaries

Think in portfolios

No single campaign is guaranteed to work. We run a small number of offers at a time, give each a fixed test budget, and judge them on the same rules. Winners get more budget gradually; losers are stopped without drama. Over time, the portfolio's results matter far more than any individual bet.

Quality is part of the economics

A lead that is later rejected is worth nothing, and too many rejected leads can end a partnership. That is why we never use incentives that attract uninterested people, never pre-fill forms with guessed data, and never promise things the advertiser does not offer. Clean traffic is not just ethical. It is the most profitable kind.