# Why does the business report success the profit line does not show? > When a measure becomes a target, it stops measuring. A campaign can report a ROAS of 3.14 and lose money on its last pound. How to tell. Source: https://barkingcat.ai/thinking/why-reported-success-does-not-show-in-profit/ ## Thinking · November 2026 · Barking Cat series Because the targets measure activity, and activity can rise while profit falls. When a measure becomes the thing a team or a partner is paid against, it stops measuring the benefit and starts measuring how good people are at moving the measure. Traffic, conversion, ROAS and revenue growth are all reasonable things to watch. As targets, each can be hit by doing something that costs the business money. The test of a measure is simple: does it move with contribution? If a team can hit it while contribution falls, it is reporting success the profit line will not show. ## When a measure becomes a target Marilyn Strathern put the general version in a paper on university audit in 1997: “When a measure becomes a target, it ceases to be a good measure.” The mechanism is not dishonesty. People find the cheapest way to move the number they are judged on, and the cheapest way is rarely the one the business hoped for. ROAS is the clearest case to work through, because the arithmetic is short. ## The working The numbers below are an illustration. A campaign spends £100,000 and is credited with £400,000 of sales: a ROAS of 4.0. After the cost of the goods and of serving the orders, the business keeps 30% of each sale, so the campaign earns £120,000 of contribution and £20,000 after paying for the media. Now double the budget. The extra spend reaches customers who are harder to win, so sales rise by 57% rather than 100%. The second £100,000 brought in £228,000 of sales and £68,400 of contribution, so it lost £31,600, while the campaign’s ROAS still reads 3.14. On a 30% margin, the break-even ROAS is 3.33: one divided by the margin. Below that line, every pound of media loses money, whatever the ratio looks like. This flatters the campaign, because it assumes every credited sale was caused by it. Some were the brand searches and the cashback visits that would have happened anyway. ## Two explanations, and the one that leads The first is attribution: the last click gets the credit, and the last click can be a customer who had already decided. That makes a ROAS figure too generous, and it is a real problem. The second is that ROAS is an average, and budgets are set at the margin. I think this one leads, because it misleads even when attribution is perfect: the working above assumes every credited sale was caused by the campaign, and the last £100,000 still loses money. What would prove me wrong is a holdout showing that the last tranche of spend causes more contribution than it costs. Where that is true, spend it. Both halves are visible in practice. I once inherited an agency that was beating its ROAS target while missing its sales target, and at another business, cutting paid media by 30% took revenue down by roughly 20% and left the bottom line stronger. ## What to do first Work out the break-even ROAS for each category from its contribution margin, after returns. It is a line of arithmetic, and it shows which campaigns are below water before anyone runs a test. Then measure what the spend causes. A holdout, with comparable regions or weeks where a campaign is switched off, shows the sales that disappear without it. The cheaper middle route is to step the budget down by 10% and read the result before going further. The [Contribution and Media Efficiency workbook](https://barkingcat.ai/thinking/contribution-and-media-efficiency/) shows what a 10%, 20% or 30% cut does to contribution before you make one. ## How you will know The measure is contribution after media on the sales the campaign caused, read against the holdout. The guardrail is new customers acquired, so that a cut which keeps the profit but stops the business finding new customers does not count as a win. Match the holdout and test periods for season, promotions and stock, and read them after at least one full buying cycle for the category. Agree before you start that if contribution after media falls by more than the media saved, the spend goes back. ## The decision The decision for the board is what each team and partner is paid against, and whether each of those measures moves with contribution. Two questions for Monday. Which of our targets could a team hit this year while contribution falls? And what did the last £10,000 of our paid media earn? This rests on one inherited agency, one cut and an illustration. It cannot tell you how much of the 20% of revenue would have come back over time, and I have no holdout from that business that would say. ## Sources - Marilyn Strathern, [‘Improving ratings’: audit in the British University system](https://www.gwern.net/doc/statistics/decision/1997-strathern.pdf), European Review, vol. 5, no. 3, 1997, pages 305 to 321 (the quotation is on page 308) Related: [The Barking Cat problem](https://barkingcat.ai/thinking/the-barking-cat-problem/); [Contribution and Media Efficiency](https://barkingcat.ai/thinking/contribution-and-media-efficiency/). ## About the author Steve Webster is an independent adviser to the boards and investors of UK retail and consumer businesses from £20m to £1bn: an operator’s judgement on ecommerce, technology and AI, before the board commits the capital. Formerly Global Ecommerce Director at Barbour, with 27 years running retail P&Ls up to £240m. ## Contact steve@barkingcat.ai Barking Cat Advisory Limited. Registered in England and Wales, company number 17338003.