How user acquisition teams’ incentives can hamper the fight against fraud, and 3 strategies to counter this
Originally published June 2019 (Mobile Dev Memo dates the guest post 15 July 2019). This preserves the original incentive argument and historical examples, rather than a current survey of fraud.
The article argues that the way user acquisition managers are measured can discourage them from investigating fraud. If you are judged only on paid installs and day 7 return on ad spend, fraud can leave your numbers looking healthy. The fix is to change the incentives.
I wrote this guest post for Mobile Dev Memo (the site has since moved from mobiledevmemo.com to mdm.co; the published title is “Mobile ad fraud can be solved with smarter team incentives”, dated 15 July 2019). It used contemporary fraud allegations to ask whether teams were rewarded for investigating suspicious results.
Why do incentives let fraud through?
The article describes teams judged on paid installs and early return on ad spend, often day 7, as a predictor of lifetime value. Click fraud that steals organic installs gets counted as paid traffic, and fake first purchases inflate day 7 revenue, so a manager’s metrics can look healthy with fraud present.
What are the three strategies?
- Include overall ROAS in evaluations. Judge managers on blended return (paid and organic together), not paid alone. It may reward organic growth they didn’t drive, but it reflects the company’s real return and gives them a stake in catching fraud.
- Backtest past ROAS. Track return at day 14 and day 30, not only day 7, and tie part of the evaluation to it. Teams can optimise on day 3 or day 7 but check downstream numbers for the truth. Its historical example from Adam Lovallo at LivingSocial involved realized revenue falling short of forecasts. The article says little of the gap was fraud; the backtest was a wider financial warning.
- Tie evaluations to overall revenue and daily active user growth. The argument is that a manager rewarded for real business growth has a reason to investigate fraud. One company used cost per daily active user increment as a KPI.
The conclusion: user acquisition is as much a financial function as a marketing one, and aligning incentives with the company’s financial health reduces the tendency to overlook fraud.
What has changed since?
Fraud tooling and iOS measurement have both changed. For a recent look at fraud data, see AppsFlyer’s fraud report summary, and for measuring honestly today, what a mobile measurement audit looks like.
Frequently asked questions
Why do user acquisition teams miss ad fraud? The article’s hypothesis is that rewarding paid install volume and early ROAS alone can make inflated attribution look successful.
How do you align incentives against fraud? The article suggests judging on blended ROAS, backtested returns beyond day 7, and overall revenue and daily active user growth.
Published
