A short-drama streaming app needed paying trial users at scale, without the incentivized and bot traffic that floods the category. In three months, our Mobile DSP nearly tripled the share of installs that became paid trials and cut fraud-flagged conversions by 92%.
StoryTV is a short-drama streaming app: vertical, episodic stories built for the phone. Viewers start with free episodes, and the business grows when they convert to a paid trial and then a subscription.
That makes the paid trial the moment that matters. StoryTV pays for that event on a CPA basis, not for installs, and every conversion is validated against the client's CRM and AppsFlyer Protect360.
Entertainment apps with free content attract low-intent installs, and CPA campaigns attract fraud: incentivized users, click spamming and bots that fake the paying event. In the first full month, one in ten paid-trial events was blocked by Protect360, and only one in twenty installs became a paid trial.
Pushing more budget into that mix would have scaled the problem with it. The job was to raise intent and remove fraud at the same time.
of paid-trial events blocked as fraud (Aug 2026)
install-to-paid-trial rate (Aug 2026)
We treated quality as the growth lever, not a constraint. Instead of chasing installs, we optimized every bid toward the probability of a validated paid trial, and we pruned sources on evidence, not on cost.
Bidding models trained on validated trial payments, not installs or registrations.
Publisher-level checks on click-to-install time, click spamming, IP, proxy, hosting and device signals, cross-checked daily against Protect360.
Budget moved to sources that held both conversion rate and fraud rate, week after week.
Android campaign live across India on the Mobile DSP; baseline measured.
Sources with fraud or weak trial rates cut; spend concentrated on proven supply.
Clean sources scaled; weekly paid trials more than doubled versus launch.
3.6x weekly paid trials versus launch week, with fraud below 1%.
The two numbers that mattered moved in opposite directions: conversion up, fraud down.
We applied the same approach to Story Max, StoryTV's sister app, with an even sharper result.
install-to-paid-trial rate (4.1% → 16.3%)
paid trials, on only +10% more installs
fraud-flagged share (8.5% → 0.4%)
Optimizing to the paying event, not the install, is what moved conversion rate. Volume followed.
Fraud control is a growth lever: every blocked fake conversion is budget returned to real users.
A playbook proven on one app transfers: Story Max saw a 4x conversion lift with the same method.
Methodology: figures from StoryTV's AppsFlyer data for the StrongMetrics campaign. Install-to-paid-trial rate = attributed trial_payment_successful events ÷ attributed installs, per month. Fraud-flagged share = Protect360-blocked paid-trial events ÷ attributed paid-trial events. Weekly index = average daily paid trials per week ÷ launch-week average × 100. Percentages and multiples only; absolute volumes withheld.
Tell us your paying event and target CPA. We'll show you where your conversion rate and fraud rate can go.