OTT and streaming app metrics
The metrics that show whether a streaming app turns installs into viewers and viewers into subscribers who stay, with the formula behind each.
The key OTT metrics are cost per install and per subscriber, first-play activation, time to first video, watch time per viewer, D7 and D30 retention, subscriber churn (voluntary and involuntary), trial-to-paid conversion, ARPPU, LTV and payback. Churn and watch time are read together, because low watch time is the earliest churn signal.
What are the key metrics for an OTT app?
The key metrics for an OTT or streaming app follow a viewer from install to first play, to regular watching, to a paid subscription that renews.
- Acquisition: cost per install and CAC per subscriber.
- Activation: activation rate on first video played, and time to first value in minutes.
- Engagement: watch time per user and session length.
- Retention: D7 and D30 retention and subscriber churn.
- Monetisation: subscription conversion, ARPPU, LTV and payback.
First play: the first video a new user starts after installing. It is the usual activation event for streaming apps because nothing else in the app delivers value before it.
How do you measure churn for an OTT app?
Count subscribers at the start of the month who cancelled or failed to renew by its end, and divide by the starting count. Split voluntary churn (cancellations) from involuntary churn (failed payments), because one needs better content or pricing and the other needs a billing fix.
- AcquisitionCost per install and per subscriber
- ActivationFirst play, and how fast
- EngagementHow much they watch
- RetentionWho renews, who leaves, and why
- MonetisationTrial to paid, price and lifetime
Involuntary churn: subscribers lost to failed payments or expired cards without choosing to leave.
How the OTT metrics connect
A subscriber's cost is CPI divided by install-to-trial and trial-to-paid rates; their value is the monthly price times margin divided by monthly churn. The two meet in payback.
- Cost per subscriber = CPI ÷ (install-to-trial × trial-to-paid).
- Subscriber LTV = monthly price × gross margin ÷ monthly churn.
- Payback = cost per subscriber ÷ (monthly price × gross margin).
- Watch time feeds churn: subscribers who watch little in a month are the ones most likely to leave at renewal, so watch time per subscriber is the leading indicator and churn the lagging one.
The calculator runs this chain on example inputs. Try raising trial-to-paid or lowering churn by a few points to see which moves payback more.
Example inputs: change them to your own numbers. The examples only make the calculator show a worked answer; they are not benchmarks or targets. Runs in your browser; nothing you type is sent anywhere.
- Cost per subscriber
- $50
- Subscriber LTV (gross margin)
- $70
- Payback period
- 7.1 months
- LTV to cost per subscriber
- 1.4x
What to watch in each OTT metric
- First play, fast. Autoplaying a trailer or opening on the title a user clicked in the ad shortens time to first value. Ads that promote one title should land users on that title, not the home screen.
- Watch time per subscriber, not per viewer. Per viewer measures depth among people who watched; per subscriber also counts paying users who watched nothing, which is the churn risk.
- Release-driven cohorts. Users who join for one big title often leave when it ends. Compare retention and churn for cohorts acquired around the same release, not across months with different slates.
- Trials measured after they end. Trial-to-paid for a cohort is only final once every trial in it has ended.
- Plan mix in ARPPU. Annual plans, discounts and family plans all move ARPPU without any change in what users watch. Report ARPPU by plan.
- Ad-supported tiers. If a free tier earns from ads, include ad revenue in ARPU and LTV, or report ad ARPU separately.
Where OTT numbers go wrong
- Autoplay counted as watching. Previews that play on their own inflate watch time and first-play activation. Count them separately.
- Churn on the wrong base. Including subscribers who joined during the month in the starting count lowers churn artificially.
- ROAS read too early. Monthly subscriptions mean revenue arrives slowly; D30 ROAS is usually the first meaningful read, and payback on margin is the real test. The App Attribution Guide's chapter on postbacks, events and ROAS explains how subscription revenue events reach ad platforms.
Further reading
Linkrunner's Metrics that Matter: OTT, media and streaming edition covers CAC, LTV, churn and content engagement for subscription and ad-supported models, and its list of events every OTT app should track sets out the events behind these formulas.
Measure the OTT and streaming tree by channel
Questions
What is a good north star metric for a streaming app?
Weekly viewing hours, or weekly viewers who finish an episode, are common choices. Both reflect value delivered and tend to move before renewals do.
How do you calculate OTT subscriber LTV?
Multiply the monthly price by gross margin and divide by monthly churn. At $10 a month, a 70% margin and 10% monthly churn, that is $70.
Why separate voluntary and involuntary churn?
They have different fixes. Voluntary churn responds to content, pricing and engagement; involuntary churn responds to payment retries and card updates.
Does watch time predict churn?
It is a common early warning sign: subscribers who stop watching are the ones who cancel at the next renewal. Check the link in your own data by comparing churn for low and high watch-time groups.
Why do tools report different values for the same metric?
Because definitions differ: what counts as active, which attribution window applies, whether organic users are included and how days are cut by timezone. Align the definition before comparing numbers.
Which currency do the calculators use?
The calculators show US dollars, but every formula works in any currency: enter your own figures in the currency you report in and read the results in that currency.
More verticals
- EdTech app metricsThe ten metrics that tell you whether a learning app is working, how each is calculated, and how they connect from the first ad to a paying learner.
- Fintech app metricsThe metrics that show whether a payments, lending, investing or savings app turns installs into funded, active customers, and how each one is calculated.
- Mobile game metricsThe metrics game teams use to judge a game and its traffic: from what an install costs, through progression and retention, to daily revenue and ROAS.