Click-to-install time (CTIT)

Click-to-install time (CTIT) is the time between a user clicking an ad and the app being opened for the first time on their device. It is one of the first metrics to check when you suspect attribution fraud in app install campaigns.

What is CTIT?

The name says “install,” but the actual end point is the first app open, because the measurement SDK only runs once the app is opened. AppsFlyer defines CTIT this way (AppsFlyer, Click to install time, accessed October 2026). Adjust also treats the first app open as the install.

For installs from Google Play, the interval can be split into two parts. The Play Install Referrer API returns the time of the click on the link that led to the store and the time the install started (Android Developers, Play Install Referrer Library, February 2026), and the SDK adds the time the app was opened. That tells you how long users thought it over on the store page, and how long they waited after the download before opening the app.

Why does CTIT matter?

Real clicks tend to turn into installs fairly fast. Adjust analyzed its data and found that more than 85% of installs from legitimate traffic happen within the first hour after the click (Adjust Help Center, Distribution Modeling, accessed October 2026). Fraud bends that distribution in two opposite directions:

• Too short: a click is slipped in right before the app opens, as with click injection and click hijacking. AppsFlyer describes this kind of analysis as usually looking for a pile-up of installs in the first 3 to 10 seconds after the click.
• Too long and flat: clicks are fired in bulk to “catch” organic installs, a scheme called click flooding. The distribution spreads almost evenly from hour 2 to hour 24 and from day 2 to day 7. Adjust starts rejecting attribution when most of a source’s conversions happen 60 minutes or more after the click.

The Media Rating Council (MRC) also requires companies that measure app installs to apply timing logic between impression, click and install, since an implausibly short gap can point to invalid activity (MRC, Invalid Traffic Detection and Filtration Standards Addendum, June 2020).

How to read a CTIT distribution

1. Export raw install data with the click time and the app open time, split by source, campaign and sub-publisher.
2. Sort installs into buckets: under 10 seconds, 10 seconds to 1 hour, 1 to 24 hours, over 24 hours.
3. Compare each source with the cleanest source for the same app, not with a single universal number. Large apps, or users on slow networks, naturally show a longer CTIT.

On iOS, this gets harder. For installs that can only be measured through SKAdNetwork or AdAttributionKit, the first postback is sent after a random delay of 24 to 48 hours once the conversion window closes (Apple Developer, AdAttributionKit, accessed October 2026). There are no per-user timestamps to build a CTIT distribution from.

Example (hypothetical)

An ecommerce app gets 2,500 Android installs from each network in a week, at a CPI of 12,000 VND. Network A has 70% of its installs in the first hour and 3% under 10 seconds. Network B has only 25% in the first hour, with 40% landing between day 2 and day 7. Network B’s 1,000 late installs alone add up to 12 million VND that needs to be reconciled, because they are most likely organic installs that the network took credit for.

CTIT vs. attribution window

See the attribution window entry for more.

Frequently asked questions

What is a normal CTIT?

There is no universal threshold. Use the Adjust benchmark (more than 85% in the first hour) as a reference point, then treat the distribution of the cleanest source in your own app as the baseline.

Does a short CTIT always mean fraud?

No. A user on a fast network installing a lightweight app can open it just over ten seconds after the click. What should raise suspicion is a source whose share of installs under 10 seconds is unusually high compared with your other sources.

If you need a CTIT report by source for your app install campaigns, talk to Adtek through the contact page.