TapCub is live — analytics, insights and live chat, free on one platform
Glossary

What is click-to-install time (CTIT)? How short CTIT flags install fraud

Click-to-install time is the gap between an ad click and the install. How it is calculated, why a spike of very short values signals install fraud, and a worked example you can recompute.

Sophia Bennett Reviewed by Oliver Reed Also called CTIT Last updated Nov 20, 2026 Glossary

Quick answer

Click-to-install time (CTIT) is the time between a user clicking an ad or tracking link and the app registering the install, usually at first open. A cluster of installs seconds after the click suggests click injection or install hijacking; a flat spread over hours and days suggests click flooding.

Formula: CTIT = install timestamp − click timestamp

What is click-to-install time?

Click-to-install time (CTIT) is the time between a user clicking an ad or tracking link and the moment the app counts that user as an install. Mobile attribution tools record both timestamps for every attributed install, so the gap can be computed for each one. A single CTIT tells you little. The shape of the distribution across hundreds of installs from one source is what separates real users from fraud.

The clock matters. In the most common setup the click time comes from the tracking link and the install time is the first time the app opens and its measurement SDK reports in, which is why a downloaded but never opened app is not an install. Google Play's Install Referrer API exposes two timestamps in seconds, the referrer click time and the time the install began, so a "click to install begin" figure can differ from a "click to first open" figure by however long the user waited to open the app. Always check which pair of timestamps your measurement tool subtracts before comparing numbers across tools.

How is CTIT calculated?

CTIT = install timestamp − click timestamp, per install. Reports then summarize it as a median, a set of time buckets or a full histogram.

Install timestamp usedWhere it comes fromWhat a short CTIT meansWhat a long CTIT can be
Install beginInstall Referrer API, installBeginTimestampSecondsDownload started right after the clickThe user left the store page and came back later
First openMeasurement SDK, first launchDownloaded and opened within secondsThe user downloaded on Wi-Fi hours earlier, or opened the app only the next day

Google states that the install referrer information is available for 90 days and does not change unless the app is reinstalled, and recommends reading it once, the first time the app runs after install. In practice that means the click and install timestamps are fixed once at first launch, so a CTIT worked out from them does not change later.

Why does a short CTIT flag install fraud?

A real person needs time between tapping an ad and having a working app: the store page loads, the download runs, the app opens. A click that arrives a few seconds before an install is suspicious for a mechanical reason, not a statistical one. Malware on the device can detect that an install has begun and then fire a click so the install is credited to the fraudster's source (click injection or install hijacking). The click is genuine as an event but it did not cause the install.

Fraud shows up in two different shapes, and they are mirror images of each other.

PatternWhat the distribution looks likeLikely causeWhat to check next
Spike at the very startLarge share of installs in the first few seconds after the clickClick injection or install hijacking, where a click is fired during an install that was already happeningShare of installs under about 10 seconds per source
Long, flat tailInstalls spread almost evenly across hours or days after the clickClick flooding or spamming, where a very large number of clicks are sent in the hope that some match a later organic installShare of installs after 24 hours, and whether the shape is flat instead of falling
Normal shapeMost installs within minutes, then a falling tailReal users clicking and installingStill compare against your own history

AppsFlyer's glossary says analysts look for a cluster of installs within a few seconds of the click when checking for install hijacking, and for a nearly flat distribution across hours and days when checking for click flooding. These are one vendor's working heuristics, not a standard: vendors use different cutoffs and some make them configurable, and a short CTIT is a signal to investigate, not proof by itself.

Worked example: two sources, same spend

Sample data: two paid sources each delivered 1,000 attributed installs for $5,000, so the reported cost per install is $5,000 ÷ 1,000 = $5.00 for both. Their CTIT buckets are very different.

CTIT bucketSource A installsShareSource B installsShare
Under 10 seconds121.2%38038.0%
10 seconds to 1 hour86186.1%42042.0%
1 to 24 hours10310.3%11011.0%
Over 24 hours242.4%909.0%
Total1,000100.0%1,000100.0%

Source B has 38.0% of its installs inside 10 seconds against 1.2% for A. That alone does not convict B, so add a second signal from after the install. Sample day-1 retention by bucket for source B: under 10 seconds, 5 of 380 returned (1.3%); 10 seconds to 1 hour, 134 of 420 (31.9%); 1 to 24 hours, 31 of 110 (28.2%); over 24 hours, 6 of 90 (6.7%). Overall that is 176 of 1,000 = 17.6%, against 26.1% for source A (261 of 1,000). The installs the CTIT test flagged almost never came back, which is what hijacked installs look like.

What this does to the price you pay: if you refuse payment for the 380 installs under 10 seconds, source B's effective cost per install is $5,000 ÷ (1,000 − 380) = $8.06, and if you also drop the 90 installs after 24 hours it is $5,000 ÷ 530 = $9.43. The "cheap" source cost 61% more per install than reported, and source A, at $5.00, is the better buy.

How to use CTIT in practice

  1. Read it per source, never as one global number. A global median of two minutes hides a source with 38% of installs inside 10 seconds. Break the distribution down by media source, campaign and, where you have it, publisher.
  2. Use buckets, not an average. Fixed buckets (under 10 seconds, 10 seconds to 1 hour, 1 to 24 hours, over 24 hours) are easy to compare week to week. An average is dragged by a few late openers.
  3. Pair it with a post-install signal. Day-1 retention, a first key event or revenue. CTIT says when the click happened relative to the install; the post-install signal says whether a person was there.
  4. Compare against your own baseline. A utility app with a one-tap install and a 2 GB game will have different normal shapes. Take the last 4 to 8 weeks of organic-adjacent sources as your reference.
  5. Set rules and write them down. For example, "a source with more than 10% of installs under 10 seconds is held for review". Choose a rule before looking at a specific partner's numbers, so you do not tune it to the result.

Common misreadings of CTIT

  • Treating one threshold as a law. The 10-second line is a heuristic. Fast networks and small apps can have real installs in under 10 seconds. Look for a spike, not one install.
  • Assuming a long CTIT means fraud. Many people tap an ad, download on the way, and open the app that evening. A smooth, falling tail is normal; a flat one at scale is the warning.
  • Comparing across tools without checking the clock. "Install begin" and "first open" can differ by hours. Compare like with like.
  • Judging with too few installs. With 30 installs a single injected click is 3.3% of the data. Wait for hundreds per source.
  • Treating CTIT as the only fraud signal. Mismatched devices, abnormal retention and implausible geographies are all stronger together than any one rule alone.

What TapCub does and does not do with CTIT

TapCub is not an attribution provider and does not match clicks to installs, so it does not compute CTIT. What it does is show what the installs from each source do afterwards, which is the second signal in the example above.

  1. Get the install data in. If you use a mobile measurement partner, point its postback at the receive address in App analytics › Configuration › Business settings. TapCub stores the callback as an $app_install event and reads the media source, campaign and match type from it.
  2. See installs by source. App analytics › Quality & attribution › Marketing attribution lists installs by media source and campaign, split into organic, non-organic and unattributed, and shows whether an MMP is connected. Take the sources your MMP flagged with a short CTIT and find them here.
  3. Check what those installs did. App analytics › Users › Channels shows, for each $channel value your SDK stores, average daily new users and next-day retention, plus a matrix of volume against stickiness. If your build stores the campaign in $channel, a flagged campaign with next-day retention far below its neighbours confirms the pattern from the example. Note that $channel is not the media source on an attribution callback, so this works only where you set it.
  4. Keep your own devices out. Test phones can be excluded with a device rule in App analytics › Configuration › App settings, so your own QA installs do not count as users.

CTIT thresholds and the decision to dispute installs stay in your attribution tool; TapCub measures the in-app outcome.

Sources

Was this explanation clear?

We use your answers to fix and expand this page.

Sophia Bennett

Product Adoption Consultant

Explores customer onboarding, live conversations, and self-service, turning common usage questions into practical guidance and improvements.

FAQ

Questions people ask

Short answers to the follow-up questions that come up most.

Still have a question?

Chat with the team behind TapCub — we usually reply within a few hours on business days.

What is a good CTIT?

There is no universal good value because it depends on app size and network. Most real installs fall within minutes of the click, followed by a falling tail. Judge a source by its distribution and by your own baseline, not by one number.

How does a short CTIT indicate fraud?

A person needs time to load the store page, download and open the app. When a large share of installs is credited within a few seconds of the click, malware may have fired the click during an install already under way (click injection or install hijacking).

Does a long CTIT mean click spam?

Not by itself. Real users often download now and open later. The warning sign is a nearly flat spread of installs across hours and days at scale, which fits click flooding, and a check on post-install retention.

Does TapCub measure click-to-install time?

No. TapCub is not an attribution provider and does not match clicks to installs. It can receive your attribution partner's install callbacks and show installs by media source and campaign, and next-day retention by the channel value your SDK stores.

See this number on your own site

Add one script tag for web and apps, and the first reports fill in within minutes.

See your data clearly. Find your growth.

Every click, backed by data. Install one line of code and see your first numbers in a minute.

No credit card · Free plans for analytics and chat · Cookieless analytics