---
title: "What Is an Attribution Window? Lookback Explained · Appy"
description: "An attribution window is how long after a click or ad view an install can still be credited to it. How windows change your numbers and why reports disagree."
source: "https://appy.to/glossary/attribution-window"
---

1.  [Home](https://appy.to/)
2.  [Glossary](https://appy.to/glossary)
3.  Attribution window

Measurement and attribution

# What is an attribution window?

Short answer

An attribution window, also called a lookback window, is the length of time after a click or an ad view during which an install or event can still be credited to it. If the install happens inside the window, the source gets the credit. If it happens later, the install counts as organic.

Summarize with[ChatGPT](https://chatgpt.com/?q=Summarize%20this%20page%20and%20list%20its%20key%20points%3A%20https%3A%2F%2Fappy.to%2Fglossary%2Fattribution-window) [Claude](https://claude.ai/new?q=Summarize%20this%20page%20and%20list%20its%20key%20points%3A%20https%3A%2F%2Fappy.to%2Fglossary%2Fattribution-window) [Perplexity](https://www.perplexity.ai/search?q=Summarize%20this%20page%20and%20list%20its%20key%20points%3A%20https%3A%2F%2Fappy.to%2Fglossary%2Fattribution-window)

## How it works

Every attribution system has to answer one question: how old may a click be and still count? Someone taps an ad on Monday and installs the app on Thursday. With a seven-day window, the ad gets the install. With a one-day window, it does not.

Windows are usually set separately for clicks and for views. A view is weaker evidence than a click, so its window is shorter. When several clicks fall inside the window, most systems credit the last one.

Events have windows too. After an install is credited, later purchases are usually tied to the same source only for a limited time. Re-engagement campaigns, which bring existing users back, often use their own shorter windows.

## Why reports disagree

Two tools that look at the same installs with different windows will report different numbers, and both will be right by their own rules. An ad platform with a long window claims installs that your own tool, with a shorter one, files as organic.

Before comparing two reports, check three settings: the click window, the view window, and whether the install is dated by the click or by the first open. Most arguments about missing installs end there.

Time zones add to the confusion. A day that ends at midnight in one tool and at midnight UTC in another shifts installs between dates, which looks like a window problem and is not one.

## Windows the platforms set for you

Some windows are not yours to choose. The Play Install Referrer keeps its value for 90 days after the install. The Apple Ads attribution token is valid for 24 hours. SKAdNetwork and AdAttributionKit count an install only if the app is launched within Apple’s own attribution time window, and then measure engagement in fixed conversion windows.

A longer window is not a better one. It moves installs from organic to paid without adding a single user. Pick a window that matches how long people really take to decide, and keep it stable so that months can be compared.

Try it here

## Appy receives your SKAdNetwork and AdAttributionKit postbacks

Point your app’s Info.plist at Appy and Apple sends a copy of every winning postback here. Appy checks Apple’s signature and shows the results on your app page.

-   Installs, re-downloads and re-engagements
-   Results by ad network and campaign
-   Fine and coarse conversion values
-   One SDK call to set the conversion value

[Read the setup guide](https://appy.to/docs/adattributionkit) [See the Appy SDK](https://appy.to/features/sdk)

[Read moreMobile Attribution in 2026: What Still Works After Privacy Sandbox DiedA map of the measurement methods that still work in 2026, and the order in which to rely on them.](https://appy.to/blog/mobile-attribution-after-privacy-sandbox)

## Frequently asked questions

### What is a typical attribution window?

There is no single standard. Each ad platform and measurement tool has its own defaults, usually measured in days for clicks and a shorter time for views. Check the settings in each tool instead of assuming they match.

### What is the difference between click-through and view-through attribution?

Click-through attribution credits an install to an ad the person tapped. View-through attribution credits an ad that was shown but not tapped. Views are weaker evidence, so they get a shorter window and lose to a click when both exist.

### What happens to installs outside the attribution window?

They are counted as organic. The person may still have been influenced by the ad, but the system has no rule that lets it give the credit, so the install lands in the unattributed group.

## Related terms

-   [Install attribution](https://appy.to/glossary/install-attribution)

Install attribution is the step that credits a new app install to the link, ad or campaign that caused it, or marks it as organic when no source is found. It joins two moments that the app store separates: the tap before the install and the first time the app is opened.

-   [Mobile attribution](https://appy.to/glossary/mobile-attribution)

Mobile attribution is working out which ad, campaign or link brought an app install or an in-app action, such as a sign-up or a purchase. It can be deterministic, based on a click ID, the install referrer or the link that opened the app, or probabilistic, based on statistical matching.

-   [Organic install](https://appy.to/glossary/organic-install)

An organic install is an app install that is not credited to any ad, link or campaign. The person found the app through store search, a recommendation or the brand name. In practice, organic also contains every install that had a source the measurement could not see.


[All glossary terms](https://appy.to/glossary)

## One link for every phone

Create a free smart link and QR code that sends iPhones to the App Store, Android phones to Google Play and desktops to your website.

[Create a free link](https://appy.to/account/register)
