What it means
A person clicks an advertisement, considers a purchase and buys six days later. A seven-day click window can include that sale, assuming the tracking system links the events, while a one-day click window would not count it for that interaction.
Start with the interaction type, because a click is not an ad view and an engaged view is not an ordinary impression, so a headline about "seven-day attribution" is incomplete without naming the interaction. Google Ads defines a conversion window as the period after an ad interaction during which a conversion is recorded.
Google Analytics describes its lookback window as the period in which a touchpoint remains eligible for attribution credit. Those are related concepts, but the reporting settings and terminology are platform-specific, so check the current settings rather than assuming a remembered default, and document the chosen window and the date it took effect.
Next define the conversion, since a purchase, lead form and subscription renewal have different buying cycles. A longer window can include more delayed conversions for a channel while a shorter one excludes later conversions, but more credited conversions do not mean the advertisements caused more sales.
The settings changed the accounting of credit, not necessarily buyer behaviour. Consider multiple touches.
Someone sees an ad on Monday, clicks search on Wednesday and buys Friday, and both interactions might fit their windows, so the attribution model decides whether one, both or neither receives credit under the platform's rules. Cross-platform totals may also overlap, because each advertising platform can claim the same sale, and adding all platform-attributed purchases is not the same as counting unique orders in the shop's transaction system.
Name the reporting date too. Some systems align conversions with the interaction date, while others align them with the date of purchase, which can make daily reports differ even when the same sale is eventually counted.
Google Ads points users to time-lag reports when choosing a window, but those reports may only include interactions the tool can observe, so apply judgment to untracked offline or cross-device journeys. Changes may affect future reporting without rewriting all past figures, because Google Ads explains that a new conversion window applies going forward.
Label breaks in the trend rather than explaining an apparent change as campaign performance alone. For an owner, the window is a measurement rule, not a lever that creates sales, so use it to understand what the performance report counts and connect that report to revenue and costs recorded outside the advertising dashboard.
In practice
Real-world examples.
Example
A shopper clicks an ad and buys six days later. A seven-day click window can include that purchase; a one-day click window would not credit that click. The advertiser therefore sees a very different conversion count for the same campaign depending on a single setting.
Example
A video platform reports engaged-view conversions and a search platform reports click conversions. The team labels both interaction types and their windows rather than comparing the headline totals alone. It then records both settings in the monthly report so that next quarter's comparison uses the same rules.
Example
Two advertising dashboards each claim the same order. Finance counts one unique paid order and treats platform attribution as overlapping explanations, not two sales. The monthly revenue figure comes from the shop's transaction system, and the dashboards are used only to compare channels.
Formula
Calculation
Illustrative eligibility test: conversion timestamp - eligible interaction timestamp <= selected window, subject to platform rules and identity matching. A seven-day click window includes a purchase six days after a qualifying click, but excludes one nine days later. Attribution allocation requires a separate model.
Now check overlap. Suppose platform A claims 120 orders, platform B claims 90 orders, and the shop's system records 150 unique paid orders, every one of which was claimed by at least one platform. The overlap is 120 + 90 - 150 = 60 orders claimed twice. At an average order value of $50, the platforms together claim 210 x $50 = $10,500, but the shop's real revenue is 150 x $50 = $7,500.Case study
Seen in the real world.
This entirely fictional case follows Vale Homeware, an invented online shop. One ad dashboard used a longer click window than another, and both reported credit for some of the same orders. Management initially added their conversion totals together and concluded that advertising had produced more orders than the shop had actually shipped. The analyst documented each interaction type, window and reporting date, then reconciled both dashboards against unique paid orders.
The shop used the reports to compare channels without claiming every attributed sale was incremental. The analyst also noted the date on which each window changed, so that a jump in reported conversions after a settings change would not be mistaken for better campaign performance. The case and numbers are invented.
Watch out
Common mistakes.
- Comparing reported return on ad spend across platforms without checking windows and conversion definitions.
- Adding platform-attributed conversions together as though overlapping credits were unique orders.
- Assuming that a longer attribution window caused the extra credited sales.
Questions
People also ask.
Is an attribution window the same as an attribution model?
No. The window controls how far back interactions remain eligible; the model allocates credit among eligible interactions.
Which window should a business choose?
Use the buying cycle, conversion goal and decision being made. Document the choice and compare like with like.
Do attributed conversions prove an ad caused a sale?
No. Attribution assigns credit under reporting rules; causal impact requires additional evidence.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%