The attribution window sets how many days a sale still counts for an ad. Too short, it misses sales that take time to happen. Too long, it credits the ads with what would have happened anyway. The choice depends on the buying cycle.
The essentials in a definition
The attribution window is the period during which a conversion can be attributed to a prior click or view. Choosing between 1, 7 and 30 days depends on the purchase cycle: short windows for quick decisions, long for lengthy ones. The setting influences reports and, in many cases, optimisation, as described in the measurement documentation.
What it is and why it matters
It is the period, after the click or the view, during which a conversion is associated with the ad. It changes the numbers reported, comparisons between channels and, in many cases, the way the algorithm learns.
Two identical campaigns with different windows appear to perform differently. So it is wise to always compare using the same window.
Linking the window to the buying cycle
For impulse, low-value purchases a short window (1 day) may be enough. For medium-consideration products (appliances, services), 7 days is common. For long decisions (B2B, property, training), 30 days or more may make sense.
Analyse the distribution of time between first contact and purchase in your data. Choose a window that covers most sales without including an irrelevant tail.
Choosing the attribution window
- Measure time between first contact and purchase
- Choose a window that covers most sales
- Use the same window in comparisons
- Review when the buying cycle changes
Clicks versus views
A click-through conversion is a stronger signal than a view-through one. Many platforms allow different windows for each. In general, shorter view windows reduce the risk of over-attribution.
Avoid mixing conversion types without distinguishing them in reports.
The right window follows the time customers take to decide.
Consistency and review
Define the company’s standard window and record it. If you change it, note the date and avoid comparing before and after periods without adjustment. Review annually or when the buying cycle changes (for example after launching a new offer).
For important decisions, validate with incrementality tests, which do not depend on the window.
Decision table
| Purchase cycle | Suggested window | Note |
|---|---|---|
| Impulse purchase | 1 to 7 days | Most happen quickly |
| Considered mid-value purchase | 7 to 30 days | Comparison and decision |
| High-value or B2B purchase | 30 days or more | Several contacts and decision-makers |
| Urgent services | 1 day or less | Immediate decision |
Common mistakes to avoid
- Window longer than the cycle. Attributes sales that did not depend on the ad.
- Window shorter than the cycle. Ignores legitimate conversions and distorts optimisation.
- Changing the window without notice. Breaks the comparability of reports.
How to measure the result
Calculate the interval between first contact and sale, by product and channel, and choose the window that covers about 80% of conversions. Review every six months.
Limits and exceptions
Choosing the window has limits: the purchase cycle varies by customer and product, and the average hides extreme cases. Long windows increase attribution uncertainty, and some platforms impose limits. Assess by cohorts, compare windows and accept that none is perfect. Document the choice and reasons, and review when the product or market changes.
Illustrative example
Illustrative example: a B2B company with an average 45-day cycle uses a 7-day window and concludes that LinkedIn “does not work”. Analysing the real interval between first contact and sale in the CRM, it sees most sales arrive 30 to 60 days later. It adjusts the window, measures by cohorts and now assesses the channel by value after 90 days.
Frequently asked questions
What is the best window?
It depends on the buying cycle. Measure the average time between first contact and purchase.
Can I have different windows by campaign?
You can, but it makes comparison harder. Prefer a common rule with justified exceptions.
Does lengthening the window improve results?
It increases attributed conversions, but not necessarily real sales.
How do I find my purchase cycle?
By calculating the time between the first recorded contact and the sale in the system, by segment.
Does the window affect automatic optimisation?
In many cases yes, because it defines the conversions the system sees; check the campaign-type documentation.
The attribution window should follow the real purchase cycle, measured in your data, not be chosen by default; changing it without a log breaks historical comparison.
Sources and recommended reading
Last editorial review: October 2026. Informational content; always confirm current conditions in the official sources before deciding.