Reader question: If a conversion is credited to a creator campaign, what exactly changes when the reporting window or attribution model changes?
A campaign readout can look like a dated object: a report run on September 30 says that a campaign generated a certain number of conversions. But the number is conditional. It depends on which earlier touchpoints were allowed to receive credit, how the credit was divided, which channels were eligible, and which clock defined the reporting date. Change one rule and the apparent story can change without a single customer repeating the action.
That is not a reason to distrust analytics. It is a reason to preserve the rules beside the result. A clean post-campaign readout should let a reviewer answer four questions: when did the touchpoint happen, when did the key event happen, which attribution settings were active, and when was the report extracted? Without those dates, “the campaign drove” is often stronger than the evidence.
Four clocks hide inside one conversion
Google Analytics uses several related controls that should not be collapsed into one generic “attribution window.” Its settings page distinguishes the reporting attribution model, the channels that can receive credit, and the key event lookback window. The same page separately describes conversion settings, including conversion windows and counting methods. Google Analytics attribution settings
| Clock or rule | The question it answers | What a reviewer should preserve |
|---|---|---|
| Click-through or engaged-view conversion window | How far back may a qualifying click or engaged view receive conversion credit? | The CTC or EVC value and the date it was configured |
| Key event lookback window | How far back may a touchpoint receive credit in Google Analytics key event reports? | The key event, window, and effective setting |
| Reporting attribution model | How is credit divided among eligible touchpoints? | The model and the report dimensions used |
| Reporting clock | Which date and time zone label the event or report row? | Event time, extraction time, property time zone, and any comparison account time zone |
The distinction matters because the rules have different scope. Google says a reporting attribution model change applies to historical and future data in reports using event-scoped traffic dimensions, while user- and session-scoped dimensions are unaffected. It says a key event lookback-window change applies going forward. Google also notes that Google Analytics uses the property time zone and Google Ads uses the Ads account time zone, which can create reporting differences. Google Analytics attribution settings
What the August 11 update changes
Google’s release log dates the conversion-window update to August 11, 2026. Eligible properties can use custom integer windows of 1–90 days for click-through conversions and 1–30 days for engaged-view conversions. The documented path is Advertising > Conversion management > Settings. Google Analytics release log The settings guide warns that conversion management may not be available to every property; check the actual property before promising the control. Availability guidance
That is a configuration capability, not a campaign outcome. A team can now align a window more closely to a considered purchase cycle, but the setting still defines eligibility for credit; it does not prove incremental lift, intent, or causation. It also does not mean every Analytics lookback setting has become the same control. The key event lookback window remains documented separately, with its own choices and forward-looking scope.
The release date and the property’s effective configuration are two different facts. A readout should record when the feature was announced, when an administrator changed the property, and which value the report actually used. If an analyst assumes that a new conversion-window value automatically rewrites older rows, that is an inference, not something established by the release note. Check the property’s change history and a bounded before-and-after report instead.
An illustrative timeline, not a result
The following is hypothetical. No campaign was run and no uplift is being reported.
Imagine a creator post published on September 2, 2026. A viewer clicks the tagged link at 12:00 UTC on September 5 and completes the defined conversion at 12:00 UTC on September 20. These hypothetical timestamps are exactly 15 days apart.
| Configuration | Eligibility for this illustrative click-through event |
|---|---|
| 7-day CTC window | Outside the window |
| 14-day CTC window | Outside the window |
| 30-day CTC window | Inside the window |
| 90-day CTC window | Inside the window |
If instead the September 5 touchpoint met the applicable platform's engaged-view qualification, with no click, the same 15-day gap would be outside a 14-day EVC window and inside a 30-day EVC window. Merely watching an ad does not establish that qualification. These labels classify an illustrative path, not a reported viewer's behavior.
Now add a second eligible touchpoint, such as a later search interaction. Under a data-driven reporting model, Analytics may distribute fractional credit between the touchpoints rather than place the whole key event in one row. Google’s documentation gives an example in which fractional key-event credit across touchpoints sums to 1.0. Google Analytics attribution settings The event did not happen twice. The allocation changed.
Build the readout as an evidence packet
For every campaign or creator link, keep a compact record with:
- publication timestamp, post ID, destination URL, and tracking parameters;
- candidate touchpoint time, key event time, and report extraction time;
- CTC or EVC conversion window, key event lookback window, eligible channels, and reporting model;
- Analytics property time zone and, if compared with Ads, the Ads account time zone;
- event count, attributed credit, and a status such as inside window, outside window, or unresolved;
- the query, export, or screenshot that shows the setting and the period it covers.
The useful unit is not just “conversions.” It is conversions under configuration X, measured on date Y, using touchpoints in period Z. If the report is rerun after a model change, retain the earlier extract and label the new one. A changed allocation is a revision to the readout’s explanation, not evidence that the underlying event count grew.
The stop condition is a missing date
Pause the conclusion when the report has a total but cannot show the active window, model, channel rule, or time zone. Do not fill the gap with a standard industry assumption. Mark the row unknown, retrieve the property change history, or narrow the claim to what the event log proves. If two systems disagree, first reconcile event time, time zone, conversion definition, and attribution scope before choosing the larger number.
The most credible campaign readout is therefore slightly less dramatic and much more reusable. It names the touchpoint, the event, the window, the model, and the extraction date. That makes a future comparison possible without pretending that a settings change was a new customer or that a credited conversion was automatically incremental.
Sources and limitations
- Google Analytics, “What’s new in Google Analytics” — checked September 2, 2026; supports the August 11, 2026 custom conversion-window update, including 1–30 day engaged-view and 1–90 day click-through ranges, the prior limitations, and the documented configuration path. Limitation: the release note describes the feature and configuration surface; it does not establish campaign lift or the historical treatment of every existing report row.
- Google Analytics, “Select attribution settings” — checked September 2, 2026; supports the distinct attribution controls, fractional credit, timing scope, and time-zone differences. Limitation: conversion management may not be available to every property; configuration, linked accounts, and report dimensions determine the live result. This is not a performance study.