GA4 Custom Conversion Windows: A Long Sales Cycle Guide for Canadian Businesses

Long sales cycles break default analytics settings quietly. A B2B enquiry in Vancouver might start with a search in March and close in June, and if your conversion window is shorter than that gap, the channel that started the relationship shows up as unproductive. Budget then moves away from it — for reasons that are entirely an artifact of a setting.

Google Analytics announced on August 11, 2026 that eligible click-through conversion windows can be set to any integer from 1 to 90 days, and engaged-view windows to any integer from 1 to 30 days. The flexibility is welcome; the risk is that people now pick a bigger number because bigger feels safer.

Map your conversion actions first

Different actions deserve different windows. Before opening any settings screen, list each conversion action and classify it:

• Immediate actions — phone calls, live chat starts, direct purchases. Short lag, short window.
• Considered actions — quote requests, consultation bookings. Medium lag.
• Research actions — guide downloads, pricing-page visits, newsletter signups. Longest lag, and the ones most damaged by short windows.
• Downstream outcomes — qualified lead, closed deal, imported from the CRM. These carry the true lag.

A single window applied across all four is a compromise that fits none of them.

Measure the lag distribution

Pull at least 12 months of closed deals from your CRM and calculate, for each, the number of days between first recorded interaction and the outcome. Then look at the distribution rather than the average — long-cycle data is usually skewed, so the mean is misleading.

What you want:

1. The median lag — the typical case.
2. The 75th percentile — where most of the volume is covered.
3. The 90th percentile — the practical upper boundary.
4. The tail beyond that — real, but not worth distorting your reporting to capture.

The following is an illustrative shape only, not measured client data: if the median is 18 days, the 75th percentile 34 days and the 90th percentile 61 days, a 60-day click-through window covers the large majority of genuine conversions without stretching attribution to the rare outlier.

Window-selection rules

• Anchor near the 90th percentile, not the maximum. The last 10% costs you clarity in exchange for very little coverage.
• Set per conversion action where the platform allows it, following the classification above.
• Keep engaged-view windows shorter than click-through windows in most cases — the interaction is weaker and the 1–30 day range reflects that.
• Do not exceed your data. If you only have six months of CRM history, a 90-day window is a guess dressed as a setting.
• Confirm eligibility and current labels in your own property before changing anything; interface wording and which properties are eligible both change over time.

Change control

A conversion window change is a measurement change, and it should be handled with the same care as a tracking change:

1. Record the current setting for every conversion action before editing.
2. Export a baseline of the last full reporting period.
3. Change one action's window at a time where practical.
4. Annotate the change date in your reporting so the discontinuity is visible to whoever reads the chart in six months.
5. Do not compare a pre-change period to a post-change period as if the setting were constant.
6. Tell stakeholders before the numbers move, not after they ask.

Ongoing monitoring

• Re-measure the lag distribution twice a year; sales cycles drift with market conditions and offer changes.
• Watch for double-counting when a lead converts on multiple actions in one journey.
• Reconcile platform conversions to CRM outcomes at least quarterly — the gap between them is the number that matters for budget decisions.
• Re-check the window settings after any account restructure or import, which are common ways settings quietly revert.

Questions to ask your agency

• What conversion window is set for each action, and why that number?
• What lag evidence supported it, and when was that evidence last refreshed?
• When did the setting last change, and is that annotated in our reports?
• How do you prevent double-counting across multiple conversion actions?
• How do platform conversions reconcile to our CRM outcomes?

If those answers are not readily available, the window length is not really the problem. Our reporting service is built to keep that reconciliation current, and our Vancouver PPC team sets windows from measured lag as part of every PPC engagement rather than leaving defaults in place.