Guide 170 · Conversions · GA4 + Tag Manager
Key event counting method: once per event against once per session, and which to pick
Google recommends once per event, and labels once per session as Legacy. The difference is exactly what happens when one person triggers the same key event five times in one session: five counts, or one. This guide works through that five-trigger session, shows where to change the setting, and records why the change only affects future data.
After the steps, the generate_lead key event appears under the Key events tab in Admin, and the standard reports show it within a day. The worked example below gives the count and the rate to expect.
Download the reference: the generate_lead key-event reference (JSON) (803 bytes, MD5 5af51c2a952d4d62e40d4a2026bbfdaa). It holds the event name, the counting method, the worked-example figures recomputed for this guide, and the checkpoints that prove the key event is recording. The figures in the file are the same figures shown on this page.
What you need
- Marketer access or above at property level. Choosing a counting method needs the Marketer role; Editor alone is not enough for this setting.
- A key event already marked, so the counting method has something to apply to. In this guide that event is generate_lead.
- A written reason for the choice. The setting is per key event, not per property, so each event gets its own decision.
The trace
- Open Admin, then Data display, then Events, then the Key events tab. The tab lists the marked events. Find the row for
generate_lead. - Open the row menu and choose Change counting method. The menu sits at the far right of the row. If the option cannot be selected, the signed-in account does not hold Marketer access or above.
- Read the two options as Google states them. Once per event counts a key event every time it is triggered, and Google marks it Recommended. Once per session counts a key event only once within a particular session, and Google marks it Legacy.
- Work through the five-trigger session before choosing. A visitor submits the lead form five times in one session. Once per event records 5 key events; once per session records 1. The difference is 4 key events, and neither setting deduplicates the person across sessions.
- Pick once per event for purchases and for leads you want to count in full. Every purchase adds revenue, so each one must count. For a lead form, once per event also counts every submission, which is the count the business compares against its own inbox.
- Record the change date. A change applies to future key events only and never recalculates past data. Write the date, the event, the old method, and the new method in the measurement log, or a later trend break has no explanation.
Dead ends
DEAD END — USING ONCE PER SESSION TO FIX A DUPLICATE TRIGGER. If the form fires twice per submit because the tag is wrong, once per session hides the second fire inside one session and leaves the duplicate in every other session. Fix the trigger; do not hide it with a counting method.
DEAD END — EXPECTING THE CHANGE TO REWRITE HISTORY. Switching methods does not restate last month. A report that spans the change mixes the old method before the date with the new method after it, which reads as a sudden rise or fall.
DEAD END — COUNTING ONCE PER SESSION TO COUNT PEOPLE. Once per session counts at most one per session, not one per person. A visitor who returns in a second session and submits again counts again. There is no counting method that deduplicates a person.
End state
The generate_lead event stays on once per event, the recommended method. The worked example below is one session with five form submissions, counted both ways, and a week of 2,400 sessions in which the choice moves the total.
| Figure | Value |
|---|---|
| Submissions in one session | 5 |
| Key events, once per event | 5 |
| Key events, once per session | 1 |
| Difference in one session | 4 |
| Sessions in the week | 2,400 |
| Key events in the week, once per event | 312 |
| Key events in the week, once per session | 268 |
| Difference in the week | 44 |
The week difference is 312 − 268 = 44 key events, recomputed in Python 3. The one-session difference is 5 − 1 = 4. The figures are a worked example, not a real property.
Sources
- Analytics Help — Change the counting method of key events — the once per event and once per session counting methods, and that a change applies to future data only
- Analytics Help — Conversions vs. key events in Google Analytics — what a key event is, the event to key event to conversion flow, and where Google Ads conversions appear
Checked 11 November 2026. The click path above is the one in Google's help pages, read 11 November 2026. The two counting methods, the Recommended and Legacy labels, the Marketer requirement, and the future-only effect of a change are as Google documents them. No step in this guide was run inside a live Analytics property, because Admin needs a Google sign-in and this site does not sign in as anyone. The figures on this page are a worked example recomputed in Python 3 with decimal arithmetic, not numbers from a real property. Google moves menu items without warning; if a path on this page no longer matches, the linked help page is the current map.