GA4 renamed Conversions to Key Events — what breaks and what doesn't

GA4 renamed "conversions" to "key events". That is genuinely all that changed inside GA4 — same events, same data, same historical numbers, a different label in the interface. What the rename did break is everything around GA4: Google Ads still says conversions and now means something slightly different, API field names moved, and every dashboard, SQL query, and internal document that hardcoded the old vocabulary now needs reading twice.

What actually changed

Inside GA4, an event you flag as important is now a key event. Previously it was a conversion. The toggle is in the same place, does the same thing, and your historical data is untouched — no reprocessing, no gaps, no need to re-mark anything.

The reason for the change is a real distinction rather than cosmetics. Google wanted two separate vocabularies:

• Key events are analytics outcomes — meaningful actions measured in GA4, whether or not they involve advertising.
• Conversions are advertising outcomes — actions a Google Ads campaign is optimising toward and bidding on.

They frequently describe the same underlying action, and that is exactly why they used to be confused. A newsletter signup can be a key event you care about analytically while not being a conversion you spend money to acquire.

What did not change

Worth stating plainly, because the rename caused more panic than it deserved:

• Historical data. Nothing was reprocessed or lost.
• Your event configuration. Events already marked stay marked.
• BigQuery export schema. The raw event tables are unaffected — another reason streaming GA4 into BigQuery is a good hedge against interface churn.
• Google Ads bidding. Campaigns optimising toward imported actions kept working throughout.

What actually breaks

Reports, dashboards, and saved views

Anything that referenced the metric by its display name. Looker Studio reports, scheduled exports, and internal dashboards that pulled "Conversions" from a GA4 source need their fields re-pointed. The data is intact; the label the report was looking for is not.

API integrations

The Data API metric name changed. Code requesting the old metric name gets an error or an empty column, depending on how gracefully the client handles unknown fields. This is the failure most likely to be silent — a nightly job that quietly writes zeros.

Cross-platform reconciliation

This is the durable one. Google Ads still calls them conversions and counts them its own way. A stakeholder comparing "conversions" in Ads to "key events" in GA4 is now comparing two differently named things that were already differently calculated. Why the two never match has not changed — but the rename removed the false comfort of a shared word.

Documentation and onboarding

Every runbook, tracking plan, and training deck written before the change. Cheap to fix, easy to forget, and the source of months of low-grade confusion for new team members.

The migration checklist

1. Inventory every consumer of the metric. Looker Studio, spreadsheets, warehouse jobs, agency reports, internal tooling.
2. Re-point dashboard fields to the key events metric and re-run a known historical period to confirm the numbers match what the report showed before.
3. Update API clients to the current metric name, and add an assertion that the returned column is non-empty so a future rename fails loudly instead of writing zeros.
4. Rewrite the vocabulary in your documentation — including the distinction, not just the word, so people understand why Ads still says conversions.
5. Re-check which key events are imported into Ads. The rename is a good excuse to confirm you are not double-counting by importing GA4 key events while also running a native Ads conversion tag for the same action.
6. Review what is marked at all. Most properties accumulate key events nobody optimises toward. Fewer, better-chosen ones make reports and bidding clearer.

The distinction worth internalising

The rename is a small event with one lasting benefit: it forces teams to answer a question they were previously able to dodge. For each important action, ask two separate questions.

Is this an analytics outcome? Something you want to count, segment, and put in a funnel. Mark it as a key event.

Is this an advertising outcome? Something worth spending money to cause, that a bidding algorithm should optimise toward. Import it into Ads as a conversion action.

Many actions are the first without being the second. Marking everything as both is how bidding algorithms end up optimising toward newsletter signups instead of revenue.

	Key event (GA4)	Conversion (Google Ads)
Question it answers	Did something meaningful happen?	Did my ad spend cause an outcome?
Attribution	Cross-channel, GA4's model	Ad-click based, Ads' model
Counting	Every occurrence	Every, or one per click
Used for	Reporting, funnels, audiences	Bidding, budget decisions

FAQ

Do I need to re-mark my conversions as key events in GA4?

No. Everything previously marked as a conversion carried over automatically, and historical data was unaffected.

Why does Google Ads still say conversions?

Because in Ads the word now means specifically an advertising outcome used for bidding. The two products deliberately use different terms for different concepts that often, but not always, describe the same user action.

Did the rename change my BigQuery export?

No. The raw event export schema is unchanged, which is one of the reasons warehouse-based reporting is more resilient to interface changes than dashboards built directly on the GA4 UI.

My Looker Studio report went blank after the rename. What do I do?

Re-point the metric field to key events and re-run a closed historical period to confirm the numbers match the previous version of the report. Blank columns are almost always a field-name mismatch, not lost data.

Should every important event be a key event?

No. Key events should be the small set of outcomes you actually make decisions on — typically three to eight. Marking dozens dilutes reporting and, if imported into Ads, actively degrades bidding.

While you are auditing what GA4 is measuring, confirm the tag is loading correctly in the first place: the free tracking audit checks GA4 presence, duplicate properties, and consent signals on any URL.