GA4 Key Events for B2B Websites: What to Mark and How

The short answer

Mark two to five key events, each tied to a sales conversation: a successful lead form submission (ideally the recommended generate_lead event), a booked meeting, and perhaps a quote request. Fire them on success, not on click. Test them in DebugView before you rely on them, and remember that marking an event as key only affects data from that point on.

Events, key events and conversions

Google Analytics describes a key event as "an event that measures an action that's particularly important to the success of your business"¹. Events that used to be marked as conversions in Analytics are now called key events, and the word conversion is kept for key events you share with Google Ads². The three terms now mean different things:

Term	What it is	Where you see it
Event	Any interaction Analytics records: a page view, a scroll, a click, a form submission	Reports and explorations, Admin > Data display > Events
Key event	An event you have marked as important to the business	The Key events and Session key event rate metrics in Analytics reports
Conversion	A key event shared with a linked Google Ads account, used for Ads reporting and bidding	Google Ads, and the advertising sections of Analytics

Session key event rate is the share of sessions in which a key event happened³. Creating a conversion from a key event needs a linked Google Ads account²; if you do not advertise on Google, you only need key events.

Which events a B2B website should mark

A key event should mean that someone has asked to talk to you, or done something your sales team would act on. If marking an event would make your key event count rise without sales noticing any difference, it should stay an ordinary event.

Action	Suggested event	Mark as key?
Demo, contact or quote form submitted successfully	`generate_lead`, with a parameter saying which form	Yes: this is the core B2B key event
Meeting booked through a scheduling tool	A custom event such as `meeting_booked`	Yes, if the booking is confirmed on your site
Free trial or account sign-up, where you offer one	`sign_up`	Yes, for product-led businesses
Gated content download (form completed)	`generate_lead` with a different form parameter, or a custom event	Often, but report it separately from sales requests
Newsletter sign-up	A custom event such as `newsletter_signup`	Usually not; never lump it with demo requests
Phone number or email address clicked	A custom click event	Only if you have checked it predicts real enquiries
Form started	`form_start` (enhanced measurement)	No: a useful diagnostic, not an outcome
Pricing page viewed, 90% scroll, file download, outbound click	`page_view`, `scroll`, `file_download`, `click`	No: signals of interest, not outcomes

generate_lead is one of Google's recommended events for lead generation, described as when a user "submits a form online or submits information offline". The same family includes qualify_lead, working_lead, close_convert_lead, close_unconvert_lead and disqualify_lead, which together measure the lead funnel and populate the Lead acquisition report⁴. Using recommended names where they fit, rather than inventing your own, keeps your setup consistent with Google's reports and documentation.

Standard properties can mark up to 30 events as key events, and Analytics 360 properties up to 50⁵. Few B2B sites need more than five. A short list keeps the Key events metric meaningful: if it adds up newsletter sign-ups and demo requests, nobody can read it.

How to set up a lead key event, step by step

Step 1: decide what "success" looks like for each form

Write down every form that leads to a sales conversation and what counts as a successful submission. The event should fire only when the submission has been accepted: on the thank-you page, or when your form tool reports success. A click on the submit button also fires when validation fails.

Step 2: send the event

Enhanced measurement can record form interactions automatically, as form_start the first time a user interacts with a form in a session and form_submit when the user submits a form⁶. It detects forms for you, so it cannot know which of your forms matter or whether a submission succeeded. Test it on your own forms before relying on it, and for key events prefer an event you send deliberately. With the Google tag, that looks like this:

gtag('event', 'generate_lead', { lead_source: 'Demo request form' });

In Google Tag Manager, the equivalent is a Google Analytics event tag named generate_lead, triggered by your form's success event or thank-you page view. Google's reference lists value, currency and lead_source as the event's parameters, and says that if you set value, currency is required⁷. Only send a value if you have an agreed basis for it, and use GBP consistently. To use a parameter such as lead_source in standard reports, register it as a custom dimension⁶.

Step 3: mark it as a key event

In Admin, under Data display, open Events and click the star next to the event. If the event has not been collected yet, use Create event, give it the exact name and turn on Mark as key event. Marking an event as a key event "affects reports from time of creation" and does not change historic data; standard reports can take up to 24 hours to show the change⁵. Mark your key events before a change you want to measure, not after.

Step 4: choose the counting method

Analytics offers two counting methods. "Once per event" counts a key event every time it is triggered and is the recommended option; "once per session" is the legacy method and counts it once within a session⁸. Once per event is the right default, provided your event fires only on genuine success. If a thank-you page reload or a double submission inflates the count, fix the trigger rather than hiding the problem with the counting method. Changes to the counting method apply to future data only⁸.

Step 5: test it in DebugView

DebugView, under Admin > Data display, shows events from devices in debug mode as they arrive; Tag Assistant or Tag Manager's preview mode turns debug mode on for your own browser⁹. Submit each form once, with valid and invalid entries, and check that:

the event fires once on a successful submission and not at all on a failed one;
the parameters, such as lead_source, carry the right values;
reloading the thank-you page does not fire it again;
it fires on mobile as well as desktop.

Step 6: reconcile with your CRM

After a few weeks, compare the key event count for each form with the leads your CRM recorded from it. They will not match exactly: some visitors decline analytics cookies or block tracking, and some leads arrive by email or phone. But a large or sudden gap is a sign that the tracking or the form has broken. Agree which number is the source of truth for each purpose: Analytics for on-site behaviour and trends, the CRM for lead counts and pipeline.

Common mistakes

Marking engagement as a key event. Scrolls, time on page or pricing page views inflate the Key events metric and make the session key event rate meaningless.
One key event for every form. If demo requests and newsletter sign-ups both fire generate_lead with no parameter to tell them apart, you cannot see which one changed.
Firing on click. The submit button click fires even when the form rejects the entry.
Double tagging. A hard-coded Google tag plus a Tag Manager tag for the same event counts every lead twice.
Counting your own team. Test submissions from staff and agencies pollute the data. Define internal traffic and filter it, and delete test leads in the CRM.
Changing definitions silently. If you rename an event, change its trigger or add a form, note the date. A key event that changes meaning mid-year breaks every comparison. Our guide to keeping a website change log covers this.

Using key events to measure website changes

Once your key events are clean, they become the primary metric for judging website changes: demo requests from a services page, say, before and after a rewrite. Most B2B pages produce too few key events for a quick verdict, which is why it matters to fix the measurement window in advance and expect some results to be inconclusive. Our step-by-step guide to measuring the impact of a website change shows how, and the guide to measuring website changes honestly explains the four verdicts we use. If a key event count and a Search Console figure seem to disagree, our guide to why Search Console and GA4 numbers differ explains which to use for which question.

Where Artiweave fits

Artiweave judges each approved change against a metric chosen in advance, so the quality of your key events decides how much its measurement can tell you. Clean key events make a verdict possible; noisy ones push more results into "inconclusive".

Frequently asked questions

What is the difference between a key event and a conversion in GA4?

A key event is an event you mark in Google Analytics as important to your business. A conversion is a key event that you share with a linked Google Ads account for Ads reporting and bidding. If you do not advertise on Google, you only need key events.

How many key events can I have in GA4?

Standard properties can mark up to 30 events as key events, and Analytics 360 properties up to 50. Most B2B websites need no more than five.

Does marking an event as a key event apply to past data?

No. Marking an event as a key event affects reports from the time you mark it and does not change historic data, so mark key events before a change you want to measure.

Should I count key events once per event or once per session?

Google recommends once per event, which counts every time the event fires. Once per session is the legacy method. Make sure the event fires only on genuine success, so the count reflects real leads.

Is form_submit from enhanced measurement good enough for lead tracking?

It is a useful start, but it detects forms automatically and cannot tell which forms matter or whether a submission succeeded. For key events, send a deliberate event such as generate_lead when the form reports success, and test it in DebugView.

Sources

Measure website changes against the leads that matter.

Artiweave publishes only the changes your team approves, with automatic rollback, and reports each outcome honestly against the metric you chose. It is in private beta.

Join the waitlist