GA4 Automatically Collected Events: Free Data for CRO

Short answer
GA4 automatically collected events: what are they and why do they matter for CRO?
GA4 automatically collected events are basic interactions that Google Analytics 4 records without extra code once your Google tag runs on the site. Page views, scrolls, outbound clicks, file downloads, and form interactions are examples. For conversion rate optimization (CRO), they give you free behavior data to build your first hypotheses.
Most businesses only look at the final conversion, such as a sale or a form submission. However, what users do on the way to that point is the raw material of optimization. This guide shows which questions the built-in events can answer.
The details also follow Google Analytics Help. Event names and settings can change, so confirm them on the automatically collected events page before you act.
How do GA4 automatically collected events differ from enhanced measurement?
The two terms overlap, so people mix them up. According to Google's enhanced measurement help page, enhanced measurement events switch on per web data stream and need no code changes. Many of them also appear in the automatically collected events list.
In practice, think of it this way. The automatic list is broad and covers both web and app. Enhanced measurement is the group of web settings that you can switch on or off one by one.
This guide only covers the web side. App events, such as first open or notification events, sit outside the scope of this CRO guide.
To change these settings you need the Editor role or higher on the property. If you use the Google tag, each event type also needs automatic detection turned on. By default, all of them are on.
Which events collect data without extra setup?
The table below lists the web events and what each one can tell you in CRO. Also, the triggers come from Google Analytics Help. The last column is our field-based starting view, so treat it as a hint, not a fixed rule.
| Event | When it fires | What it tells you in CRO |
|---|---|---|
| page_view | A page loads or the browser history state changes | Shows which pages act as entry doors |
| scroll | A user first reaches 90% vertical depth | Shows who reaches the end of the content |
| click | A user clicks a link to another domain | Shows where visitors leave your site |
| file_download | A user clicks a link to a file with a common extension | Shows pre-decision document needs |
| form_start | A user first interacts with a form in a session | Measures form intent apart from submission |
| form_submit | A user submits a form | Confirms form conversion |
| video_start, video_progress, video_complete | An embedded YouTube video starts, hits 10, 25, 50, 75 percent, or ends | Shows how persuasive the video is |
| view_search_results | Site search shows up through a URL query parameter | Shows what visitors look for and cannot find |
| user_engagement | A page stays in focus for at least one second | Separates real interest from page views |
For a wider introduction, read what GA4 is and its benefits. To confirm that your events collect correctly, use our checklist on whether your GA4 setup is correct.
To turn event counts into rates, the free conversion rate calculator helps. That way you move from "how many people" to "what percentage".
What does the scroll event tell you about CRO?
The GA4 scroll event fires the first time a user reaches 90% vertical depth. Google's help page says it fires only once per page. So it does not show intermediate points such as 25% or 50%.
So that limit shapes your reading. The event counts people who reach the end. For example, a low rate on a long service page may mean that users decided earlier, not that they ignored the content.
If your pricing or contact block sits in the middle of the page, a visitor can convert without scrolling to the end. So read scroll together with form start and click events, never alone.
- Compare by page type, because a blog post and a service page set different scroll expectations.
- Read mobile and desktop separately.
- If scrolling feels broken, read our guide to designing the scroll experience.
The first screen and the section order decide most of this result. If you want to rethink that structure, professional web design services turn the data straight into page architecture. Otherwise you can start with small copy changes.
Does the click event only measure outbound links?
Yes. The automatic click event only measures clicks on links that lead away from your current domain. Links to domains set up for cross-domain measurement do not trigger it. Its parameters include link classes, domain, ID, and URL.
So you cannot see internal menu or button clicks with this event. Those need a custom event, which we cover later in this guide.
Still, outbound click data has value. For example, if visitors go to a third-party booking page, you may lose conversions at that exact point. You can also watch map, social, or payment provider links one by one.
- First, report the link_domain parameter as a custom dimension.
- Then list the external domains that draw the most exits.
- Finally, judge the reason for each exit next to the page content.
What decisions can the file_download event support?
The file_download event fires when a user clicks a link to a file with a common extension. Google's page lists documents, archives, executables, media, and presentations. Its parameters include file name, extension, and link text.
For CRO, this event shows pre-decision information needs. People who download a catalog, price list, or technical brochure may sit close to a purchase. That is an assumption, so check it against your own data.
On the other hand, a download does not replace a contact. If users take the document and leave, there is a gap between download and inquiry. You can close it by adding a clear next step on the download page.
The file name parameter also tells you which content draws interest. For example, if one brochure clearly beats another, visitors care more about that topic. Moving that document to a more visible spot can be your first test.
How do form_start and form_submit explain form conversion?
The form_start event fires on the first form interaction in a session. Then the form_submit event fires when a user submits the form. As a result, the gap between the two shows your abandonment rate, which is the cheapest way to spot friction in a form.
However, be careful. The form_start event counts the first interaction per session, so it may not separate several forms in one session. If a page has both a search box and a contact form, check the form_id and form_name parameters.
Once you see the gap between start and submit, you can ask these questions:
- Does the form have more fields than it needs?
- Are the error messages clear?
- Do keyboard and field types fit mobile use?
For layout ideas, read our guide to lead form design for booking, quote, and demo pages. Also remember that parameters only show in reports after you create custom dimensions for them.
How do video events show the persuasion power of a page?
For embedded YouTube videos, GA4 collects three events: video_start, video_progress, and video_complete. Progress fires at 10, 25, 50, and 75 percent. Google's page says this works only for YouTube embeds with JS API support.
The data shows whether a video gets watched at all. For example, if a product video starts often but viewers leave early, the opening seconds may be weak. When the video is the main persuasion tool on the page, that gap matters.
You can also compare the conversion rate of video viewers with that of non-viewers. Do not read the result as cause and effect, because an interested visitor tends to watch and convert anyway. Still, the gap helps you judge where the video belongs.
Videos embedded from other platforms may not fire these events. In that case you need custom tracking. Before you move a video to the first screen, save the current viewing data so you can compare afterwards.
How do you read on-site search intent with view_search_results?
The view_search_results event detects site search through a URL query parameter. According to Google's page, the default keys are q, s, search, query, and keyword. You can add more keys in the advanced settings. The search term arrives in the search_term parameter.
Site search is the one place where visitors tell you in their own words what they want. That makes it very valuable for CRO. Frequent searches with no useful result point to missing content or poor labeling.
As an example scenario, imagine visitors on a service site keep searching for a price. If the site has no pricing page, that term shows a page need. Not publishing prices is a business decision. Still, visitors need a page that explains the next step.
However, if your search box does not use a URL parameter, the event will not fire. In that case, check the search mechanism first. Pages with zero results are also a useful signal, because they show the exact moment a visitor fails to find something.
What foundation do page_view, session_start, and user_engagement build?
These three events form the ground floor of every other analysis. The page_view event fires when a page loads or the browser history state changes. Google's page says it collects automatically and you cannot turn it off. The session_start event begins the session when a user engages with the site.
The user_engagement event fires when a page stays in focus for at least one second and carries engagement time. That helps you separate a tab that opened and closed from real interest. The first_visit event also marks a first visit.
Because of that, the CRO value is simple but important. First you see which pages act as entry doors. Then you see where interest drops. To find the leak in your funnel, try our guide to GA4 funnel exploration.
Instead of hunting for one problem on one page, start with your three highest traffic pages. That way you spend limited energy where the impact is largest, because small pages rarely move the totals. A change on a low traffic page rarely moves overall conversion in a visible way.
Why do GA4 automatically collected events go missing in reports?
First, the most common reason is that parameters do not appear in reports by themselves. Google's page says parameters show up only after you create custom dimensions for them. You see the event count, but the file name or form ID never reaches the table.
The second reason is settings. You may have switched off enhanced measurement events at the web data stream level. Also, with the Google tag, automatic detection must stay on. And you need the Editor role to change these settings.
The third reason is site structure. In single page apps or forms inside an iframe, events may not fire the way you expect. For missing or double counted conversions, read our guide to why GA4 conversions go missing.
- First, test that the event really fires in the real-time view.
- Then define the parameter as a custom dimension.
- Finally, wait a few days and check the report again.
How do you make event parameters reportable?
A custom dimension makes a parameter usable in reports. First decide which parameter the decision needs. Adding everything only creates clutter.
In practice, a short starting list is enough for CRO. Use file name and extension for file_download, link domain for click, form ID and name for form events, and the search term for site search. These five cover most hypotheses.
- Enter the parameter name exactly as the official documentation spells it.
- Then choose the event scope for the custom dimension.
- After you define it, wait for data to build up, because it does not backfill history.
Menu labels change, so we do not give an exact click path here. Look for the custom definitions area in the admin panel and follow Google's current help page.
Why should you read mobile and desktop event data separately?
For example, the same page gives two different experiences on two devices. A button above the fold on desktop can sit below it on mobile. So total numbers can hide the real problem. Split by device category first, then interpret.
For example, if scroll is high on mobile but form_start is low, visitors may read the content and still fail to find the form. On desktop you may see the reverse. Those are two hypotheses and two different fixes.
Clicks also behave differently on touch screens. Accidental taps, brief tab opens, and quick returns can rise. So read the click event with care on mobile.
- Split each event by device category.
- Then start with the device that brings the most traffic.
- Fix that device experience first when you find a gap.
Which page types gain the most from GA4 automatically collected events?
Not every page gets the same value from the same events. Service and contact pages benefit from form events. Blog posts benefit from scroll and outbound clicks. Download pages benefit from file_download. Knowing this match tells you which event to check where.
On a service page, the main question is whether visitors see, start, and submit the form. In contrast, on a blog post, reading depth and the move to the next page matter more. Meanwhile, on a catalog or product page, downloads and video events stand out.
- Service page: form_start, form_submit, scroll.
- Blog post: scroll, outbound click, user_engagement.
- Catalog or price list page: file_download, click.
- Presentation page: video events, form_start.
The home page is a special case. There, events show which direction visitors take. Look at it through the lens of a conversion funnel.
How do you turn automatic events into CRO hypotheses?
However, data alone is not a decision. A hypothesis turns an observation into a testable sentence: on page X we see problem Y, and change Z should lift conversion. The table below maps event patterns to hypotheses. The examples are illustrative and promise no result.
| Observation | Possible reading | Hypothesis to test |
|---|---|---|
| Many form_start, few form_submit | Form friction exists | Fewer fields will lift submissions |
| Low scroll, high exit | The first screen promises too little | A new headline and first screen will lift engagement |
| Many file_download, few contacts | Visitors take the file and leave | A clear call after the download will lift inquiries |
| Frequent view_search_results term | Menu or content is missing | A page for that term will cut searches |
| High outbound click | Visitors leave to decide | Moving the information on page will cut exits |
| Video starts but viewers drop early | The first seconds are weak | A shorter intro will lift watch rate |
After you write the hypothesis, rank it. Pick the one with high impact and easy delivery. For testing, our guide on getting started with A/B testing shows the way.
How do you tie event data to business results with a conversion rate calculator?
Event counts mean little alone. They gain meaning as rates. Open the tool to calculate your conversion rate, enter a page's visitor count and the number of target events, and note the result. That rate becomes the baseline for your hypothesis.
For example, enter a service page's sessions and its form_submit count. Then repeat the math for form_start. The gap between the two rates shows how many users the form loses. Running the math separately for mobile and desktop gives an even clearer picture.
Then, after a change, repeat the same math. For the comparison to mean anything, the periods must look alike. Small samples swing naturally, so also check with the ab test calculator whether a difference is meaningful.
That way event data becomes a concrete number you can defend in the next meeting. If you save the tool, you can repeat the math quickly after every test.
Which events should become key events?
Not every event is a conversion. A key event represents a business result, such as a form submission, a booking, or a purchase. File downloads or scrolls are usually micro conversions, not the main goal.
This split keeps reports clean. If you mark everything as a key event, your conversion count inflates and optimization points the wrong way. For goal selection, read our guide to setting website conversion goals.
Still, tracking micro conversions has value. They show the in-between steps of the decision journey. For e-commerce examples, our post on micro conversions and GA4 tracking helps.
- Main goal: one or two events that produce a business result.
- Support events: form start, document download, video watch.
- Information events: scroll, outbound click, search.
How do you verify event data with tests?
An event that appears in a report did not necessarily fire correctly. So test each event by hand. Browse the site, trigger the event on purpose, and watch it in the real-time view. This simple step saves months of wrong interpretation.
- First, open the page and confirm the page_view event arrives.
- Then scroll to the bottom and wait for the scroll event.
- Next, click an external link and check the click parameters.
- After that, click a file link and verify file_download.
- Finally, start the form, submit it, and compare the two events.
Run the test on mobile and on desktop. Also try both consent choices, accepting and rejecting cookies, because measurement behavior can change with your consent setup.
If you find a problem, check the tag setup first, then the data stream settings, and the site structure last. That order solves most issues fastest.
What privacy points should you watch with these events?
Google's help page asks you to make sure you collect no personally identifiable information. Form events are sensitive in this respect. Never send form field values, such as a name, email, or phone number, as parameters.
Also, if page URLs carry personal data in query parameters, that data can leak into event parameters. Search terms can sometimes contain personal details too. So review your data stream on a regular basis.
Cookie consent and measurement behavior vary by country and by your settings. This article is not legal advice. Talk to a specialist about the rules that apply to you, including consent management and retention.
On the technical side, test your tag management and consent settings together. A badly wired consent flow causes problems for both compliance and data quality.
When do automatic events stop being enough?
Built-in events measure general behavior, not behavior unique to your site. Internal menu clicks, the use of one specific button, tab changes, or a product comparison all need custom events. In those cases a tag management tool steps in.
First write down which decision the event will serve, then build it. An event with no decision behind it only clutters the report. For setup, our guide to Google Tag Manager covers the basics.
Before you add custom events, make sure you use the built-in ones fully. Most businesses never read even half of the free data. Reading first and adding later is the healthier order.
Standardize event naming from day one. Consistent names make future reports easier to build and make team communication smoother.
How do you fit event data into a weekly CRO routine?
Data is useless if nobody looks at it on a schedule. A thirty minute weekly routine is enough for most businesses. Ask the same three questions every week: what changed on the top page, which event drifted from expectation, and which hypothesis are we testing?
This routine makes you plan instead of react. Do not take weekly swings too seriously. Look at the trend over three or four weeks and allow for seasonal effects.
- First, check the event rates of your top three pages.
- Then flag drifts against last week and last month.
- Next, pick one hypothesis and write the change plan.
- Finally, record the result of the previous test.
Keeping a log is very valuable. If you do not know which change happened when, you cannot tie a rise or drop to the right cause. A simple table is enough.
What does a GA4 automatically collected events checklist look like?
The list below is a quick audit before you use the data for CRO. Go through it in order. If you note the date and result next to each item, you can see later what changed.
- Does the Google tag load on every page?
- Are enhanced measurement events on in the web data stream?
- Do scroll, click, file_download, and form events fire in a test?
- Did you define the needed parameters as custom dimensions?
- Does the site search URL parameter get detected?
- Do your key events match business results?
- Is personal data kept out of parameters?
- Do you have a written hypothesis list with priorities?
Run this list again every three months. Events change as your site changes. A theme update can quietly break a form trigger.
How do you turn data into page design, and when do you need help?
Event data shows a problem, not the fix. The fix lives in page structure, message, and flow. You can test small changes yourself: a new headline, a button moved up, one field removed.
But when the problem sits in page architecture, the change needs design skill. At that point a design partner helps you turn data into visual and structural decisions. As Talha Aslan and our team, we handle measurement and page structure together. We suggest testable steps and make no promise of results.
For the broader CRO approach, our sibling guide on when conversion rate optimization pays off is a good companion. To see the whole process, also read our guide to the CRO process for a business website.
If you want to go alone, use the checklist in this guide. If you need a deeper redesign, take a look at our business website design page.
Core guide: What Is Digital Marketing? Its types, examples and how it is done, on one page with an interactive film that travels through the brain.



