Website Traffic Analytics in GA4: Seven Reports an Owner Needs, and How to Read Them Without an Analyst
Website traffic analytics in GA4 for owners: 7 reports, key events for HoReCa, services, and e-commerce, UTM discipline, and a 15-minute weekly review.
In Short
Website traffic analytics exists so an owner can answer three questions: where do people come from, what do they do on the site, and how many of them turn into money. GA4 gives you these answers for free, on two conditions: key events are set up on the site, and every ad and owned link is tagged with UTM parameters. Below: seven reports an owner should check, the events to track for HoReCa, services, and e-commerce, how to pull GA4, Search Console, and ad accounts into one picture, and a 15-minute weekly routine that replaces an hour of wandering the interface.
Why an Owner Sees Numbers but Not Answers
A typical scene: an owner opens GA4, sees “users: 4,812, sessions: 6,300, average engagement time: 48 seconds,” and closes the tab. There are numbers, no decision. Three reasons why.
No definition of a result. GA4 doesn’t know what counts as success for your business. For a restaurant it’s a table booking and a phone-number click, for a clinic it’s a booking form, for a store it’s a paid order. Until you’ve marked these actions as key events, the system counts visits, not results.
Traffic sources aren’t tagged. Half of your visits are marked direct or unassigned, because a link from a newsletter, a QR code on a menu, and an ad in a Telegram channel all look identical to GA4. You can’t say which channel brought customers and which brought only views.
No habit of reading the same thing. The GA4 interface has dozens of reports and an exploration builder. If you open different sections every week, you’ll have nothing consistent to compare against. The working approach is a fixed set of reports in a fixed order.
Once these three things are handled, website traffic analytics turns from a pile of numbers into a decision-making tool: where to add budget, which page to rework, which channel to shut off.
The Seven GA4 Reports an Owner Actually Needs
You can spend an hour in GA4 and never find what matters. Here are the seven reports that cover 90% of an owner’s decisions. Names are given exactly as they appear in the menu, with a note on what to look at there.
1. Traffic acquisition by channel (Reports → Acquisition → Traffic acquisition). Shows where sessions came from: organic search, paid search, social, referral, direct, email. Don’t look at session count. Look at the key events column: a channel with less traffic and more leads is worth more than a channel with lots of traffic and zero results. We covered how to split effort across channels in detail in the article on website traffic sources.
2. Landing pages (Reports → Engagement → Landing page). This is the first page a visitor sees. Compare the engagement rate and key-event count by page: if a page gets a lot of traffic and few leads, the problem is the page, not the ad.
3. Key events (Reports → Engagement → Events, filtered to key events). How many calls, forms, bookings, and purchases happened over a period, and from which sources. This is the main report, and it’s the one that’s empty in most of the accounts we see during audits.
4. Engagement (Reports → Engagement → Overview). Engagement rate, average time, views per session. Don’t judge these numbers on their own. Compare across channels and pages. A 5-second session from an ad means the ad promises something the page doesn’t deliver.
5. Geography and devices (Reports → User → Demographics and Tech). For a local business, this is a sanity filter: if a restaurant in Poltava gets a third of its traffic from other regions, the ad targeting is too broad. The device breakdown shows whether the site is breaking on mobile: if conversion on phones is several times lower than on desktop, while phones make up most of the traffic, you’re losing leads to a form or a speed problem.
6. User paths (Explore → Path exploration). Where people go after the landing page, and at which step they drop off. For e-commerce it’s the funnel “product card → cart → checkout → payment”; for services it’s “service page → pricing → contacts.” A gap in the funnel gives a developer or designer a concrete task.
7. Retention (Reports → Retention). How many people return to the site after their first visit. For food delivery and stores, this is a direct signal of product and service quality; for services, it’s a reason to look at remarketing, which brings back people who left without converting.
These seven reports, taken in this order, form a route: first where people came from, then where they landed, then what they did, then why they didn’t.
UTM Discipline: Without It, Half Your Traffic Is “Direct”
UTM parameters are five values in a link that tell GA4 where a visit came from: source, medium, campaign, plus content and term. Without them, the system guesses, and guesses badly.
What usually stays untagged and disappears into direct:
- links in newsletters and Viber or Telegram channels;
- QR codes on menus, receipts, storefronts, and printed materials;
- links in Instagram and TikTok bios;
- influencer ads and partner posts;
- offline advertising with a shortened link.
Rules worth writing down for the whole team:
- One shared vocabulary. The source is always lowercase and consistent: instagram, not Instagram or insta. GA4 treats those as three different sources.
- Medium matches purpose. For paid ads, medium is cpc or paid_social; for newsletters, email; for partners, referral; for QR codes, offline or qr.
- Campaign names with a date or period. A year from now you won’t remember what promo2 was, but lunch_menu_2026_10 reads without explanation.
- One generator, one spreadsheet. All tags are created from one template and logged in a shared spreadsheet. Whoever makes the link logs it.
- No UTM tags on internal links. A tag on a button inside the site cuts the session short and attributes the visit to a new source.
For Google Ads, tags are added automatically with auto-tagging enabled, so in a Google Ads setup from MOVE, UTM tagging and the GA4 connection are one line item: without it, optimizing for leads isn’t possible.
Key Events: What Counts as a Result in Different Niches
A key event in GA4 is an action you’ve marked as valuable. It’s what becomes a “conversion” in reports and a signal for ad platforms. The set of events depends on the business. Here’s the baseline minimum for the three types of clients we work with most.
| Business type | Required key events | Additional events for deeper analysis |
|---|---|---|
| HoReCa: restaurants, cafés, delivery | Phone click, table booking, completed delivery order, messenger button click | Menu view, get directions, address click, click to Instagram |
| Services: clinics, auto shops, security, education | Booking or inquiry form submitted, phone click, callback request, widget booking | Pricing page view, doctor or specialist page view, price list download, form start |
| E-commerce: online stores | Add to cart, checkout started, purchase with amount passed | Product view, add to wishlist, filter used, promo code applied, cart return |
Three principles for setup:
- An event has to represent money. “Viewed contact page” isn’t a result. A call or a form is.
- No more than five key events. If everything is key, ad platforms optimize for the cheapest action instead of the lead.
- A form counts as submitted, not opened. The event has to fire after a successful submission, otherwise you’re counting attempts, not leads.
Technically, events are set up through Google Tag Manager so you don’t need a developer for every change. In the Parodent dental network case, the 23% drop in cost per lead and the 30% growth in bookings became visible precisely because leads were counted as events, not eyeballed by the front desk: without that, comparing periods and proving the result would have been impossible.
GA4 + Search Console + Ad Accounts: One Picture Instead of Three
GA4 on its own answers “what did people do on the site,” but not the other two questions: “which queries find us” and “what did this cost.” That’s why an owner needs three sources connected together.
Search Console shows the queries the site shows up for in Google, clicks, impressions, average position, and technical indexing errors. It’s the only place you see organic demand before the click. Connecting it to GA4 lets you see queries directly inside acquisition reports.
Google Ads and Meta Ads know spend, impressions, and clicks. GA4 knows what those clicks did on the site. Connecting Google Ads to GA4 lets you see cost per key event by campaign, not just cost per click. For Meta, the pixel and Conversions API play the same role, and channel comparison happens in GA4 via UTM tags.
What the combined picture gives you:
| Question | Where the answer is |
|---|---|
| How much did we spend on a channel? | Ad account |
| How many leads did the channel bring? | GA4, key events report by source |
| What does a lead from each channel cost? | Ad account spend divided by GA4 events |
| Which queries find us for free? | Search Console |
| Which queries are worth covering with ads, and which are already covered by organic? | Search Console + Google Ads campaign report |
| Are there technical problems eating into traffic? | Search Console, Indexing and Core Web Vitals |
The line between paid and free traffic becomes concrete in this picture: you see which queries you’re paying for even though you already rank at the top, and the reverse. For the balance between the two, read paid vs. organic traffic.
A 15-Minute Weekly Review
Website traffic analytics works when it’s regular. Here’s a routine an owner or manager runs every Monday, following the same script. Compare the week against the previous week and against the same week a month ago, to tell a trend from a seasonal blip.
| Minutes | What to open | What to look at | What decision might come out of it |
|---|---|---|---|
| 0–3 | GA4 → Traffic acquisition | Sessions and key events by channel, change vs. last week | A channel dipped or grew: a question for the vendor, budget reallocation |
| 3–6 | GA4 → Key events | Number of leads, calls, purchases, and their split by source | Fewer leads on the same traffic: check the form, phone tracking, the site |
| 6–9 | Ad accounts | Weekly spend, cost per lead by campaign | A campaign costs more than the target: pause or rework it |
| 9–11 | GA4 → Landing pages | Top 10 pages by traffic, their engagement rate and events | A page with traffic but no leads goes on the rework list |
| 11–13 | Search Console → Performance | Clicks and impressions for the week, new queries in the top 20 | A query is close to page one: reinforce the page with content |
| 13–15 | Search Console → Indexing | New errors, pages dropped from the index | A technical task for the developer |
These 15 minutes are for a pulse check; deep analysis happens separately. Once a month, spend an hour with whoever runs the ads looking at user paths, retention, and a full campaign breakdown. If traffic overall is falling and you need a growth plan, start with the article on how to increase website traffic.
Data Traps: When the Numbers Lie
Even a properly configured analytics setup can mislead. Here are the traps we find in most audits.
Bot and junk traffic. A spike in sessions with an engagement rate near zero, odd geography, and one-second duration has nothing to do with customers. GA4 automatically filters some known bots, but referral spam and scanners are worth excluding manually through filters and referral exclusion lists. Otherwise the site’s average metrics look worse than they are, and the direct share looks bigger.
No consent for data collection. If the site has a cookie banner that blocks analytics until consent, part of your visits never reach GA4 at all. That’s normal and lawful, but it means GA4’s numbers are lower than reality, and ad accounts and GA4 will never match to the last cent. A gap of tens of percent between ad-account clicks and GA4 sessions means you should check how consent is actually implemented before looking for a mistake in the ads.
Duplicate tags. The most common technical mistake: the GA4 code sits both in the site’s template and in Google Tag Manager. Every pageview gets counted twice, the bounce rate drops to near zero, and event counts double. Check it in a minute using GTM’s preview mode or a tag-debugging extension.
Untagged campaigns. Covered above: ads without UTM parameters, or with a tag made “manually from memory,” fragment into several sources or vanish into direct. Comparing channels becomes meaningless.
The team’s own visits. A developer, marketer, front desk, and owner open the site dozens of times a day. Internal traffic needs to be excluded by IP or a separate parameter, otherwise on small volumes it skews everything.
Cross-domain transitions. If payment or booking happens on a different domain, without cross-domain tracking set up, the session breaks and the purchase gets attributed to a “referral” from the payment system instead of the ad.
That’s exactly why GA4 and GTM setup is one line item in MOVE’s Google Ads setup: we tag things once and correctly, instead of trying to explain later why the report shows twice as many leads as the CRM.
Six Mistakes in Traffic Analytics
- Watching traffic instead of results. A 40% rise in sessions with no rise in leads is a reason to ask questions, not celebrate. Judge channels by key events and cost per event.
- Judging a site-wide average. An average engagement rate or average time says nothing: one page holds people for five minutes, another loses them in three seconds. Look at it broken down by page and channel.
- Drawing conclusions from small numbers. Two leads last week and five this week look like “2.5x growth,” but that’s noise. Compare periods with enough events to be meaningful.
- Changing several things at once. You update the site, change the creatives, and add a channel in the same week, and now you can’t tell what actually worked. Change things one at a time and mark change dates with annotations.
- Not reconciling GA4 with the CRM. GA4 counts forms, the CRM counts customers. If there’s a gap between them, the problem is either in lead handling or in event setup. Both are worth digging into.
- Setting it up once and forgetting it. The site gets updated, buttons get renamed, forms get moved, and events stop firing. Check once a month that every key event is still tracking.
What to Do With the Findings
Website traffic analytics should end in a list of actions, not a report. After every weekly review, one to three tasks should come out of it, each with a named owner: the ad vendor, the developer, the front desk staff who answer calls. If you’re weighing Google Ads against targeted ads as your main paid channel, the comparison in Google Ads vs. Meta targeting will help you decide based on data instead of habit.
If the site isn’t ready for measurement at all (no forms, calls aren’t tracked, pages are slow), sort out the site itself first: development and redesign with analytics built in from the start is cheaper than trying to measure something that wasn’t built to be measured.
Not sure your GA4 numbers are telling the truth, and don’t know where to start? We check exactly this as part of our ad account and GA4 audit: we go through events, UTM tagging, and ad-account connections, show you where the data is skewed, and write out the tasks: what to fix first, and who owns it. Once analytics tells the truth, every next hryvnia spent on ads works better.