Google Analytics Analysis: Decode GA4 Without a Data Expert
Analyzing Google Analytics well comes down to six GA4 reports, not the entire dashboard. Open Traffic acquisition, Pages and screens, Events, Conversions, Path exploration, and Retention. Track just four numbers: active users, engagement rate, conversions, and traffic source. The golden rule: a raw number means nothing on its own, so always compare one period to the previous one and break results down by channel. Without conversions set up, GA4 can't tell a visit from a sale. And never draw conclusions from same-day numbers: processing can take up to 48 hours.
The essentials
- GA4 no longer measures bounce rate the old way: the core metric is now engagement rate (sessions longer than 10 seconds, 2 page views, or a conversion).
- 90% of the value of an analysis sits in 6 reports: Traffic acquisition, Pages and screens, Events, Conversions, Path exploration, and Retention.
- A number only means something once you compare it: always look at one period against the previous one, and break results down by acquisition channel.
- Without conversions set up (the key: events marked as conversions), GA4 can't tell a visit apart from a business result.
- GA4's data processing delay can reach 24 to 48 hours: don't draw any conclusions from same-day numbers.
Table of contents
- What GA4 actually measures (and what changed)
- The 6 reports to open first when you run a small business
- Reading a metric without getting it wrong: the 4 numbers that matter
- Comparing your data to give it meaning
- Setting up conversions: the step 80% of small businesses skip
- Interpretation mistakes that skew your decisions
- Building a readable dashboard from GA4
- Systematizing your web data analysis without a data expert
- Frequently asked questions
What GA4 actually measures (and what changed)
GA4 measures your visitors' actual behavior as events, not as isolated page views like the old version did. That's the fundamental shift to understand before anything else. If your account isn't set up yet or the tag isn't installed, start with our guide GA4 for beginners: getting started step by step, then come back here for the analysis.
Universal Analytics, the previous generation that Google shut down in July 2023, counted sessions and page views. GA4 thinks in actions: a click, a scroll, a download, a video play. Every interaction becomes a named event. That logic fits how people actually browse today, moving between mobile and desktop, sometimes between an app and a website within the same hour.
The habit most people are missing the first time they open a GA4 account: looking for bounce rate as the main reference point. It still exists, but it no longer drives anything. The new core metric is called engagement rate: the share of sessions that last more than 10 seconds, generate at least two page views, or trigger a conversion. A session that used to count as a "bounce" can now count as "engaged" if the visitor read your page for 30 seconds.
What can you analyze with Google Analytics?
With Google Analytics, you analyze four broad categories of questions: where your visitors come from, what they do on the site, what brings them back, and what turns into a business result. You see traffic sources (Google, social media, email, direct links), the most visited pages, the devices used, the countries, and the actions that actually matter, like purchases or submitted forms. It's a thermometer for your online presence, not a guessing tool.
Why old Universal Analytics benchmarks no longer apply
Universal Analytics benchmarks no longer apply because GA4 redefines the very basis of counting. The old version counted any single-page session as a failure, even if the visitor had read an entire article for two minutes. Today, that same session counts as engaged.
In practice, a content site could show a seemingly disastrous 70% bounce rate while readers were perfectly satisfied. The official Google Analytics help documentation on engagement rate details these thresholds. Judging a GA4 report by an old bounce benchmark leads to bad decisions: it's like comparing Celsius to Fahrenheit, nothing more.
The 6 reports to open first when you run a small business
Six reports hold 90% of the value of a GA4 analysis for a small business: Traffic acquisition, Pages and screens, Events, Conversions, Path exploration, and Retention. Everything else is detail you can explore later, if you genuinely have the time.
Most guides sort these reports by menu, the way Google organizes them. That's the wrong angle. A business owner doesn't think in "menus," they think in questions: who's coming to me, what are they doing, what's paying off. I've made a habit of reorganizing these six reports around business questions instead. Here's the matrix I use on every account I work with.
| Business owner's question | GA4 report | What it reveals |
|---|---|---|
| Where do my visitors come from? | Traffic acquisition | Channels, sources, campaigns |
| Which pages are working? | Pages and screens | Page views, time spent, exits |
| What exactly are they doing? | Events | Clicks, scrolls, videos, forms |
| What pays off? | Conversions | Sales, quotes, qualified leads |
| What path do they take? | Path exploration | Movement between pages |
| Do they come back? | Retention | Loyalty, repeat visitors |
Acquisition: where your visitors come from
The Traffic acquisition report answers the most strategic question: which channels actually fill your site. You see the split between organic traffic (Google, unpaid), direct, social media, email, and paid campaigns.
On the small business accounts we've audited, the useful habit is to look not at raw volume but at volume weighed against engagement. A channel bringing 5,000 visitors with a 20% engagement rate is worth less than one bringing 800 with 70%. To dig deeper into this logic, our guide to the channels that actually drive your sales covers attribution suited to small businesses.
Engagement: what they do on the site
The Pages and screens and Events reports tell the story of actual behavior: how long people stay, which pages hold their attention, where they drop off. It's your friendly surveillance camera.
Spot your entry pages with high traffic but low engagement. Those are your leaks. A homepage viewed 3,000 times with an 8-second average time on page signals a problem with the promise, the speed, or the clarity. On the other hand, a product page with a 2-minute average read time deserves more traffic sent its way.
Conversions: what actually pays off
The Conversions report is the only one that connects your data to your revenue. Without it, you're measuring activity, not results. It lists the events you've flagged as important: a purchase, a quote request, a newsletter signup, a phone call.
A point many people miss: GA4 doesn't guess what matters to you. You have to tell it. We'll come back to this in detail further down, because it's the step that changes everything.
Reading a metric without getting it wrong: the 4 numbers that matter
Four metrics are enough to run a small or medium business: active users, engagement rate, conversions, and traffic source. Guides that drown you in 20 metrics turn you into a spectator, not a decision-maker.
That's a deliberate stance on my part, against the grain of a lot of content out there. After five years reading GA4 accounts, I've noticed that owners who run their business well never track 20 numbers. They track four, every week, and they act on them. Piling on metrics doesn't create clarity, it creates paralysis. Four numbers you understand beat twenty numbers you glance at without deciding anything, as our list of 10 marketing KPIs to track when you run a small business also points out.
Engagement rate, sessions, active users: what's the difference?
These three terms describe distinct things that get mixed up constantly. An active user is a unique person who visited your site during the period. A session is a visit: the same person can generate three sessions by coming back three days in a row. Engagement rate measures the share of quality sessions, the ones where the visitor does more than just glance around.
The classic mistake is celebrating a spike in sessions without checking engagement. Picture a store where a thousand people walk in and out in two seconds. Plenty of foot traffic, zero sales. Engagement rate is the percentage of people who actually did their shopping.
Why your GA4 numbers differ from Search Console
GA4 and Search Console show different numbers because they don't measure the same thing, or at the same moment. Almost no top-ranking guide explains this clearly, yet it's the most common source of confusion.
Search Console counts clicks from Google's results, meaning what happens before someone reaches your site. GA4 counts what happens after, once the person has landed. Gaps show up between the two: a visitor can click and leave before GA4 loads, or a cookie blocker can prevent GA4 from measuring the visit while Google still counted the click. A 10 to 20% gap between the two tools is normal. To connect both readings, our complete guide to running your SEO with Search Console shows how to cross-reference the two sources without losing your mind.
Comparing your data to give it meaning
A single GA4 number means nothing on its own: only comparison gives it meaning. Is 5,000 visitors good or bad? Impossible to say without a reference point. Always compare along two axes: over time, and across segments.
Period-over-period comparison is the first move. GA4 offers a "Compare" option in the top right of every report. Turn it on every time. That lets you set the last 30 days against the previous 30, or this month against the same month last year to cancel out seasonality. A toy store comparing December to November will always see a misleading spike.
On a client project in the professional equipment sector last year, traffic seemed to collapse by 15% from one month to the next. The year-over-year comparison actually showed a 22% increase. The dip was just a normal summer lull. Without the right comparison window, the owner was about to cut an ad budget that was working perfectly well.
Period over period, segments, and channels
Segmenting means splitting your audience into comparable groups to spot what's working. Compare the engagement rate of organic traffic to that of social media, or the conversion rate on mobile to desktop. The gaps reveal your priorities. Here are engagement benchmarks observed by channel, to adjust to your own industry.
| Acquisition channel | Benchmark engagement rate | What it means |
|---|---|---|
| Organic traffic (SEO) | 60 to 75% | Qualified visitors, clear intent |
| 55 to 70% | Audience that already knows you | |
| Social media | 35 to 50% | Curiosity, more fragile attention |
| Display / banner ads | 20 to 40% | Clicks often low in intent |
These ranges are ballpark figures drawn from accounts we've observed, not universal standards. Your industry, your offer, and your site speed all shift them. The goal isn't to hit an absolute number, but to understand which channel brings you people who are genuinely interested.
What are the 4 types of data analysis?
The four types of data analysis are: descriptive (what happened), diagnostic (why), predictive (what's going to happen), and prescriptive (what to do). GA4 mainly equips you for the first two. The descriptive layer shows you the drop in conversions. The diagnostic layer, by cross-referencing segments, tells you it's coming from mobile on the paid channel. Predictive and prescriptive analysis need a layer of human interpretation, or a tool that connects the sources together. Our method for making fast decisions from web data walks through these steps one by one.
Setting up conversions: the step 80% of small businesses skip
Without conversions set up, GA4 can't tell a visit apart from a sale, and your entire analysis stays blind. This step matters more than exploring fifty reports.
Exploring fifty reports without telling GA4 what matters to your business is like studying the weather in a thousand cities without knowing where you live. In our work with small businesses, a large majority of the accounts we take over have no conversions set up at all. Traffic gets measured, but the business result stays invisible. The tool is running on empty.
In GA4, a conversion is an event you flag as important: a form submitted, a purchase completed, a click on your phone number. Until you've flagged it, GA4 treats it as just another ordinary event among hundreds of others.
Marking an event as a conversion
To flag a conversion, go to Admin, then Events, and toggle "Mark as key event" on the event you want. That's it. GA4 keeps the data already collected and starts surfacing it in the Conversions report. The step-by-step procedure is in the official Google Analytics help on key events.
Before this step, check that the event actually exists. Some are collected automatically (scroll, first visit), others need to be set up through Google Tag Manager, Google's free tool for adding tags without touching your code. For an online store, the "purchase" event needs to be wired up correctly, or you won't measure any sales at all.
Connecting GA4 to Google Ads and Search Console
Connecting GA4 to your other Google tools turns separate silos into one overall view. In Admin, under "Product links," connect Google Ads and Search Console in a few clicks. That link lets you see which paid campaigns actually generate conversions, and which Google queries bring in engaged traffic. It's also where attribution plays out: which channel deserves credit for a sale. Without these links, you judge each tool in isolation, like reading three pages of the same book out of order.
Interpretation mistakes that skew your decisions
Five reading mistakes show up in almost every account we audit, and they cost real marketing decisions. Knowing them is worth more than a course on metrics. They come on top of the 8 GA4 configuration mistakes that skew small business data, which need fixing before any interpretation.
The first: drawing a conclusion from too little traffic. With 40 sessions on a page, a conversion rate that goes from 2% to 4% proves nothing (that's just one more sale). At that scale, randomness outweighs the signal. Wait for at least a few hundred sessions before judging a trend.
The second is about real time. GA4 shows an appealing "Realtime" report displaying visitors present right now. Never run your business off it. It's a verification tool, not an analysis one: the real data takes 24 to 48 hours to stabilize, according to Google Analytics help. Judging a day by its morning numbers is a straight path to a mistake.
Internal traffic, bots, and misleading real-time data
Your own traffic and bot traffic inflate your numbers if you don't filter them out. When you and your team check the site ten times a day, GA4 counts you as visitors. On a small site, that skews everything. In Admin, set up an internal traffic filter based on your IP address. GA4 also offers automatic filtering of known bots, but it doesn't catch everything. A sudden, unexplained spike of direct traffic at night? Often bots, not customers.
Drawing conclusions from too little traffic
The rule I apply on the ground: no marketing decision below 100 conversions or 1,000 sessions over the period analyzed. Below that, statistical noise dominates. An A/B test on a low-traffic page can look decided after a week when it actually needs a month. Comparing a Tuesday to a Sunday, or a sale week to a slow week, also skews the diagnosis. Patience is an analysis skill in its own right.
Lysible applies this same discipline automatically: it compares your periods at a comparable volume and waits for a significance threshold before alerting you.
Building a readable dashboard from GA4
A good GA4 dashboard fits on a single view and takes fifteen minutes to read each week. If you have to click through ten menus to understand your activity, the dashboard has failed. The full build, in GA4 Explorations or Looker Studio, is covered step by step in our guide to building a Google Analytics dashboard.
Bring your four core metrics together on a single page: active users, engagement rate, conversions, and the breakdown by traffic source. Add the comparison with the previous period on each one. Nothing more to start with. You should be able to answer three questions at a glance: is it going up, where's it coming from, and is it paying off.
Which metrics to bring together in a single view
Bring together the metrics that call for action, not the ones that just look good. Total page views is a nice number but rarely actionable. Conversion rate by channel, on the other hand, tells you where to put your next dollar. Here's a simple priority matrix, ranked by impact on the decision.
| Metric | Impact on the decision | Show on the main view? |
|---|---|---|
| Conversions by channel | High | Yes |
| Engagement rate by source | High | Yes |
| Active users (trend) | Medium | Yes |
| Most visited pages | Medium | Yes |
| Total page views | Low | No |
| Detailed demographic data | Low | No |
From GA4 report to weekly decision
Turn your GA4 review into a fifteen-minute routine every Monday. The first week, write down your baseline numbers. In the following weeks, only look at what's changed: what moved, in which direction, and why. This discipline beats a deep-dive analysis once every three months.
To structure this routine and automate the reading, our Google Analytics analysis method that turns GA4 into decisions offers a reusable template. GA4 remains free in its standard version, which makes it an accessible starting point for any small business. The alternative Matomo exists for businesses that are sensitive to privacy and cookie concerns.
Systematizing your web data analysis without a data expert
Reading GA4 every week, cross-referencing Search Console, checking your Ads campaigns: the method works, but it takes time and means juggling multiple tabs. That's exactly the wall business owners hit once the routine is in place.
Centralizing these sources into a single read, translated into clear actions, saves the hours a small business doesn't have. That's the approach we built with Lysible: connecting GA4, Search Console, and Google Ads to go from raw numbers to decisions, without hiring a data specialist or depending on an agency. You keep control, you understand what's happening, and you act fast.
Frequently asked questions
What can you analyze with Google Analytics?
Google Analytics covers four broad categories of questions about your site: where your visitors come from (Google, social media, email, direct access), what they do (pages viewed, time spent, navigation path), whether they come back (new or returning), and the business results (purchases, form submissions, calls). You also see the devices used, geographic locations, and which pages keep people or make them leave. It's a complete thermometer of your online presence. For it to actually help, set up your conversions so GA4 can tell a simple visit apart from a real result for your business.
What are the 4 types of data analysis?
The four types are descriptive (what happened), diagnostic (why it happened), predictive (what's likely to happen), and prescriptive (what to do about it). GA4 mainly equips you for the first two: it shows your numbers and, by cross-referencing segments, helps explain their causes. The descriptive view reveals a drop in conversions; the diagnostic view traces it to mobile or a specific channel. Predictive and prescriptive analysis need human interpretation, or a tool that connects your data sources and suggests concrete recommendations.
Is Google Analytics free?
Yes, GA4 offers a standard version that's entirely free, which covers the needs of nearly every small business. You measure traffic, conversions, sources, and behavior at no cost. A paid version exists, Google Analytics 360, aimed at very large companies with massive volumes and advanced needs: its cost runs into the tens of thousands of euros a year. For a small business, the free version covers everything. The only real investment is learning time, which resources like Google Analytics Academy help cut down.
What's the difference between users, sessions, and engagement rate in GA4?
An active user is a unique person who visits your site. A session is a visit: the same person generates several sessions if they come back on different days. Engagement rate measures the quality of those sessions, meaning the share of visits that last more than 10 seconds, include at least two page views, or trigger a conversion. Mixing up these three concepts leads to mistakes: a spike in sessions with no engagement looks like a store where a thousand people walk in and out in two seconds. Plenty of foot traffic, but no purchases.
Why don't my GA4 numbers match Search Console?
Because the two tools don't measure the same thing. Search Console counts clicks from Google's results, so it captures what happens before someone reaches your site. GA4 measures what happens after, once the visitor has landed. Gaps appear between the two: someone can click and leave before GA4 loads, or block tracking cookies while still being counted by Google. A 10 to 20% gap is normal and doesn't signal a problem. Use each tool for what it measures best, rather than trying to make the numbers match exactly.
How do you set up a conversion in GA4?
Go to Admin, then Events, and toggle "Mark as key event" on the event you want to track, for example a form submission or a purchase. GA4 then surfaces it in the Conversions report. First check that the event is actually being collected: some are tracked automatically, others need to be set up through Google Tag Manager. For an online store, the purchase event needs to be wired up correctly, or no sales get measured at all. It's the most overlooked step, and yet the most decisive one: without a conversion set up, GA4 has no idea what matters to your business.
Which GA4 reports should a small business check first?
Six reports cover the essentials: Traffic acquisition (where your visitors come from), Pages and screens (which pages perform), Events (what people do), Conversions (what pays off), Path exploration (where they go), and Retention (whether they come back). To start, focus on Acquisition and Conversions: they connect your traffic to your results. There's no need to explore the dozens of other reports until you've mastered these six. Six reports understood and checked every week beat an entire dashboard skimmed once a quarter with no decision to show for it.
How long does GA4 take to show data?
GA4 shows some numbers in real time in its dedicated report, but stabilized data generally takes 24 to 48 hours to process. Don't draw any conclusions from same-day numbers: they're incomplete and misleading. The "Realtime" report is there to confirm a new setup is working, like a button firing correctly, not to run your business. For a reliable read, wait two days and always compare a full period to the previous one. Patience is part of the method: judging too soon is like tasting a dish before it's done cooking.
Isaac SIKORSKI
With Lysible, I want to give businesses back control of their online presence. A website you actually understand is one that brings in real opportunities.
