Before opening Analytics: what you need to be clear about
Google Analytics doesn’t measure people, it measures sessions and events. That distinction, which sounds academic, completely changes how you interpret the numbers. If your website receives 2,000 visits a month but only 400 users, you don’t have 2,000 interested people: you have 400 people who came back several times. Confusing both figures is the most common mistake for beginners.
Before looking at any report, check two things. First: that the measurement tag (gtag.js or Google Tag Manager) is installed on all pages, not just the homepage. Second: that the property is GA4, because Universal Analytics stopped processing data and many old tutorials are still circulating.
GA4 reorganized the interface compared to the previous version. Instead of “Audience,” “Acquisition,” and “Behavior,” you now have Reports with a different structure. For a beginner, the most worthwhile path is this one:
1. Realtime
It’s used to verify that the installation works. Open your website on your phone, go to Realtime, and check that your visit appears. If it doesn’t appear, the problem is with the tag, not the report.
2. Reports > Acquisition > Traffic acquisition
Here you see where people come from: organic (search engines), direct, referral, social, email. It’s the most consulted report and the most misinterpreted one. “Direct” traffic isn’t always someone who typed your domain: it also groups visits without a referrer, links from apps, PDFs, and part of the dark traffic that no tool resolves well.
3. Reports > Engagement > Pages and screens
Which pages are viewed and for how long. Keep in mind: average engagement time in GA4 only counts active tabs, so a long read with your phone locked may register only a few seconds.
A specific case that teaches more than twenty definitions
Imagine a physiotherapy clinic with 1,200 monthly sessions. The acquisition report shows 55% organic traffic and 30% direct. The temptation is to celebrate the SEO. But when segmenting by device, it turns out that 80% of the direct traffic comes from mobile and coincides with the Tuesday booking peak. Most likely these are patients who saved the website or arrive from the clinic’s WhatsApp link: it’s branded traffic, not discovery. Without that segmentation, the conclusion would have been wrong.
This kind of nuance can only be detected by cross-referencing it with Search Console, which does tell you which queries generated clicks from Google. Analytics tells you what happened on the website; Search Console tells you what happened before arriving.
Limitations you should accept from the start
- Cookies and consent: with a poorly configured consent banner, GA4 loses a portion of visits. According to Google’s official documentation, data modeling helps estimate that gap, but it’s an estimate, not an exact count.
- Ad blockers: a portion of users will never run the tag. No configuration solves this 100%.
- Bounces: in GA4 a bounce is a session without engagement, not a single-page visit. If you compare data with old Universal Analytics screenshots, the numbers won’t match and it doesn’t mean you’ve lost traffic.
- Bots: automatic traffic filtering doesn’t catch everything. A sudden spike with zero duration and one hundred percent bounce rate is usually noise, not a viral campaign.
Three habits that make a difference after two months
First, note in a digital notebook the date of any change to the website (new landing page, redesign, template change). When traffic drops, that notebook is worth more than any diagnostic tool.
Second, define at least one realistic conversion: form submission, phone click, visit to the thank-you page. Without conversions, Analytics only counts curious onlookers.
Third, review once a month, not every day. Daily variations are statistical noise and lead to hasty decisions.
What not to expect from this tool
Google Analytics won’t tell you why a user left, or what they thought when they saw your price, or whether the competition is beating you. Nor does it replace a CRM: cross-channel attribution is still imperfect and no dashboard fully resolves the journey of someone who discovers you on Instagram, searches for your brand on Google, and buys from their laptop three days later. For small projects, sometimes a simple Search Console dashboard plus the business’s own data is enough; Analytics adds value when there’s volume and several channels to compare.