You will be able to spot and avoid the common tagging errors that split or overwrite your data.
Farah's assistant had a bright idea. Their homepage had a banner for the new work trousers, and she wanted to know how many people clicked it, so she tagged the banner link with utm_source=homepage and utm_medium=banner. A month later Farah noticed something odd in GA4. Sales from her newsletter had almost vanished, and a new source called "homepage" was suddenly one of her best performers.
Nothing had changed in the newsletter. The banner tag had quietly taken the credit. This lesson covers that mistake and the other common ones, each of which can make a report wrong without anyone noticing.
When a visitor clicks a tagged link, GA4 reads the tags as a statement about where this visit came from. That is fine for a link in an email. On a link inside your own website, it is a false statement, because the visitor was already there.
Here is what happened to Farah's data. A subscriber clicked the newsletter, landed on the homepage and was recorded as coming from newsletter, email. Then they clicked the banner. GA4 read the new tags and recorded the rest of the visit as coming from homepage, banner. When they bought the trousers a few minutes later, the purchase was credited to the banner, and the newsletter got nothing.
The rule is simple: tags go only on links that point to your site from somewhere else. To measure clicks on your own banners and buttons, use GA4 events instead, which module 3 covers. They record the click without rewriting where the visitor came from.
The same goes for links in your own online shop's emails, such as order confirmations, if they point back to the site. Those are fine to tag, because they come from outside. Links between pages of the site never are.
Lesson 2.2, A naming convention your whole team can follow, explained why "Email" and "email" are different values. In practice the damage comes from several people tagging links over months. One person writes "insta", another "instagram", a third "ig". The campaign name "2026-11-work-trousers" turns into "nov-trousers" in the hands of a freelancer who never saw the rules.
Each variation becomes its own row in the report. A campaign that brought 400 visits shows up as five rows of 80, and the one row you happen to look at looks weak. Worse, the split is invisible unless you go looking for it. Nobody sees an error message.
The fix is the shared rules page from lesson 2.2 and the tagging sheet in lesson 2.4, plus a habit: once a month, sort your sources and campaigns alphabetically in the traffic acquisition report and scan for near-duplicates. Misspellings sit next to each other when sorted, which makes them easy to catch.
A tagged link is long and ugly. Pasted into a WhatsApp message or an Instagram story, it looks like spam, and people are less inclined to tap a link that looks like a ransom note. Some people respond by trimming the tags off before sharing, which defeats the purpose.
Use a link shortener instead. A shortener gives you a short address that redirects to the full tagged link, so the tags still arrive at your site. Many website builders and email tools have one built in, and there are free stand-alone ones. QR codes work the same way: generate the code from the full tagged link, or from a short link that redirects to it.
Two cautions. First, check that your shortener passes the full link through, by testing it as described below. Second, if you use a free shortener and later lose access to the account, every printed QR code pointing at it can break. For anything printed, pick a service you control or a short page on your own domain.
A tagged link with a typo still works for the visitor, so nobody complains. The page loads, the customer buys, and the visit is filed under the typo or under direct. You only find out weeks later, if at all.
So test each new link before it goes out. Open a private or incognito browser window, paste the link, and check that the page loads with the tags visible in the address bar. Then open GA4 and go to the realtime report, which shows activity from the last few minutes. Within a minute or so you should see a user with the source and medium you set. If you see direct instead, something stripped the tags, usually a redirect or a shortener.
This takes about two minutes per link. Farah now tests the newsletter link, the bio link and every new QR code before they go out, and has not lost a campaign since.
One caution for later: once you set up an internal traffic filter in module 3, GA4 will exclude your own visits, and your tests may stop appearing in realtime. Lesson 3.2 explains how to test with the filter in place.
Doing the test takes less time than reading about it did, and the first time you see your own visit appear with the right labels, the whole tagging system stops feeling abstract.
Click one of your tagged links, then find the visit in GA4's realtime report and note the source and medium it shows.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).