UTM tags: measure campaigns without naming chaos

A normal link such as example.com/contact can carry a query string: ?utm_source=newsletter&utm_medium=email&utm_campaign=summer-offer. The destination page stays the same, while Analytics can read the campaign labels.

These labels describe the link used. They do not prove that a particular person read a newsletter, that every click was human or that the campaign alone caused a sale. A forwarded link may retain the original labels, and a security scanner may open it before the recipient does.

Google's campaign URL documentation explains the parameters and where their values appear in Traffic acquisition. Keep actual account registrations or enquiries as separate, verified outcomes.

Direct traffic can include a typed URL or bookmark, but also visits whose origin cannot be determined. Apps, messaging services, documents and unlabelled email links may not send a useful referrer. Redirects that strip parameters can make the picture less complete too.

The dark-social research cited in the original guide illustrates that some messaging and social visits can lose their origin. It is not a fixed percentage you can apply to every website today. There is no universal maximum share of direct traffic that proves your tracking is broken.

Investigate your own pattern: which pages receive direct traffic, which links were distributed, whether parameters survive the redirect and whether analytics was permitted and loaded. A long landing URL is a reason to investigate, not proof that the visit was misclassified.

Set source, medium and campaign together and document any additional fields. Use a channel naming scheme that matches how your reporting is configured. Google Ads auto-tagging can provide richer campaign information when correctly linked to GA4; verify that setup before adding manual labels that might conflict with it.

Build links with Google's Campaign URL Builder or a shared spreadsheet, and test the final destination before publishing.

Parameter values are case sensitive: Newsletter and newsletter can create separate rows. The same problem occurs when one person writes paid_social and another writes social-ad for the same channel.

1. Use lowercase values.
2. Choose one separator, such as hyphens, and apply it consistently.
3. Maintain an agreed list of source and medium values.
4. Give each campaign a stable name and each creative a distinct content label.
5. Store the complete link, date, destination and purpose in one log.

Share the convention with agencies and colleagues before they publish. Avoid putting names, email addresses or other personal details in campaign parameters: URLs can enter browser history, analytics and server logs.

UTM parameters belong on incoming campaign links, such as newsletters, external social posts, QR codes and partner promotions. They should not label your own menu, footer or a button between two pages of the same website.

Internal tagging can pollute campaign and attribution information. It is not accurate to say that a new UTM value always starts a new GA4 session; the problem is inconsistent measurement and interpretation, not a guaranteed doubling of sessions.

• Use campaign labels on the link distributed outside your site.
• Use an internal click event if you need to measure a particular button.
• Keep navigation and contextual article links clean.
• Check whether redirects preserve the original incoming labels.

In GA4 Traffic acquisition, inspect Session source/medium and Session campaign alongside the outcomes you configured. Compare equivalent date ranges and distinguish events from users, sessions and confirmed accounts. Do not add different measures together as if they described the same people.

Ask three practical questions: which campaigns produce a measurable next step, which bring visits without that step, and where the reporting loses coverage. A low recorded result may mean poor performance, missing consent, a broken event or a mismatch between systems. Test those possibilities before deciding why a campaign failed.

For organic search, use Search Console beside GA4. Do not attach UTMs to your own Google search results or use campaign labels to invent missing organic queries.

Open a labelled link in the browsers and apps your audience uses. Check the destination, consent behaviour, form, confirmation and final account or enquiry record. Use test data clearly identified as internal so it can be excluded from customer results.

A public website analysis can identify visible signals, but cannot prove that every campaign event is received correctly. Start with Marketing Scan for website findings and keep a separate measurement check for your campaign's complete journey.