UTM Parameters: A Naming Convention That Survives a Team

Created on 16 November, 2026 • Link tracking & analytics • 7 views • 5 minutes read

What each UTM parameter is for, which ones Google Analytics 4 actually reports, and how to write a convention that does not collapse the moment a second person uses it.

UTM parameters are query-string tags appended to a destination URL so your analytics can attribute the visit to a campaign. They are old, boring, and still the most reliable attribution mechanism available, because they travel in the URL itself rather than depending on a referrer that half the internet strips.

They also fail constantly, and almost never for technical reasons. They fail because two people on the same team wrote facebook and Facebook, and the report now shows two sources.

The parameters, and what each is actually for

Google Analytics 4 recognises the following campaign parameters:

  • utm_source - where the traffic came from: a site, a newsletter, a partner, a physical placement. newsletter, linkedin, conference-badge.
  • utm_medium - the channel type: email, cpc, social, qr, affiliate. This is the one that feeds channel grouping, so it is the one most worth standardising.
  • utm_campaign - the campaign, promotion or product the link belongs to.
  • utm_term - the paid keyword.
  • utm_content - which specific creative or placement, used to tell apart two links that are otherwise identical: header-button versus footer-text.
  • utm_id - a campaign ID, meant to match the IDs used in campaign cost-data uploads.
  • utm_source_platform - the platform that sent the traffic, such as an ad-management platform.
  • utm_creative_format and utm_marketing_tactic - creative type and targeting type. Google documents both but notes they are not currently surfaced in GA reports, so treat them as forward-looking rather than useful today.

In practice utm_source, utm_medium and utm_campaign carry nearly all the value, and Google recommends always setting all three. Omitting one produces (not set) in reporting, which is the analytics equivalent of a shrug.

The rule that matters more than all the others

UTM values are case sensitive. utm_source=google and utm_source=Google are two different sources in GA4, and they will appear as two rows forever. The same applies to SpringSale versus spring_sale versus spring-sale.

So: lowercase everything, always, with no exceptions and no judgement calls. This is the single rule that prevents most UTM damage, and it is the one that breaks first when someone builds a link by hand in a hurry.

Second rule: pick one word separator and never use the other. Hyphens are the usual choice because underscores can be visually swallowed by link underlining. Either is fine; mixing them is not.

Third rule: no spaces, ever. A space becomes %20 or + depending on who encodes it, and you now have two values again.

A convention that holds up

The goal is that any person on the team, given the same campaign, produces the same string. That means a closed list, not a style guide.

Fix the medium list. This should be short and never grow casually. Something like: email, social, cpc, display, affiliate, referral, qr, print, sms, push. Ten values. If someone needs an eleventh, that is a conversation, not a decision they make alone in a URL field.

Fix the source list per medium. Sources are allowed to grow, but each new one gets written into the shared list the first time it is used.

Give campaigns a date-first structure. 2026-03-spring-sale sorts chronologically, is unambiguous when the same campaign runs annually, and makes a year-old report readable. Compare with springsale, which in two years' time means nothing.

Use utm_content for the thing you will want to compare. If you will ever ask "did the header or the footer link do better?", that distinction belongs in utm_content, not in three separate campaigns.

Keep the list where people actually work. A shared spreadsheet with one row per link - date, owner, destination, each parameter, final URL - is unglamorous and works. The alternative is six months of archaeology.

Building the URLs by hand is where typos enter. A UTM link generator removes that step, and a slug generator is useful for turning a campaign name into a consistent lowercase, hyphenated value.

UTMs and short links together

These are complementary, not alternatives, and the order matters. Put the UTM parameters on the destination URL, then shorten the whole thing. The shortener stores the full URL including the query string and hands it to the browser, so the parameters arrive intact and your analytics sees the campaign.

Do not try to put UTM parameters on the short link itself. Some services pass unknown parameters through to the destination and some do not, so the behaviour is service-specific and will surprise you at the worst moment. Test it once with your own service if you need to rely on it.

This combination also fixes a specific annoyance: without UTMs, a redirect can cause your analytics to attribute the visit to the shortening domain as a referral, because that is the last thing the browser saw. With UTMs on the destination, campaign attribution wins. The mechanics are in how URL shorteners actually work.

Things that will bite you

UTM parameters are visible to the user. They appear in the address bar, in shared links, and in anything someone copies and pastes. Do not put anything in them you would not want read aloud - no internal codenames you have not announced, no personal data, no pricing experiment names.

People share tagged URLs. When someone copies the link from their address bar and posts it, your campaign tags travel with it, and their visitors are attributed to your email campaign. This is unavoidable and usually small, but it explains odd long tails in reporting.

Tagging internal links corrupts sessions. Putting UTM parameters on a link from one page of your site to another starts a new campaign attribution mid-visit and destroys the original source. Never tag internal navigation.

Some platforms strip or rewrite query strings. Certain apps and email clients modify links in transit. If a channel's numbers look impossible, test the actual delivered link in that app before blaming analytics.

And remember that the number of tagged sessions will not equal the click count on your short link. That gap is normal and has its own explanation - see why clicks and sessions never match.