UTM parameters explained, and the naming rule that saves you

Aug 22
3 min read

Updated: 5 days ago

Introduction

A UTM parameter is a label you bolt onto a link so your analytics knows where the click came from. Without them, a visitor from a newsletter, a paid advert and a partner's website can all arrive looking identical.

They are simple to add and easy to ruin. Almost every messy analytics report is a tagging problem rather than a tool problem.

1. UTM parameters explained: the five you can use

They are query strings appended after a `?` on any link.

`utm_source` — where it came from. `newsletter`, `facebook`, `partner-name`. `utm_medium` — the kind of channel. `email`, `cpc`, `social`, `referral`. `utm_campaign` — the specific effort. `spring-launch`, `ramadan-offer`. `utm_content` — which version, when you are testing more than one. `hook-a`, `hook-b`. `utm_term` — the keyword, mostly for manual search tagging.

Source, medium and campaign carry almost all the value. Content earns its place when you are testing creative. Term is rarely needed.

2. The rule that matters more than the tags

Lower case, always. Hyphens, never spaces. One spelling per thing.

`Facebook`, `facebook` and `FaceBook` are three different sources to your analytics. So are `fb` and `facebook`. Nothing warns you — the report simply shows six rows where there should be one, and the totals you were trying to compare become meaningless.

Decide the spellings once, write them down, and use that list every time. This single convention is the difference between a usable report and a pile of near-duplicates.

3. Do not tag internal links

The most damaging mistake. Adding UTMs to a link between two pages of your own site starts a brand new session and attributes the visitor to whatever the tag says.

The effect is that your own website becomes your best-performing traffic source, and the real source of the visit disappears. Only tag links arriving from somewhere you do not control.

4. Do not tag Google Ads either

Google Ads already passes its own click data to Analytics automatically when the accounts are linked. Adding UTMs on top overwrites that with something less detailed.

Link the accounts and leave the URLs alone. Tag the channels that have no automatic connection: email, newsletters, partner links, printed QR codes, messaging apps, anything offline.

5. Tag the offline things too

This is the underused half. A QR code on a bill, a poster, a printed card or a receipt can carry a tagged link, which turns an untrackable channel into a measurable one.

Use a short redirect if the tagged URL is too long to print sensibly. It is the cheapest way to find out whether the printed thing is doing anything at all.

6. Keep a single sheet of every tag you have used

One spreadsheet: the campaign, the tagged URL, when it went out, who made it.

It takes seconds per link and it prevents the two failures that make tagging useless — two people spelling the same source differently, and nobody remembering what `campaign-3` referred to six months later.

It also lets a new person pick up the routine without inventing a second convention alongside yours.

7. Check the report, not the link

Tagging is only correct if the report says what you expected. After sending anything tagged, open the traffic acquisition report and confirm the source and medium appear as intended.

Mistakes are cheap to fix on the next send and impossible to fix retrospectively — the data is already recorded under the wrong label.

8. Decide who owns the convention

Tagging fails at the second person. One person's `email` is another's `newsletter`, and the report splits.

Give the convention an owner and put the agreed list where links get made, not in a document nobody opens. When a new channel appears, that person decides the spelling and adds it to the list.

It is a small piece of governance and it is the difference between tagging that stays useful for years and tagging that degrades within two months.

Conclusion

Use source, medium and campaign on every external link, add content only when testing versions, and fix your spellings in writing before you start.

Never tag internal links, let Google Ads pass its own data, tag the offline channels that otherwise stay invisible, log every tag in one sheet, and check the report after the first send rather than assuming it worked.