UTM parameters tracking guide

UTM parameters are five query-string tags appended to a URL. Used well, they are the cleanest source of marketing attribution you have. Used badly — and most teams use them badly — they produce a data swamp where the same campaign shows up under a dozen names and nobody trusts the numbers.

The five parameters and what each is for

There are five standard UTM parameters. Each one answers a specific question, and the discipline is to keep that meaning consistent across every link your team ever ships.

• utm_source — the site or platform sending the traffic. Examples: facebook, newsletter, partner_blog. Always lowercase, always one canonical name per source.
• utm_medium — the channel category: email, cpc, social, display, referral, affiliate. This is the bucket your analytics platform groups by — pick a small fixed list and never deviate.
• utm_campaign — the specific campaign name. Examples: q2_launch, holiday_sale_2026. This is the most-abused parameter; standardize the format or it becomes chaos.
• utm_content — distinguishes variants within a campaign. Use it for ad creative variants, link positions in an email, button-vs-link splits.
• utm_term — originally for paid-search keywords. Most teams now use it for audience or segment tags.

Three of the five (source, medium, campaign) are required for the link to be useful. The other two are optional but worth filling in when you'll want to slice the data later.

A taxonomy beats clever names

The fastest way to ruin marketing analytics is to let everyone invent their own UTM names. By month three, you have facebook, Facebook, fb, FB, meta, and Facebook_paid in utm_source, all referring to the same channel. Reports become unusable and trust collapses.

The fix is a published taxonomy: a short document everyone references before tagging a link. The taxonomy specifies the allowed values for each parameter, the format conventions, and the naming pattern for campaigns.

• Source — fixed list. facebook, instagram, linkedin, google, newsletter, partner, etc. Add to the list with explicit approval, never on the fly.
• Medium — fixed list of categories matching how your analytics platform groups channels. Usually six to ten values total.
• Campaign — formatted pattern: quarter_initiative_year or type_topic_date. Pick one and stick to it.
• Format rules — lowercase only, underscores not spaces, no special characters, no trailing punctuation.

The document fits on one page. The hard part is enforcing it.

Governance is what separates teams

Taxonomy without governance is a doc nobody reads. The teams that maintain clean tracking data over years do three concrete things.

1. Centralize link generation. Build or adopt a UTM builder — a shared spreadsheet, an internal tool, or a feature in your campaign tooling — that enforces the taxonomy. Free-text UTMs in production are how the swamp starts.
2. Audit weekly for the first quarter. Pull the unique values from utm_source, utm_medium, and utm_campaign in your analytics platform. If you see two spellings of the same source, fix the link or rewrite the data. Catching it in week one is a five-minute fix; catching it in month six is a project.
3. Document campaigns at launch. Every campaign with a UTM should have a one-line entry in a shared sheet — campaign name, owner, launch date, expected duration. When someone asks "what is q2_launch_v2?" three months later, the answer exists.