The UTM builder that shows you how your URLs will be tracked

The platform, site, or newsletter name where the click came from. The most specific identifier of where traffic originates.

Good values: facebook, newsletter, google, partner-site.com. Keep it lowercase, one concept per source.

Common mistake: putting a campaign name here. utm_source=spring-sale is wrong. The source is where the link was clicked (e.g. facebook), not what the click was about.

The channel or vehicle. This is the single strongest signal analytics tools use to classify traffic.

Standard values: cpc or ppc (paid search/social), social (organic social), email, newsletter, referral, affiliate, display, video, organic, podcast, sms.

Anything outside these values lands in Referral or a generic bucket. If your team invents utm_medium=blog-share, analytics treats it as Referral, not Social. Pick from the standard list.

The named marketing initiative. spring-2026-launch, black-friday, webinar-product-demo.

Lowercase, hyphens not spaces. Prefix with year or quarter if you re-run annually: 2026q2-launch. Otherwise last year's launch merges with this year's launch in reports.

Do not cram creative variants or source info here. Same campaign with three creatives: same utm_campaign, different utm_content.

Which creative or variant was clicked. This is your A/B testing field.

Examples: banner-a vs banner-b, footer-cta vs hero-cta, v1 vs v2.

If you're running one campaign with three ads, the three links share utm_campaign and differ only in utm_content. That way your analytics can roll the campaign up or break it down by creative on demand.

Historically for paid search keywords. Google Ads auto-populates this with {keyword} in your ad URLs, so the exact keyword the searcher used shows up in analytics.

Leave this blank for anything other than paid search. Some teams use it for retargeting audience IDs, which is valid, but do not invent a use for it on every campaign. An email campaign with utm_term is usually a tagging mistake someone will spend two hours untangling later.

Facebook and facebook are two different sources to analytics. Tools do not normalize case for you.

Result: the same campaign appears as two separate rows. Totals are wrong. Reports split.

Fix: lowercase everything. Always. Enable the auto-lowercase toggle in the builder.

utm_campaign=spring sale gets URL-encoded to spring%20sale in the URL. Browsers handle it, but your dashboard shows it literally with the %20. Ugly in reports, and any tool that re-encodes the URL (email clients, ad platforms) may produce spring%2520sale (double-encoded).

Fix: hyphens. spring-sale.

Characters &, =, ?, #, +, % are reserved URL syntax. Using them inside a parameter value breaks parsing.

utm_campaign=Q3&Q4 becomes two parameters: one called utm_campaign with value Q3, and a separate one called Q4. Half your data vanishes.

Fix: q3-q4, or percent-encode deliberately (rare in practice).

Destination URL already has ?utm_source=google, and you add &utm_source=facebook. Most analytics tools keep one at random; some concatenate; some drop both.

You will not know which one was used until the numbers look off. The hygiene score flags this automatically.

Fix: strip existing UTMs from your destination URL before building.

utm_term was designed for paid search keywords. Adding it on an email blast or social post does not break anything technically, but it misleads whoever reviews the report.

A teammate opens the Campaigns panel, sees utm_term=early-access on an email, and spends an hour figuring out why an email is categorized as paid search.

Fix: leave utm_term blank unless the medium is cpc, ppc, or similar.

utm_campaign=launch three years in a row gives you three rows called launch that merge in any report without a date filter. Year-over-year comparisons become impossible.

Fix: prefix with year or quarter. 2024-launch, 2025-launch, 2026q2-launch. Year-over-year analysis becomes a one-click filter.

Click IDs are ad-platform identifiers appended automatically: gclid (Google), msclkid (Microsoft), fbclid (Facebook), ttclid (TikTok), li_fat_id (LinkedIn), twclid (X), rdt_cid (Reddit), pnclid (Pinterest), dclid (Google Display), irclickid (Impact).

They carry information UTMs do not (specific ad, ad group, keyword) and they influence classification. A gclid on a Facebook-tagged URL still classifies as Paid Search, because the click came through Google Ads.

Do not strip click IDs from URLs. Let them ride alongside your UTMs. Clickport uses both.

If your destination is a redirect (site.com/r/abc123 → site.com/landing), make sure UTMs land on the final URL, not just the redirect entry point. Most ad platforms strip custom query params across redirects unless you explicitly append them.

Build the tagged URL against the final destination. Let the redirect preserve the query string via server config (most do) or append them in the redirect logic.

SPAs use client-side routing (pushState). UTMs in the initial URL load fine. If the visitor navigates to /page-b via internal links, the URL changes and the UTMs are stripped.

Clickport captures UTMs on the first page-load event per session and pins them to the session. Internal SPA navigation does not lose attribution.

Shorteners like bit.ly and TinyURL preserve query parameters across the redirect if the long URL has them. Build the shortened link from the tagged URL, not the clean one. Otherwise you lose attribution at the redirect.

QR codes work differently. They encode the full URL directly (no redirect). Tagged URL in, tagged URL scanned.

Consent Mode v2 (Google) affects cookie-based tracking. UTMs are URL-based, not cookie-based, so they work identically regardless of consent state.

The analytics tool reading the UTMs may be gated by consent. The tags themselves are not. Clickport does not require consent, so UTMs are captured on every session.

UTMs label a single click. Whether that click gets credit for the conversion depends on the attribution model.

First-touch: the first UTM of the session wins. Good for campaigns whose job is to acquire a visitor.

Last-touch: the last UTM of the session wins. Good for campaigns whose job is to convert a visitor.

Clickport uses first-touch per session by default (see UTM tracking docs).