UTM Governance: Naming Conventions That Keep Attribution Clean

Every marketing report you will ever read starts with a decision someone made while pasting a link: what to type after utm_source=. Get that decision right consistently and your Sources report tells you exactly which channels make money. Get it wrong — Facebook here, facebook there, fb in the ESP — and the same channel splinters into three rows, your channel grouping misfires, and every downstream decision is made on fragmented data.

This is what UTM governance means: a small set of written rules, applied by everyone who creates a link, forever. This guide gives you the complete playbook — naming conventions, a source/medium taxonomy that matches Sealmetrics's default channel definitions, the errors that silently corrupt attribution, a spreadsheet template your team can adopt today, and a walkthrough of how Sealmetrics actually processes UTM parameters so you understand why the rules matter.

It matters even more in Sealmetrics than in cookie-based tools. Sealmetrics reads UTM parameters directly from the URL on every page load, without cookies — attribution works for 100% of your visitors, including the ones who reject or ignore consent banners. That's the good news. The flip side: your UTMs are the primary attribution signal, so their quality is your attribution quality.

The Five UTM Parameters, and What Each Is Actually For

Sealmetrics recognizes the five standard UTM parameters:

Parameter	URL format	Answers	Example values
Source	utm_source=	Which site or platform sent the click?	google, facebook, newsletter
Medium	utm_medium=	What kind of channel is it?	cpc, organic, email
Campaign	utm_campaign=	Which initiative paid for or produced the link?	spring-sale, brand-2025
Term	utm_term=	Which keyword (paid search)?	running+shoes
Content	utm_content=	Which variant of the ad or link?	banner-a, text-link

A fully tagged URL looks like this:

https://example.com/landing?utm_source=google&utm_medium=cpc&utm_campaign=spring-sale&utm_term=running+shoes&utm_content=banner-a

Two of these carry almost all the analytical weight. Source and medium together determine which row of the Sources report the visit lands in and — critically — which channel it's assigned to (more on that below). Campaign is how you compare initiatives against each other. Term and content are refinements: keyword-level analysis and creative A/B testing respectively.

The most common conceptual mistake is blurring the source/medium boundary. A useful mental test:

• Source is a proper noun — the name of a specific place: google, facebook, mailchimp, partner-blog.
• Medium is a category — a type of traffic that many sources can share: cpc, email, social, affiliate.

If you find yourself typing utm_medium=facebook, stop — Facebook is a source. The medium is social (organic posts) or cpc/paid (ads).

What happens when a parameter is missing

When a UTM parameter is absent from the URL, Sealmetrics reports "(not set)" for that field — except utm_medium, which shows as "(none)", aligned with GA4's convention. Untagged visits aren't lost: a visit arriving from an untagged external link is attributed to its referring domain with medium referrer (e.g. reddit.com / referrer), and a visit with no referrer at all is direct. See Understanding the "(not set)" value and Referral vs Direct Traffic.

Governance implication: you only need to tag links you control (ads, emails, social posts, partner placements). Organic search, untagged referrals, and direct traffic classify themselves.

Naming Conventions: The Non-Negotiables

These four rules prevent 90% of UTM chaos. Put them at the top of your governance doc.

1. Always lowercase

Sealmetrics reports group by the exact string value. Facebook, facebook, and FACEBOOK are three different sources — three separate rows in your report, three separate inputs to channel classification. There is no automatic case normalization at ingest (and deliberately so: silently rewriting your data would be its own problem).

The fix is procedural, not technical: every UTM value is lowercase, no exceptions. It's the single highest-leverage rule in this guide because case errors are invisible in the URL bar and only surface weeks later as duplicate rows.

2. snake_case or hyphens — pick one separator, never spaces

Multi-word values need a separator. Use underscores (black_friday_2026) or hyphens (black-friday-2026) — both are fine; what matters is that your team picks one and uses it everywhere. Mixed separators (black_friday in email, black-friday in ads) split one campaign into two rows exactly like a case mismatch does.

Never use spaces. A space in a URL gets encoded as %20 or + depending on the tool that built the link, which means "summer sale" can arrive as summer%20sale, summer+sale, or break the URL entirely. Spaces are how you end up with unexplainable garbage values in the Campaigns tab.

Avoid other special characters too (&, ?, #, = have structural meaning in URLs). Stick to a–z, 0–9, and your chosen separator.

3. One canonical name per entity

Decide the one true spelling of every recurring source and medium, write it down, and outlaw the synonyms:

• facebook — not fb, not facebook.com, not meta
• google — not google.com, not adwords
• newsletter — not news_letter, not nl

Domains make bad source values (facebook.com vs facebook) because they look plausible and thus survive review. Pick the bare platform name.

4. Stable, structured campaign names

Campaign names live forever in your historical data, so make them self-describing. A structure that scales well:

{year}_{initiative}_{detail}
2026_black_friday_early_access
2026_brand_awareness_q3
2026_product_launch_alpine

Prefixing the year keeps recurring campaigns comparable across editions (2025_black_friday vs 2026_black_friday) without ambiguity. Whatever structure you choose, the governance rule is the same: campaign names are created in the spreadsheet first (see below), never improvised in the ad platform at publish time.

A Source/Medium Taxonomy Aligned with Sealmetrics Channels

Here's why medium discipline pays off directly in Sealmetrics: the Channels tab of the Sources report buckets every session into a marketing channel by matching its utm_source, utm_medium, and utm_campaign against channel grouping rules. Sealmetrics ships with default channels — from Paid Search and Paid Social down to Referral and Direct — whose rules match on exactly the source and medium values in the canonical table on that page. Traffic that matches no rule lands in Unassigned.

So the taxonomy below isn't arbitrary: each recommended medium is one the default channel rules already recognize. Use these values and your Channels tab works out of the box, with zero custom configuration.

Traffic type	utm_source (examples)	utm_medium	Default channel it lands in
Paid search ads	google, bing	cpc	Paid Search
Paid social ads	facebook, instagram, linkedin, tiktok	cpc (or paid)	Paid Social
Paid video ads	youtube, vimeo, twitch	cpc (or a paid/video medium)	Paid Video
Display / banners	ad network name	display	Display
Shopping ads	google, bing	shopping	Paid Shopping
Organic social posts	facebook, instagram, linkedin, x	social	Organic Social
Email campaigns	newsletter, mailchimp, klaviyo	email	Email
AI assistants	chatgpt, perplexity	ai	AI
Affiliate partners	partner name	affiliate	Affiliates
SMS campaigns	your brand or tool name	sms	SMS
Push notifications	your tool name	push	Push
Podcast / audio	spotify, podcast name	audio medium	Audio

A few notes on using this table:

• Don't tag what classifies itself. Organic search (Organic Search), untagged referrals (Referral), and direct visits (Direct) are detected without UTMs. Tagging is for the links you place.
• The default rules are forgiving; your taxonomy shouldn't be. The Paid Search rule matches several paid mediums (cpc, ppc, paid, sem) precisely because teams are inconsistent. Pick one — we recommend cpc for all paid click inventory — so your Mediums tab stays clean even though the channel layer would have caught the synonyms.
• The full, canonical channel-by-channel matching table lives in the Sources report documentation. If you go beyond the defaults — Brand vs Non-Brand splits, per-partner affiliate channels — that's what custom Channel Grouping rules are for, with a safe draft → test → publish workflow.

The Errors That Silently Corrupt Attribution

Every one of these comes from a real-world pattern. Screen your links against this list.

Case and synonym drift

Covered above, but worth restating as the #1 offender: utm_source=Facebook vs facebook, fb vs facebook, Email vs email. Each variant is a separate row and a separate input to channel rules. The symptom is a Sources report where the same logical channel appears multiple times with split metrics.

Spaces and encoding surprises

utm_campaign=spring sale becomes spring%20sale or spring+sale depending on which tool encoded it — potentially both, in the same campaign, from different tools. Use your separator instead.

UTMs on internal links — never

UTM parameters describe how a visitor arrived at your site from outside. Putting them on internal links — a homepage banner tagged utm_source=homepage_banner, a menu item tagged utm_campaign=nav_test — is the most destructive mistake on this list: mid-visit, you're overwriting the acquisition signal with navigation trivia. The visitor who actually came from your expensive Google Ads campaign now shows up in reports attributed to "homepage_banner". Your acquisition data gets polluted by on-site behavior, and no report can un-mix them afterwards.

For measuring internal elements, use microconversions (sealmetrics.micro('cta_click')) or Content Grouping — that's what they're for. UTMs are for external links only. No exceptions.

Medium/source role confusion

utm_medium=facebook (a source in the medium slot), utm_source=cpc (a medium in the source slot). Both cause the channel rules to see nonsense — the Paid Social rule expects a social network in source and a paid medium in medium, so an inverted pair falls through to Unassigned. Apply the proper-noun/category test from above.

Redundant or overloaded parameters

Stuffing everything into one parameter (utm_campaign=facebook_cpc_black_friday_banner_a) wastes the structure that makes reports filterable. That string contains a source, a medium, a campaign, and a content variant — put each in its own parameter and the Sources report's drill-downs (source → campaigns → terms) work for free.

Tagging your own domain in emails-to-site handoffs

Watch out for tools that append their own tracking parameters (some affiliate networks and CMSes do). If a vendor's parameter carries real attribution value under a non-UTM name, don't rewrite your links — map it with UTM Mapping (see below).

The Governance Spreadsheet

The difference between having conventions and having governance is a living document. One shared spreadsheet, one owner, and a rule: no tagged link ships unless its campaign row exists in the sheet.

A battle-tested structure — one row per tagged link (or per link family):

Column	Example	Notes
Date created	2026-07-23
Owner	maria	Who to ask when the row is unclear
Landing URL	https://example.com/landing	Without parameters
utm_source	facebook	Dropdown from your canonical source list
utm_medium	cpc	Dropdown from your canonical medium list
utm_campaign	2026_black_friday_early_access	Free text, but validated against the naming pattern
utm_term	running_shoes	Paid search only
utm_content	banner_a	Only when A/B testing creatives
Final URL	(formula)	Concatenated automatically — humans never hand-assemble URLs
Status	live	draft / live / ended

Three practices make the sheet actually work:

1. Dropdown validation on source and medium. Free-text fields are where Facebook sneaks in. Data-validation dropdowns make the canonical list the only possible input — the spreadsheet becomes your enforcement mechanism, not just your record.
2. A URL-building formula. The Final URL column concatenates the landing URL and parameters (lowercasing on the way is cheap insurance). Marketers copy the finished URL; nobody types ?utm_ by hand.
3. A named owner. Someone approves new campaign names, prunes ended rows, and owns the canonical value lists. Governance without an owner decays in about one quarter.

Review the sheet against your live data monthly: open the Sources report, sort each tab, and look for rows that shouldn't exist — an unfamiliar source, a (not set) where a campaign should be, a near-duplicate value. Every anomaly traces back to either an untagged link (add it to the sheet) or a rule violation (fix the link, remind the human).

How Sealmetrics Processes Your UTMs

Understanding the pipeline turns the rules above from dogma into engineering. When a visit arrives, Sealmetrics processes attribution in this order:

URL arrives  →  UTMs extracted  →  Your UTM Mappings applied
             →  Channel rules evaluated (custom → default)
             →  Channel assigned to the visit

Stage 1: UTMs are read from the URL — every load, no cookies

The tracker reads the query string on each page load. There is no consent gate and no cookie dependency, which is why Sealmetrics attributes campaigns correctly even for visitors that cookie-based tools lose entirely. Your tagged URL is the attribution record.

Stage 2: UTM Mappings translate custom parameters

If your ESP, CRM, or ad platform emits its own parameter names — campaign_id, email_source, ad_medium — UTM Mapping lets you declare, per site, "treat campaign_id as utm_campaign". The mapping is applied at ingest, before classification, with well-defined semantics: explicit UTMs always win over mapped values, mappings copy (never move) the value, and empty parameters are a no-op. Mappings take effect in ~5 minutes and apply to future traffic only.

Governance implication: UTM Mapping is your integration escape hatch, not an excuse to skip conventions. The values still need to follow your taxonomy — mapping renames the parameter, it never rewrites Facebook into facebook. There's deliberately no value transformation at ingest.

Stage 3: Channel rules classify the visit

With final UTM values in hand, Sealmetrics evaluates channel grouping rules in priority order — your account's custom rules first, then the defaults — and the first match wins. This is where your source/medium discipline cashes out: rules are regex patterns over utm_source, utm_medium, and utm_campaign, so consistent values mean deterministic channels.

Two governance-relevant features of the channel settings:

• Test before you trust. The Test a visit tool in Channel Grouping settings lets you type any source/medium/campaign combination and see exactly which rule catches it — including draft rules that aren't live yet. Make it a habit: when you add a new taxonomy value, test it before the campaign launches, not after a week of Unassigned traffic.
• Drafts → test → publish. Custom rules are created as drafts and only affect real classification when a human publishes them, so experimenting is safe. Rules can also be bulk-managed via CSV import/export.

Where it all shows up

The Sources report presents the result across seven tabs — Channels (the default view), Sources, Mediums, Campaigns, Terms, Content, and Referrers — each with entrances, bounce rate, conversions, and revenue per row. Clean UTMs in, clean seven-way breakdown out.

Your UTM Governance Checklist

Condensed version to copy into your team wiki:

• All UTM values are lowercase, always
• One separator (_ or -), no spaces, no special characters
• One canonical name per source and medium — synonyms are outlawed in writing
• Mediums come from the default-channel-compatible taxonomy
• Campaign names follow a documented structure and are created in the spreadsheet first
• Zero UTMs on internal links — use microconversions or Content Grouping instead
• Links are generated by the spreadsheet formula, never hand-typed
• Custom vendor parameters are handled via UTM Mapping, values still follow the taxonomy
• New values are checked in the Channel Grouping "Test a visit" tool before launch
• Monthly audit of the Sources report against the spreadsheet, with a named owner

Related documentation

• Sources Report — the canonical default-channel table and all seven analysis tabs
• UTM Mapping — custom parameter names → standard UTMs at ingest
• Channel Grouping — custom channels, the rule tester, and the draft workflow
• How to Track Google Ads Campaigns — UTM templates for paid search
• How to Track Social Ads Campaigns — UTM tagging for Meta Ads
• How to Change UTM Parameters — the UTM Mappings UI walkthrough