What a UTM tracking code is
UTM stands for Urchin Tracking Module — the analytics company Google bought in 2005 — and the tags are just extra text appended to a URL after a question mark. When a shopper clicks a tagged link, Google Analytics strips those tags off and records them against the session.
There are five parameters. Three are effectively required for clean reporting, two are optional:
- utm_source — where the click came from (facebook, google, klaviyo).
- utm_medium — the type of traffic (cpc, email, social).
- utm_campaign — the specific push (bfcm-2026, spring-restock).
- utm_content — which ad or link variant (video-a, header-link).
- utm_term — the paid keyword, mostly for search.
So a single Black Friday email link might carry utm_source=klaviyo, utm_medium=email, utm_campaign=bfcm-2026. That is the entire mechanism. No pixel, no script change — just disciplined text on the end of a link.
How Google Analytics actually reads your UTMs
This is where most how-to articles stop and where operators get burned. GA4 doesn't just store your tags — it runs them through a fixed set of rules to assign each session to a default channel group, and those rules are Google's, not yours. You can add custom groupings, but you cannot override the defaults (the GA4 default channel rules are documented here).
The rules are exact and case-sensitive. A session tagged utm_medium=cpc from a recognized source lands in Paid Search; utm_medium=paid may not, because "paid" isn't a value the rule recognizes. utm_source=Facebook and utm_source=facebook become two different sources in your reports (case drift is the classic self-inflicted reporting break). Anything that matches no rule falls into catch-alls — "Referral" or "Unassigned" — which is where an affiliate link missing utm_medium=affiliate or an untagged influencer link quietly goes to die.
One more thing worth knowing before you obsess over model settings: GA4 removed first-click, linear, time-decay, and position-based attribution models in November 2023, leaving data-driven (the default), paid-and-organic last click, and Google-paid-channels last click (per Google Analytics Help). You literally cannot pick the older models anymore. And data-driven attribution needs volume to be stable — a commonly cited practitioner threshold is around three hundred-plus conversions a month before its fractional credit splits settle down (ReferralCandy's 2026 attribution guide). Below that, the model wobbles with noise no matter how clean your tags are.
The naming rules that keep reports usable
UTMs are the only attribution input you fully control, which also makes sloppy tagging the most common wound store owners inflict on themselves. A finite, documented vocabulary is the whole game. The near-universal conventions:
- All-lowercase, always. The single highest-leverage rule; case drift fragments every downstream report (utm.io's best-practice list).
- Hyphens or underscores, never spaces. Spaces turn into plus signs or %20 and break consistency.
- A fixed source and medium list. facebook, not fb. Draw utm_medium from the GA4-recognized set — cpc, email, social, affiliate, referral, display — so channel grouping works by construction.
- Never tag internal links. Putting UTMs on your own site navigation or in-site banners restarts the session's source mid-journey and overwrites the real acquisition source. This is the classic attribution reset that turns a paid order into a "Direct" one.
Keep one shared sheet or a builder that everyone making links uses. If you want a faster path than a spreadsheet, our UTM tracking code builder walks the structure field by field. For the full picture of how these tags fit alongside pixels and server-side order data, the attribution and tracking guide is the hub to start from.
Why UTMs matter more now, not less
A common belief is that Apple's privacy changes killed link tracking. It's backwards. Since iOS 17 (September 2023), Apple's Link Tracking Protection strips user-level click IDs — fbclid and gclid — from links opened in Apple Mail, Messages, and Safari Private Browsing, but UTM parameters are left intact because they identify campaigns, not individuals (Klaviyo's breakdown of Link Tracking Protection).
That matters because of how much of your traffic is Apple traffic. Safari runs roughly fifty-five percent of US mobile sessions (per StatCounter data summarized by Backlinko), and ATT opt-out means the industry-average opt-in for cross-app tracking had climbed only to about thirty-five percent by the second quarter of 2025 (according to Adjust). Click IDs are the fragile layer. Manual UTMs are the durable one — the redundancy that survives when the platform's own tracking gets clipped. Your Facebook pixel setup handles the pixel side; UTMs are the half you control by hand.
What a UTM tells you — and the number it can't
Here's the worked example that the tutorial pages skip. Say you run a store doing 340 orders a month at a $31 average order value, with $2,800 a month in Meta spend. You tag your Black Friday prospecting campaign, and GA4 reports it drove 40 sessions that converted. Great — you now know the source.
What the UTM cannot tell you is whether those 40 orders made money. Walk the real per-order math. Your shirt sells for $31. Printify charges you $12 for the blank and print, shipping runs $4, and payment fees take about $1.20. That's $31 − $12 − $4 − $1.20 = $13.80 of gross profit per order. Forty orders is 40 × $13.80 = $552 in gross profit. If that tagged campaign spent $600 to get them, it lost money — even though GA4 proudly credits it with 40 conversions.
The UTM answered "where did they come from?" It had nothing to say about product cost, supplier fees, or whether the campaign cleared its own ad spend. That cost side is exactly what lives outside Google Analytics, split across Shopify, your print supplier, and your ad account.
Where Victor fits
This is the gap PodVector AI's Victor is built to close. Victor is an AI employee who connects to Shopify, Meta Ads, Google Ads, Printify, Printful, Gelato, and Klaviyo, so he sees every order, every ad dollar, and every supplier cost in one place — and computes the true per-order profit a UTM report can never show you. Victor knows the product cost from Printify before the first sale, reads the Meta and Google spend behind the campaign, and tells you which push actually cleared.
He doesn't just report it, either. Victor pauses the losing ad set, shifts budget to the campaign that's actually profitable, and builds the Klaviyo email flow with properly tagged links so your email traffic stops landing in "Unassigned." When you're ready to stop guessing which campaigns earn their spend, start with Victor.
For the email side specifically — where UTM discipline breaks most often — see how to get it right in Klaviyo UTM tracking. And if you're weighing whether a dedicated attribution tool belongs in your stack, marketing attribution tools and ROI tracking compares the options.
FAQs
Where do I see UTM data in Google Analytics?
In GA4, your tagged traffic shows up under the Acquisition reports, broken out by the source and medium you set, and rolled into the channel group GA4 assigned. Switch the primary dimension to "Session source / medium" or "Session campaign" to see your exact tags rather than the grouped channel. If a campaign you tagged isn't showing, the usual culprit is a typo or a case mismatch in the tag.
Does case really matter in a UTM code?
Yes, and it's the most common mistake. utm_source=Facebook and utm_source=facebook are two separate rows in GA4, and a medium GA4 doesn't recognize can drop the whole session into "Unassigned." Standardize on all-lowercase for every tag, every time, and don't let two people tag the same campaign with different capitalization.
Should I put UTM codes on my paid ads if Meta already adds fbclid?
Yes. Meta and Google auto-append their own click IDs, but Apple's Link Tracking Protection strips fbclid and gclid in Mail, Messages, and Safari Private Browsing, while leaving UTMs untouched (Klaviyo). Manual UTMs are your backup when the platform's own tracking gets clipped — which, for a US store with heavy Safari traffic, happens constantly.
Can a UTM code tell me if a campaign was profitable?
No. A UTM records the traffic source and the conversion, but it has no view of your product cost, supplier fees, shipping, or payment processing — so it can't tell you the profit on an order. That calculation lives across Shopify, your print supplier, and your ad account. Victor connects those, computes the true per-order profit, and tells you which campaign actually cleared its spend.
Why is so much of my traffic showing as "Direct" or "Unassigned"?
Rising Direct is usually lost attribution, not people typing your URL — stripped click IDs, expired Safari cookies, or untagged links. "Unassigned" is almost always a tagging problem: a medium GA4 doesn't recognize, or a link with no UTMs at all. Audit your tags before you assume the data is broken; most mystery traffic is a vocabulary fix, not a measurement mystery.
Do I need to tag links between pages on my own site?
No — and you should not. Internal UTMs restart the session's source mid-visit and overwrite the real channel that brought the shopper in, turning a paid or email order into a "Direct" one. Only tag links that bring people to your site from outside it: ads, emails, social posts, and partner links.