UTM for internal links: the short answer is no
Should you use utm for internal links, the links that go from one page of your own site to another? No. Almost never. UTM parameters exist to label traffic arriving at your site from somewhere else, a Reddit post, a newsletter, an ad, each tagged with a source and medium. Put them on your own internal navigation, your header menu or a pricing-page button, and you are not tracking a new visitor. You are tampering with the record of the one you already have.
This trips up a lot of people because tagging an internal link feels like more tracking, and more tracking sounds like more data. It is the opposite. It quietly corrupts data you already had.
What actually breaks, and it is not what most guides tell you
Here is where almost every article on this gets the details wrong, so it is worth being precise.
In the current Google Analytics, GA4, tagging an internal link does not immediately hijack the visit in progress. As the GA4 specialists at Analytics Mania put it: "The current session remains unaffected. However, if the next session is direct, the UTM value from the previous session will be inherited." So the real damage is delayed. The utm_source you stuck on your own menu link can get inherited by that visitor's *next* visit, one that may have had nothing to do with any campaign, silently relabelling it.
You will still read, all over the internet, that tagging an internal link "starts a new session and reattributes the visit immediately." That was true, but in the old Universal Analytics, the version Google retired in 2023. In UA, a change in campaign values mid-visit did trigger a fresh session on the spot. GA4 does not work that way. If you are running a current GA4 property, the session-restart warning is describing a tool that no longer exists, and the inheritance problem is the one you actually have.
Either way, the conclusion is identical. Both versions punish tagging your own links. Only the mechanism differs.
The SEO angle, sized honestly
There is a second reason you will see cited: every internal link carrying UTMs is a different URL pointing at the same page. yoursaas.com/pricing and yoursaas.com/pricing?utm_source=header are two addresses for one page, and search engines have to crawl and reconcile the duplicates. A canonical tag helps, but it is a hint to Google, not a command, and crawlers still have to visit each variant to find it.
Here is the honest part most SEO guides skip: this problem mostly matters at scale. If you run an enterprise site with a hundred thousand pages, parameter variants genuinely eat crawl budget. If you run a twenty-page SaaS site, Google will crawl the lot regardless and you will never feel it. So do not let anyone scare you about crawl budget you do not have. On a small site, the attribution mess above is the real reason to avoid internal UTMs, not the SEO one.
So how do I track how people move around my site?
This is the useful question underneath, and it has a clean answer once you see the split.
A UTM answers one thing: where did this visit come from. That is a question about the outside world, about the boundary where someone crosses onto your site. How a visitor moves *once they are already inside*, which pages, which buttons, in what order, is a completely different question, and UTMs are simply the wrong tool for it. Bolting a UTM onto an internal link to answer it is like putting a "welcome" mat between two rooms of the same house.
The right tools for on-site movement are analytics events (a recorded click or pageview that does not touch attribution) or a proper journey view. So when you are creating utm links, keep the rule to one line: only tag links that live somewhere other than your own site. Everything internal gets tracked a different way.
This is exactly the on-site question a UTM cannot answer and ScoutVibe is built for: not just which link someone arrived from, but the actual path they took through your site afterward, per person.
Movement inside your site is the question UTMs cannot answer. ScoutVibe shows you the actual path each visitor took after they landed, per person, not just which channel they arrived from. The free tier covers your next launch.
Try ScoutVibeThe one exception people ask about
"But I want to A/B test two buttons on my own homepage. Different UTMs would tell them apart, right?"
They would, and they would also corrupt your attribution exactly as described above, because GA4 cannot tell a testing UTM on an internal link from a real one. The better answer is an event: fire a distinct event, or one event with a label, for each button. You get the click comparison you wanted, and none of the attribution damage. So even here, the answer is still no UTMs on internal links.
FAQ
Should I put UTM tags on links between my own pages?
No. UTM parameters are for labelling traffic that arrives at your site from an external source like a Reddit post or a newsletter. On an internal link, they do not track a new visitor, they corrupt the attribution of the one you already have. The rule is simple: only tag links that point to your site from somewhere else.
What actually breaks if I do?
In GA4, the current visit is not immediately hijacked, but the stale UTM value can be inherited by that person's next visit, mislabelling a session that had nothing to do with any campaign. Older guides say it "starts a new session" instead, but that describes Universal Analytics, which Google retired in 2023. On a small site the SEO concern about duplicate URLs is minor; the attribution problem is the one that matters.
If not UTMs, how do I track how people move around my site?
With analytics events or a journey view, not UTMs. A UTM answers where a visit came from, a question about the boundary of your site. Movement between your own pages is a different question, and the right tools for it are events that record a click or pageview without touching attribution, or a per-person path view of what each visitor actually did after landing.