HubSpot GA4 Attribution Mismatch? Fix Source/Medium Drift First

HubSpot says the email, paid social, or landing page did its job. GA4 says the same traffic arrived as Direct, Unassigned, or three different source / medium rows. That is the classic HubSpot GA4 attribution mismatch pattern. In most accounts, the first problem is not the attribution model. It is upstream labeling drift: one campaign tagged five ways, a landing page that drops UTMs, or a HubSpot handoff that makes one acquisition path look like several.

Why HubSpot and GA4 disagree so often#

HubSpot and GA4 are not trying to answer the same exact question. HubSpot cares about campaign influence, lifecycle progression, and CRM-linked outcomes. GA4 cares about sessions, channels, and on-site behavior. Some difference is normal. The real problem is when the difference becomes too big to explain with model logic alone. If HubSpot says email created demand but GA4 spreads that same traffic across hubspot / email, email / email, direct / (none), and Unassigned, you are not looking at a subtle modeling disagreement. You are looking at a classification problem.

That matters because GA4 is case-sensitive, so email, Email, and EMAIL are different mediums. On top of that, 10–20% of GA4 sessions commonly land in Unassigned, and as much as 26% of conversions can be credited to the wrong channel when UTM data is fragmented. Once the account has that kind of drift, every HubSpot-to-GA4 comparison becomes noisier than it should be.

Run the HubSpot GA4 mismatch audit in 6 steps#

1. 1

   1) Export raw source / medium and campaign rows from GA4

   Start with the raw acquisition rows, not a dashboard screenshot. Look for the same traffic lane split across values like hubspot, HubSpot, newsletter, email, hs_email, or (direct) / (none). If the export is already fragmented, the mismatch is visible before you touch any campaign settings.

2. 2

   2) Compare one HubSpot email link and one landing-page CTA link

   Open the exact live URLs. Confirm utm_source, utm_medium, and utm_campaign are present on the final landing-page URL, not only in the builder. If the final URL is clean for one path but not the other, you have narrowed the break to a specific template or redirect chain.

3. 3

   3) Standardize utm_medium before you do anything else

   A team often invents descriptive mediums like newsletter, workflow, hs-email, or lead-nurture. They feel useful, but they split one lifecycle lane into several GA4 rows. Use one stable medium policy and move extra detail into utm_campaign or utm_content. UTM naming convention for teams explains the governance side.

4. 4

   4) Check whether campaign names are drifting by case, spaces, or encoding

   A campaign should not appear once as q3_demo_push, once as Q3 Demo Push, and once as q3-demo-push. That is not three campaigns. It is one campaign reported three ways. This is the same pattern behind many GA4 UTM not working complaints.

5. 5

   5) Trace the conversion path after the first page view

   If the landing page keeps the UTMs but the final conversion still lands in Direct or Unassigned, the leak is later in the journey. Check consent timing, redirects, embedded forms, booking links, subdomains, and checkout or meeting-booking handoffs. The original session may be arriving cleanly and then getting reset before the conversion is recorded.

6. 6

   6) Grade the drift before you rewrite templates

   Do not make changes one link at a time without preserving the before-state. Grade the export first, cluster the biggest duplicates, and rank them by traffic impact. That gives you a clear cleanup queue and a proof asset you can share with operators, leadership, or clients.

What to score inside a HubSpot source/medium audit#

The fastest agency or RevOps-friendly audit is a source/medium audit, not a giant attribution workshop. You are checking whether acquisition labels are stable enough for leadership decisions. If one email program appears under four mediums and two sources, the reporting layer is already unstable even if every dashboard still loads.

A mini example: one HubSpot campaign, four different realities#

This is the pattern agencies and in-house ops teams keep running into. One team launches an email, a paid-social retargeting push, and a landing page using slightly different naming habits. HubSpot can still show campaign influence because the activity exists in the CRM. GA4, meanwhile, receives several versions of what should be one clean path. The mismatch feels mysterious until you line up the actual query strings.

The 5 most common root causes behind HubSpot GA4 mismatch#

• The same lifecycle or paid campaign uses multiple utm_medium values across HubSpot assets.
• Case-sensitive source and medium drift splits one channel into several GA4 rows.
• Campaign names change format between email, landing page, and ad handoff.
• Redirects, embedded forms, meeting links, or subdomains strip or reset the original UTMs.
• Teams compare modeled influence in HubSpot against unstable raw acquisition labels in GA4 and mistake that instability for a pure attribution-model issue.

Notice that only the last item is really about models. The first four are data-hygiene problems. That is why a quick audit before a QBR or a one-account pre-audit scorecard often closes the gap faster than another cross-functional reporting meeting. Clean the input labels first. Then compare models on top of a dataset you can actually trust.

How agencies and RevOps teams should use this in practice#

If you sell cleanup, do not position this as a vague analytics health check. Position it as a short proof object: export the rows, grade the drift, show the top clusters, and recommend the next cleanup step. That works especially well for HubSpot-heavy teams because the buyer already feels the commercial pain. The CRM says one story. GA4 says another. A source/medium audit gives them a concrete place to start.