Klaviyo GA4 Attribution Mismatch? Fix Source/Medium and Missing Flow Revenue

Klaviyo says the campaign worked. GA4 says email barely existed. Or campaigns appear in acquisition reports while flows disappear into Direct, Unassigned, or a weird split like klaviyo / campaing. That mismatch usually is not a philosophical attribution debate. It is a tagging and session-integrity problem you can audit: inconsistent utm_medium values, missing flow campaign parameters, encoded campaign names, or landing-page and checkout paths that stop preserving the original email session.

Why Klaviyo and GA4 disagree in the first place#

Klaviyo and GA4 are answering different questions. Klaviyo is optimized to explain email performance inside the send and flow system. GA4 is trying to preserve session and purchase context across landing pages, consent rules, redirects, and checkout. When those layers stay clean, the numbers still will not match perfectly. But when the gap is extreme such as email revenue in Klaviyo versus almost nothing in GA4, the first thing to audit is not the model. It is whether the original email session reached GA4 intact and stayed labeled the same way all the way to conversion.

Run the Klaviyo to GA4 mismatch audit in 6 steps#

You do not need to reverse-engineer the whole lifecycle stack on the first pass. The fastest route is to test one campaign link and one flow link, then compare how GA4 stores them. The goal is to prove exactly where the label changes.

1. 1

   Start in GA4 with Session source / medium

   Open Traffic acquisition and switch the primary dimension to Session source / medium. Look for email traffic split across values like klaviyo / email, email / email, klaviyo / flow, klaviyo / campaign, or (direct) / (none). If the rows are already fragmented, fix the labeling problem before you trust any top-line revenue comparison.

2. 2

   Compare one campaign link and one flow link side by side

   Send yourself a live campaign email and trigger one real flow email. Click both. Watch the final URL and confirm each keeps utm_source, utm_medium, and utm_campaign all the way to the landing page. If flows fail while campaigns succeed, you have narrowed the problem to the flow template or the flow-specific path.

3. 3

   Check whether utm_medium is stable

   GA4 can only classify cleanly if the medium is predictable. If one link says email, another says klaviyo-email, and a third says flow, the email lane fragments immediately. Align this with the governance rules in UTM naming convention for teams.

4. 4

   Inspect session campaign values for encoding drift

   Recent public Klaviyo threads showed GA4 storing campaign names as encoded variants like 4th$2520of$2520July.... If one campaign appears under two spellings, compare the sent URL, the landing-page URL, and the exact campaign token GA4 stores. One campaign split into two rows is still a taxonomy problem even if the click technically arrived.

5. 5

   Trace the post-click path through Shopify or your checkout stack

   If email sessions look tagged on the landing page but revenue still shifts to Direct, the leak may happen after the click. Audit the checkout or form path exactly as you would in Shopify direct traffic spike in GA4: consent timing, redirects, app scripts, and session persistence to purchase.

6. 6

   Export the raw rows before changing templates

   Pull source / medium and session campaign data into CSV before you change anything. That gives you the before-state, shows every split token at once, and makes it obvious whether the mismatch is mostly a Klaviyo setup issue or a broader GA4 hygiene problem like GA4 UTM not working.

The 5 most common root causes in Klaviyo + GA4#

1) Flows often break where campaigns do not

This is the exact pain pattern repeated in current Klaviyo-community evidence: campaigns pull through to GA4, but flows stay missing. That usually means the email channel is not broken in general. One lane is tagged cleanly and the other is not. Treat that as good news. It narrows the fix to flow-specific URL construction, a missing campaign token, or a post-click path unique to flow destinations.

2) Medium drift makes one email lane look like three

Teams often overcomplicate this by inventing a new medium for every message type. email, flow, newsletter, and sms-email might all feel descriptive, but they force GA4 to treat one lifecycle channel as several acquisition lanes. Put the detail somewhere safer, such as utm_campaign or utm_content, and keep the medium policy stable. If you need channel-level granularity for SMS, create an intentional rule for it instead of letting templates invent their own labels.

3) Encoded campaign names create fake duplicates

Several same-day Klaviyo threads describe campaign names mutating into encoded strings inside GA4. Even when revenue is real, that formatting drift makes reporting tedious and can hide performance under multiple rows. The practical fix is simple: keep utm_campaign lowercase, avoid spaces, and use one repeatable token format. UTM campaign naming structure gives you a structure that survives handoffs without becoming unreadable.

What a clean Klaviyo attribution policy should look like#

You are not trying to make Klaviyo and GA4 identical. You are trying to remove preventable drift so the remaining difference is explainable. A clean policy is boring: one source convention, one medium convention, campaign names that do not fork, and a QA habit for both campaigns and flows. Once that is in place, you can have a real attribution conversation instead of a labeling conversation disguised as an attribution one.

Audit the export before you rewrite every template#

A manual spot check catches one broken link. An export audit catches the whole pattern: every medium variant, every encoded campaign token, and every row that fell into Direct or Unassigned after a Klaviyo send. That is why UTM Drift Guard starts with the CSV and grades the mess before it suggests a canonical taxonomy. The Claude API clusters spelling, case, and token variants so you can review the problem as a system, not as one email at a time. Nothing writes back to Klaviyo or GA4 automatically.

For lifecycle teams, that matters because the same mismatch rarely stays isolated to one campaign. The exact same drift usually appears in flows, promo blasts, SMS-to-email hybrids, and ecommerce handoffs. Once the export is clean, outreach and reporting get easier because the proof asset is specific: you can show exactly which source / medium and campaign tokens are poisoning the report.