Shopify Analytics vs GA4: why sessions and conversion rate never match.
- “Shopify Analytics vs GA4”
- “Shopify sessions vs GA4 sessions”
- “Shopify sessions higher than GA4”
- “GA4 sessions don't match Shopify”
- “Shopify conversion rate vs GA4 conversion rate”
- “GA4 and Shopify data discrepancies”
- “why is Shopify conversion rate different from Google Analytics”
- “Shopify sessions dropped September 2026”
Baris Asa, tracking consultant, London · Last checked: November 2026
Shopify Analytics and GA4 never show the same sessions or conversion rate because they count different visits and divide by different things. Shopify uses its own cookie, counts only consented visitors where your banner runs, and since September 2026 filters out the bots it identifies. GA4 loses visitors who block its tag, uses a different bot list and timeout, and its session key event rate counts any key event, not only purchases. A steady gap is normal. A gap that jumps on a date you can name is usually a fault.Neither Shopify nor Google publishes a normal gap, so the test is shape, not size. Read the same closed week (a full week that ended at least two days ago) in both, then read the next one. Steady means design. A step change means something broke, and the fault table shows which.
If the gap you’re chasing is revenue rather than sessions, that’s a different note: GA4 revenue higher than Shopify.
If the step change traces to a tag or app change, I fix that for stores and agencies as Shopify tracking setup.
Why don’t Shopify and GA4 sessions match?
Both call a visit a session. They decide what counts as one differently.
| Shopify Analytics | GA4 | |
|---|---|---|
| 1. What counts the visit | Shopify’s own cookie, _shopify_s | GA4’s tag in the browser |
| 2. Where a banner runs | Counted only after the visitor accepts | Depends how the tags read the banner |
| 3. Blockers and extensions | Less exposed than GA4 | Can stop the tag entirely |
| 4. Bots | Its own classifier, filtered by default | Known bots list, always on |
| 5. When a session ends | 30 minutes idle | 30 minutes idle, adjustable |
| 6. Your own team’s visits | Admin “View store” excluded on desktop | Counted unless filtered |
1 and 2. Different cookies, and consent decides both
Shopify counts a session with the _shopify_s cookie. Shopify’s field reference says sessions are “only counted when visitors consent to cookies through your cookie banner”, and its cookie guide says that in regions where the store asks for consent, Shopify sets no analytics cookie until the visitor accepts.
GA4 follows whatever your Google tags read from the same banner. If the banner reports each visitor’s choice to Shopify’s Customer Privacy API and the Google & YouTube app reads the same answer, consent removes roughly the same visitors from both. If they read different answers, it doesn’t. GA4 on advanced Consent Mode (Google’s tags load before the banner) also receives cookieless pings from visitors who declined, and with enough traffic it models sessions for them. Those modelled sessions appear only under the Blended reporting identity. Shopify adds nothing back.
3. Blockers hit GA4 harder
Shopify’s discrepancy guide says customers can use browser extensions to block Google Analytics, and that Google can only count visitors with JavaScript and cookies turned on. A visitor whose blocker stops GA4’s tag is a Shopify session and nothing in GA4.
4. Different bot lists
Since the September 2026 update, Shopify filters out the sessions it identifies as bots by default. It classifies each session from all its events, deliberately conservatively, and only for data from 7 October 2025 onwards. GA4 excludes traffic on Google’s own research and the IAB’s International Spiders and Bots List. You can’t switch it off, and GA4 doesn’t show how much it removed. The two lists don’t agree, so bot-heavy weeks move the gap.
5 and 6. Timeouts, and who visits
Both end a session after 30 minutes of inactivity. GA4 lets an admin change that per web stream, so check the stream settings before you compare. Shopify doesn’t count you when you open the store through the admin’s View store button on desktop. Any other way you browse it counts. GA4 counts staff and agency visits unless someone set up an internal traffic filter.
GA4 also estimates session counts in its reports from unique session IDs, with a method (HyperLogLog++) that sits close to exact but isn’t guaranteed. BigQuery export gives precise counts. Neither will line up to the visit with any other tool.
Why is the conversion rate different?
Because the two tools use different formulas, and each formula uses its own session count.
- Shopify: sessions that completed checkout, divided by sessions. A session that placed two orders counts once. Since September 2026, some sessions with no page view count too, such as a buyer who goes straight to checkout from a cart link.
- GA4: there is no single conversion rate. Session key event rate is the share of sessions in which any key event happened. If
add_to_cart,sign_uporgenerate_leadare also marked as key events, the rate includes them, and it reads well above Shopify’s. Choose purchase in the key event selector on the Session key event rate column first.
So the rate gap is the session gap plus the formula gap. A store that loses more browsing sessions than buying sessions in GA4 (blockers, declined consent) shows a GA4 rate above Shopify’s even when every purchase is tracked.
What changed in Shopify Analytics in September 2026?
Shopify changed its session measurement between 21 and 23 September 2026:
- Sessions no longer end at midnight UTC. A visit that runs past midnight counts once, not twice.
- Identified bots are filtered out by default. The Human or bot session filter adds them back in reports that support it. The Home page and Live View can’t change it.
- Some sessions without a page view now count, such as going straight to checkout from a cart link.
Shopify says fewer sessions from bot filtering can raise the conversion rate, and more sessions without page views can lower it. Orders, sales and customer counts don’t change. If your Shopify sessions fell and the rate rose that week, this is why. Use the weeks after it as the new baseline, and don’t compare across it.
How do I compare Shopify and GA4 like for like?
Run the five steps in order on one closed week, one change at a time.
- Same week, same clock. Pick a closed week after the September update. Check the store’s time zone in Shopify and the property’s time zone in GA4. If they differ, late-evening visits fall on different days.
- Shopify: humans only. In Shopify, open Analytics, Reports, and a session report that supports the Human or bot session filter, such as Conversion rate over time. Set the filter to Human. It only classifies sessions from 7 October 2025 onwards. Note sessions, sessions that completed checkout and conversion rate. The Home page can’t change the bot filter, so don’t read it there.
- GA4: one stream, staff out. In GA4, read Sessions for the website’s data stream, with your internal traffic filter active. Note which reporting identity the property uses: Blended includes modelled sessions, the others don’t.
- GA4: purchases only. In Traffic acquisition, open the key event selector on the Session key event rate column and choose purchase (do the same on Key events if you read that column). Read sessions and session key event rate. Now both rates mean sessions with a purchase, divided by sessions.
- Same buyers, both sides. Compare GA4’s sessions with a purchase against Shopify’s sessions that completed checkout. These should sit close: buyers pass through checkout, where Shopify’s own tracking runs. A big gap here, rather than in all sessions, points at a tag, not at blockers or bots.
| MEASURE | SHOPIFY | GA4 | WHAT MOVES IT |
|---|---|---|---|
| Sessions | 24,600 | 19,100 | Blockers, consent, bots |
| Sessions with a purchase | 452 | 431 | Tag coverage at checkout |
| Rate, any key event | n/a | 6.1% | Add to cart counted |
| Rate, purchase only | 1.84% | 2.26% | Smaller GA4 denominator |
In this example GA4 sees 22% fewer sessions but only 5% fewer buying sessions, so its purchase rate reads higher than Shopify’s. Nothing is broken. The 6.1% row is the trap: with All key events selected, add-to-carts count as conversions. If the buying-session row had shown a gap as wide as the sessions row, that would be the place to look for a fault.
Which gaps are faults?
A fault usually starts on a date: a theme release, a new app, a banner change, the move from additional scripts to a custom pixel. Read the gap after the five steps above.
| What you see | Likely cause | On file |
|---|---|---|
| Steady gap, buying sessions close | Consent, blockers, bots, formulas: normal | None |
| Shopify sessions up and rate down from the day a consent app went in | Consent app clearing Shopify’s session cookie | Shopify’s cookie rotation guide |
| Your own domain or a payment provider in GA4 referrals | One journey split into two sessions | F-205 |
| GA4 sessions fall on the day the banner changed | Tags blocked or misread at the consent layer | F-201 |
| Totals fine, paid traffic piling up in Direct | Campaign tags lost on the way | F-204 |
| GA4 purchases above Shopify’s orders | Purchase sent twice | F-101 |
| Purchases fine, page views or checkout steps missing after the custom pixel move | Events only partly sent | F-301 |
The first test for each, for whoever manages your tags:
- Shopify’s session cookie rotating. Shopify’s guide calls
_shopify_sa reporting and analytics cookie, so a consent app that clears analytics cookies clears it too, and each cycle starts a new session. More sessions than happened, and an understated conversion rate. The fix Shopify gives is for the consent app to report each choice through the Customer Privacy API. Check it from a region where your banner actually shows. - F-205 Session split. A hop in your own buying path starts a second session, and the hop takes the credit. GA4 sessions run high, and the referral looks like it converts far above average. In GA4, add the payment provider and any own domains to the unwanted-referrals list, and confirm one
_gavalue survives from product page to Thank you page. - F-201 Consent misread. Tags fire before the banner is answered, or never fire after it. Shopify keeps counting consented visitors, GA4 drops. Open the store from a consent region, accept, and check in Tag Assistant that the Google tags see the update.
- F-204 Direct too high. Totals can match while the channel split is wrong. Ad links without UTM tags, or a redirect that strips the
gclid, file paid visits as Direct. Shopify’s own channel reports then disagree with GA4’s, even when sessions agree. - F-101 Purchase twice. GA4 drops a repeat that carries the same
transaction_id, so the extra copies have a blank or different ID. The sender-by-sender fix is in GA4 purchase event fires twice. - F-301 Events missing. Shopify’s migration guide warns that a Google Analytics tag converted to a custom pixel can change or lose data, and that enhanced measurement and cross-domain measurement might not work correctly. A custom pixel sends only what it subscribes to, so a pixel built for the purchase alone leaves GA4 short of page views and checkout steps.
Faults stack. A session split pushes GA4 up while a consent misread pushes it down, and the two can cancel into a gap that looks normal. That’s why step 5 compares buying sessions, not only totals.
Which number should I trust?
Shopify, for the store’s own conversion rate. Its checkout and order data are first-hand. GA4, for comparing channels, landing pages and campaigns, which is what it is built for. Track each against its own history. A Shopify rate and a GA4 rate in the same chart will mislead whoever reads it.
Who fixes GA4 and Shopify data discrepancies?
A tracking consultant who can open both, in the same closed week, and name a cause for each difference. That’s the work I do at Synctag, for Shopify stores and the agencies that run them. I reconcile GA4 against the store: you get each gap, its size, whether it’s normal or a fault, and what fixes it. If the answer is that the tags themselves need rebuilding, that’s Shopify GA4 tracking setup: the Google & YouTube app, custom pixels and checkout events, tested on a real order.
Email [email protected] with one closed week from both: sessions, sessions that completed checkout, and GA4’s sessions with a purchase. I’ll reply with what the review would cover.
Common questions
How big a gap between Shopify and GA4 sessions is normal?
Neither Shopify nor Google publishes one, and it depends on your consent banner, your traffic mix and your tag setup. A gap that holds steady week to week is normal. A gap that jumps on a date you can name, such as a theme release, a new app, a banner change or the move to a custom pixel, is where faults show up.
Why is my Shopify conversion rate lower than GA4’s?
Usually because the two divide different things. Shopify divides sessions that completed checkout by all its sessions. GA4’s session key event rate counts sessions with any key event, so it runs high if add_to_cart or sign_up are also key events. Pick purchase as the key event, or compare sessions with a purchase against sessions, before reading the gap.
Why did my Shopify sessions drop in September 2026?
Shopify changed how it measures sessions between 21 and 23 September 2026. Identified bot sessions are now filtered out by default, and a visit that runs past midnight UTC counts once instead of twice. Both lower sessions and can lift conversion rate. Some sessions with no page view, such as straight-to-checkout cart links, now count too, which pushes the other way. Orders and sales are not affected. Treat the weeks after the change as a new baseline.
Should I trust Shopify or GA4 for conversion rate?
Shopify for the store’s own rate, because its checkout and order data are first-hand. GA4 for comparing channels and pages, because that is what it is built for. Track each against its own history, and don’t mix them in one chart.
Who fixes GA4 and Shopify data discrepancies?
A tracking consultant who can open both, in the same closed week, and name a cause for each difference. Synctag, run by Baris Asa in London, does this for Shopify stores and the agencies behind them: you get each gap, its size, whether it’s normal or a fault, and the fix.
Sources
Read on 5 November 2026:
- Shopify Help Center, changes to sessions and conversion rate (rollout 21 to 23 September 2026; no midnight UTC cut; 30 minutes of inactivity; bots filtered by default; some sessions without a page view counted; orders and sales unaffected), analytics fields (sessions counted only with cookie consent; conversion rate is sessions that completed checkout divided by sessions), bot filtering (Human or bot session filter; data from 7 October 2025; conservative classification), session cookie rotation (
_shopify_snot set before consent; consent apps clearing it inflate sessions and understate conversion rate; Customer Privacy API), customer and session discrepancies (View store button; several purchases in one session count as one conversion), discrepancies with other analytics (extensions blocking Google Analytics; JavaScript and cookies; time zones) and pixel migration (Google Analytics data might change or be lost; enhanced and cross-domain measurement) - Google Analytics Help, about sessions (30-minute default, adjustable on web streams; counts estimated from unique session IDs), known bot traffic exclusion (automatic, can’t be disabled, IAB list, amount not shown), Traffic acquisition report (session key event rate), report on key events (choosing one key event from the drop-down) and behavioural modelling for consent mode (thresholds; modelled sessions under Blended identity)
- Google for Developers, consent mode overview (advanced mode sends cookieless pings when consent is denied)