The opportunity
Shopify and Google Analytics 4 answer different questions. Shopify records commerce outcomes under the merchant's order definition; GA4 records sessions, events, acquisition fields, and modeled behavior under its measurement settings. A difference between their revenue totals is not automatically lost tracking or poor channel performance.
BizSidekick aligns the reporting window and revenue definition, keeps unmatched records visible, and classifies the best-supported explanations. The result is a tracking plan, not an invented single source of truth.
Keep the source definitions visible
| Source | Evidence used | Boundary |
|---|---|---|
| Shopify | Paid-order time, landing page, referrer, UTM fields, discounts, refunds, and net sales | Commerce record under the selected order definition |
| GA4 | Session source, medium, campaign, landing page, purchase events, and revenue | Measurement record subject to consent, configuration, and attribution settings |
| Task rules | Complete dates, store time zone, currency, refund treatment | Makes the comparison repeatable |
Privacy-safe matching may leave orders or sessions unmatched. Those records remain an explicit evidence class instead of being assigned to a channel by guesswork.
Classify before fixing
The review separates:
- expected timing or attribution-model differences;
- consent-related measurement loss;
- missing or inconsistent campaign parameters;
- payment-domain or cross-domain return issues;
- event implementation problems;
- evidence that remains insufficient.
This classification prevents a measurement problem from being mistaken for a budget problem.
How it works
- Align the comparison — Use complete dates, one time zone, one currency, and explicit Shopify and GA4 revenue definitions.
- Read both evidence sets — Keep source, medium, campaign, landing page, purchase, order, refund, and net-sales fields reviewable.
- Build comparable cohorts — Match only on available privacy-safe evidence and retain all unmatched rows.
- Rank explanations — Prioritize causes by the amount of affected evidence and the confidence of the diagnosis.
- Prepare the tracking plan — Fix the highest-confidence measurement issue first, then repeat the same cohort after a stable observation window.
Safety boundary
The example values are illustrative and no analytics setting is changed. The Task does not claim perfect attribution, assign identities without evidence, or recommend a channel budget change before measurement issues are investigated.
Expected outcome
An evidence-ranked tracking plan before channel budgets change
Try this prompt
Paste it into Claude to start this use case.
@BizSidekick Compare Shopify paid orders with GA4 purchase sessions for the last 30 complete days. Show revenue classified as Direct or Unassigned, explain likely causes, and list tracking checks in priority order.
