TL;DR
Connect Shopify to turn store data into governed profitability, inventory, and revenue-reconciliation decisions—without treating estimates as fact.
The ecommerce question is not “what sold?”
Most Shopify reports can tell a merchant what sold. The harder, more useful questions are whether the sale contributed margin, whether the underlying cost data is complete, whether inventory will cover likely demand, and why a marketing platform’s revenue does not match the store. Those questions cross systems and contain uncertainty. A dashboard that hides that uncertainty is not a decision system; it is a prettier spreadsheet.
nqzai’s Shopify connector is designed around a simple operating rule: Shopify is transaction truth; acquisition tools describe behaviour and attribution. The connector brings store facts into nqzai, keeps their freshness visible, and lets merchants ask focused questions without asking an LLM to invent a number or make a store change.
This is the seed article for nqzai’s Ecommerce Intelligence series. It explains the connector’s current read-only capability, the evidence behind each answer, and the guardrails that keep “AI commerce” from becoming confident guesswork.
What the Shopify connector does today
After a merchant connects a Shopify store, nqzai can build a store-level view of catalog, variants, inventory, orders, discounts, refunds, and recorded product costs. The first value is deliberately operational rather than promotional: confirm that the connection and data are healthy before drawing conclusions.
Today’s capability has five practical jobs:
- Connection and sync status: show whether the store is connected, which data is available, when catalog and order data last synced, and whether the connector is healthy.
- Store data health: identify missing product costs, untracked inventory, sync gaps, and which analyses the current data can honestly support.
- Profitability analysis: separate gross sales, discounts, refunds, net revenue, and contribution margin covered by recorded cost data.
- Inventory risk: estimate demand rate, days of stock cover, stockout exposure, revenue at risk, and inventory with no observed sales history.
- Shopify-to-GA4 reconciliation: compare the store’s recorded purchase economics with GA4 purchase tracking and name the reasons a gap may exist.
The distinction matters. nqzai does not represent a margin calculated from only some products as the store’s total margin. When a variant lacks a recorded Shopify cost, the report discloses the coverage percentage. If an optional estimate is shown, its assumption is labelled as an assumption—not a discovered fact.
Why Shopify is the transaction source of truth
GA4 is useful for answering how sessions, channels, campaigns, and landing pages behaved. It is not the same record as the store’s order ledger. Consent choices, ad blockers, tag implementation, timezone differences, refunds, cancelled or test orders, tax, and shipping can all create valid differences between the two systems.
For a revenue question—“how much did we sell?” or “which products made money?”—nqzai starts with Shopify. For an acquisition question—“which channel brought the traffic?”—it uses the analytics and search connectors. For a discrepancy question, it puts the two beside each other instead of forcing a false merge.
That follows Shopify’s own platform model: the Admin API exposes store resources, while analytics implementations record their own events. The connector uses the Admin GraphQL API, bulk operations for larger imports, and webhooks for timely change signals. Shopify documents those patterns in its guides on authentication and authorization, bulk query operations, and webhooks.
A decision workflow that shows its working
A credible commerce answer needs more than a total. nqzai follows a short, inspectable sequence.
1. Connect with scoped access
The connection validates the store domain and OAuth response, then stores access credentials per shop rather than in a generic connector record. That makes the implementation ready for merchants who operate more than one store, while keeping the connector card simple.
Access is progressive. Catalog and inventory questions need different data than order and refund analysis; write permissions are not required for the current read-only reports. Shopify treats some order-related resources as protected customer data, so the product remains explicit about the scopes and data required rather than implying that every report is available immediately. Shopify’s protected customer data guidance is the relevant policy reference.
2. Import facts, then reconcile them
Initial catalog and order history can be large. nqzai uses Shopify bulk operations to import records efficiently, then maintains freshness through webhook-driven updates and reconciliation. Webhooks accelerate an update; they are not treated as the only proof that data is complete. The connector records sync state so a merchant can distinguish a live finding from a stale one.
This is also why large products need care: a product webhook cannot be assumed to contain every variant. The connector can request a targeted refresh when the payload signals that more variants exist. That is a small engineering detail with a big reporting consequence: profitability and inventory are only as complete as the variants beneath a product.
3. Check data coverage before recommending action
Before asking which product to promote or reorder, ask whether the data can support that claim. Store Health makes missing cost records and inventory coverage visible. It also identifies what is unlocked by the current connection.
For example, a product can have strong revenue but no recorded unit cost. nqzai can report the revenue; it cannot truthfully rank that product’s contribution margin. The useful next action is not a synthetic score—it is a direct link back to the product in Shopify so the merchant can record Cost per item. The next sync incorporates that fact.
4. Answer the business question with qualified evidence
Profitability reports distinguish gross sales, discounts, refunds, net revenue, and known contribution margin. Inventory reports distinguish variants with enough sales history from those where a forecast would be speculative. Reconciliation reports compare like with like: Shopify’s gross-at-purchase figure, including the relevant tax and shipping composition, against the GA4 purchase value for the same period.
The answer should include its time window, source timestamps, currency, coverage, and caveats. That is a better merchant experience than giving a single precise-looking number that cannot be audited.
5. Keep action controlled
The connector’s current work is intentionally about truth and diagnosis, not autonomous price changes. nqzai can explain a decision path, but it does not use a language model to calculate financial metrics or write directly to a store. The broader product direction is governed recommendations: evidence first, constrained proposal second, explicit approval before any future execution, and a reversible record of what changed.
The five questions to ask nqzai
| Merchant question | What nqzai checks | What a good answer includes |
|---|---|---|
| “Is my Shopify store connected and current?” | Store connection, granted scopes, sync timestamps, product/order counts, errors | Connection state and data freshness—not a vague success message |
| “Can I trust my store data?” | Cost coverage, inventory coverage, reconciliation state, sync errors | The missing fields and the analyses they prevent |
| “How profitable was the store?” | Shopify sales, discounts, refunds, net revenue, recorded cost coverage | Known contribution margin plus coverage; estimates clearly labelled |
| “What is likely to stock out?” | Variant inventory, recent demand rate, days of cover, history sufficiency | Risk horizon, revenue exposure, and an explicit insufficiency warning where needed |
| “Why does GA4 disagree with Shopify?” | Comparable Shopify purchase value and GA4 revenue for the same window | Named factors such as refunds, test orders, tax/shipping, consent, and timezone |
How to get a useful first report
- Connect the correct myshopify.com store. Use the canonical store domain and approve only the scopes needed for the first questions.
- Wait for the initial sync to complete. Check connection and Store Health before treating any profitability result as comprehensive.
- Fill material cost gaps in Shopify. Record Cost per item for sold variants that lack it; nqzai synchronizes the field and reports the coverage impact.
- Ask a narrow question with a timeframe. “Show product profitability for the last 30 days” is more auditable than “what should I sell?”
- Use reconciliation as a diagnostic, not a contest. If Shopify and GA4 differ, inspect the named factors before changing tags or declaring either system wrong.
- Treat forecasts as risk signals. Reorder decisions also require supplier lead time, cash constraints, and merchant judgement; a short sales history should reduce confidence, not increase automation.
What makes this different from a generic AI dashboard
The advantage is not that nqzai promises a magic number. It is that the product makes the boundary between observation, estimate, and decision visible.
- Typed sources instead of chat-memory arithmetic. Store facts, GA4 behaviour, and search data each retain their own role.
- Coverage before certainty. Missing COGS and sparse sales history change what the report can claim.
- Store-level and product-level questions stay distinct. A whole-store summary cannot substitute for a ranked product-margin view.
- Reconciliation explains rather than overwrites. Shopify and GA4 are expected to disagree in understandable ways.
- No hidden write path. The current connector reads, diagnoses, and guides. Any future write capability must be a separate approved, constrained, and reversible workflow.
Shopify’s API also has practical operating constraints—cost-based throttling and bulk-operation limits among them. nqzai queues work by shop and uses the platform’s returned throttle information rather than letting one large import impair every merchant. Shopify explains its current API rate-limit model in its developer documentation.
What this connector does not claim
It does not claim that every connected store has enough history for a reliable demand forecast. It does not infer a true contribution margin when product costs are absent. It does not promise that GA4 and Shopify will match. And it does not automatically change prices, discounts, inventory, or storefront behaviour.
Those limits are a feature. A merchant should be able to see what data supports a recommendation, decide whether the uncertainty is acceptable, and retain control over the outcome. That is the foundation for trustworthy inventory, merchandising, experimentation, and profitability work.
The Ecommerce Intelligence series
This seed article anchors a 30-post ecommerce cluster. The follow-on guides will go deeper into the individual capabilities and the questions merchants actually search for: Shopify COGS coverage, product contribution margin, refund-adjusted revenue, GA4 versus Shopify revenue, stockout risk, dead stock, reorder signals, catalog data quality, ecommerce attribution, and governed AI recommendations.
Each article will use the same standard: say what the data can establish, link the relevant evidence, identify the operational decision, and name the uncertainty. That gives merchants something more useful than a list of ecommerce buzzwords—a repeatable way to make better decisions with their own store data.
Frequently asked questions
Does nqzai replace Shopify reporting?
No. Shopify remains the transaction system of record. nqzai adds a governed analytical layer that connects store facts to data health, profitability, inventory risk, and acquisition reconciliation.
Can I use it if some products have no cost recorded?
Yes, but the margin analysis will disclose its recorded-cost coverage. Add Cost per item in Shopify for the affected variants to improve what can be measured. nqzai does not silently turn incomplete cost data into a definitive margin claim.
Why would Shopify and GA4 show different revenue?
They are different systems with different event definitions and failure modes. Refunds, discounts, cancelled or test orders, tax and shipping composition, consent, ad blockers, tag configuration, and timezone can all contribute. The reconciliation report names those factors for the selected period.
Will nqzai automatically reorder stock or change prices?
Not in the current connector. It is read-first: it helps a merchant understand the data and risk before any future action is proposed. The product direction is approval-based and reversible, not autonomous store mutation.
What should I do after connecting?
Start with connection status and Store Health, correct any material COGS gaps, then run a bounded profitability, inventory, or reconciliation question. That sequence yields an answer a finance, operations, or growth teammate can inspect.
Final takeaway
Shopify analytics become useful when they preserve the difference between sales, margin, attribution, and risk. nqzai connects those facts without pretending that incomplete data is complete or that a forecast is a promise. Start with a healthy store record, ask a precise question, and use the resulting evidence to decide what deserves attention next.
Explore the Shopify capabilities
Use these focused guides when you need a deeper workflow for a specific Shopify question:
