Shopify Analytics: which numbers matter each week?

Shopify Analytics becomes useful when a question comes before the dashboard. “Why are mobile orders down?” needs a different analysis from “Which product leaves the best contribution?” Do not start by copying every available metric.

Read at three levels

First, review orders and net sales over a comparable period. Next, separate sessions and orders by source, device and landing page. Finally, inspect product families: views, cart adds, units sold and returns. Check each metric's definition in the tool; gross sales, net sales and average order value answer different questions.

Illustrative example: sessions double after a video campaign but orders stay flat. The overall rate falls, without proving the store got worse. Separating video traffic from search traffic shows that campaign visitors read a guide then leave. The team improves the guide's link to a relevant category before altering checkout.

Keep an event log

Record promotions, stockouts, payment failures, new content and tracking changes beside the numbers. A difference may come from a campaign or a missing analytics tag. Check consistency between actual orders, payments and reports, especially after an app installation or migration.

Choose one action for the week

End the review with one task: clarify a category, test mobile cart or repair a missing shipping option. Write the hypothesis, expected measure and review date. The dashboard directs work; it does not alone prove the cause of a change.

Read every metric with its denominator

Revenue growth may come from more orders, larger carts, a higher price or a change in refunds. Start with paid orders and net revenue for a defined period, then inspect sessions, product views, carts and checkouts. Use the same dates and time zone for comparisons. A conversion rate based on ten visits moves dramatically after one order; avoid firm conclusions from tiny samples.

Next rank products by contribution rather than revenue alone. A high-revenue item may consume much of its value in shipping or support. Inspect heavily viewed products with few orders, then read their questions, stock status and returns. A metric directs attention; it does not explain cause by itself. Consult the Shopify report definitions available to your plan because presentation and available metrics can vary.

A practical weekly read

In an illustrative case, sessions rise from 500 to 700 while orders move from 20 to 21. Traffic is up, orders barely so. Before rewriting the homepage, the team checks sources: the increase came from a campaign sending visitors to an incomplete product page. They inspect mobile rendering, delivered price and variants. If new visitors leave at the product page, the likely issue differs from abandonment at the last payment step.

Keep a dated journal of promotions, campaigns, added items, stockouts and theme edits. Without context a chart invites a false explanation. Reconcile refunded and canceled orders with the sales view so the team does not celebrate revenue it will not keep.

Turn the report into a decision

Pick one operating question each week: “Are extra visits to this page producing orders with margin?”, “Why do carts from this region fail?” or “Which variant sells slower than expected?” Record the baseline, a hypothesis, an action and review date. If data are sparse, watch user sessions and speak to buyers before assigning a cause. The point of reporting is a checkable action, not another dashboard.

Questions when opening reports

Which figure comes first? Paid orders and net revenue for a defined period, then contribution if you can calculate it. Sessions and conversion rates help locate a problem but cannot replace economic results.

Is higher traffic progress? Only if visitors could be buyers or the traffic teaches something useful. Compare source, landing page, engagement, cart and orders. A broad campaign can double visits without bringing one additional sale.

Why do figures differ from the ad platform? Tools may use different time zones, attribution windows and sales definitions. Record differences and use actually paid orders as the operational reference.

How often should reports be checked? A weekly review often suits routine decisions; payment and stock checks may be daily. With low order volume, extend the analysis period before claiming a trend.

A twenty-minute weekly review

Use the same weekday and date range each time. Start with paid orders, refunds and net revenue; then inspect product groups by contribution if your cost data allow it. Compare traffic source, landing page, product views, carts and checkout for the one question you selected. Avoid reading all available reports without a decision to make. A chart becomes useful when it points to a page or process you can inspect.

Keep a simple change log next to the numbers. Record campaigns, promotions, new products, stockouts, price changes and theme updates. When a metric moves, check that log before assigning a cause. Choose one action, give it an owner and decide when to review the result. If order volume is very low, make the observation period longer and supplement it with customer conversations. Weekly reporting should reduce uncertainty, not manufacture confidence from small samples.

Related guides

• Ecommerce conversion: diagnose the journey before testing
• Ecommerce SEO audit: the checks that come first

Further reading: official documentation.