Shopify Analytics Reports: What Each One Tells You

Shopify gives you seven native report sections. Most operators use maybe three. The other four sit untouched because they either answer questions nobody asked, or they answer them so badly that the answer is worse than no answer at all. Understanding what each report actually tells you - and more importantly, what it hides - is the difference between a store that scales and one that gets confused at $500K revenue and never figures out why.

What ShopifyQL exposes vs. what it hides.

Every report you'll see in Shopify Admin is rendered from a query language called ShopifyQL. Get specific about what ShopifyQL actually carries before you read each tab. We confirmed this against a live store on 2026-03-09 while building Sentinel. Eight datasets are queryable: sales, orders, products, customers, sessions, payment_attempts, web_performance, and benchmark. Four metrics that operators expect to find are not in ShopifyQL at all: bounce rate, exit rate, page-level analytics, and reliable session duration (the field exists but reads “0 seconds”). The Behavior report below is the visible symptom of that gap. The shopifyqlQuery field landed in the GraphQL Admin API in version 2025-10; custom-app access has required OAuth since January 2026.

The seven reports, in order.

Overview. The dashboard you live on. Revenue, sessions, conversion rate, average order value, trending over seven days. It answers "are things going up?" Nothing else. Useful for spotting day-of crashes. Useless for diagnosis.

Sales. Revenue broken down by day, product, collection, and sales channel. The time-series is clean and accurate. The granularity stops there. You see total revenue per product. You don't see revenue per customer per product, or which customer cohorts drive which products, or whether product mix is shifting because demand changed or because acquisition changed.

Customers. Total customer count, repeat rate, and cohort-level revenue aggregation. This is where the first major blind spot appears. Repeat rate tells you what fraction of your customers came back. It doesn't tell you whether repeat customers are actually more profitable, whether retention is improving or declining, or whether the repeat rate is declining because acquisition improved (more new customers diluting the total) or because product degraded.

Products. Sales volume and revenue per product. No metadata attached - no supplier, no margin, no fulfillment time, no connection to inventory. You see which products sold. You don't see which products are actually profitable after you account for cost of goods and fulfillment labor.

Acquisition. Traffic sources, device, geographic, landing page. Integrated from GA4 if you wired it. This connection is fragile. Any tracking setup slip and you lose two weeks of visibility. More importantly, Acquisition answers "where did the session come from?" It does not answer "did that session convert?" because the attribution link between GA4 and Shopify is weak. You can see 100K sessions from Meta and $50K revenue attributed to Meta. You have no way to know if those are the same customers.

Behavior. Page views, bounce rate, average time on page per landing page. This report is the most direct evidence of the ShopifyQL gap above — none of those metrics live in the Shopify data layer, so the report is patched together from a fragile GA4 link. Don't use it. Everything you need from it you get better from GA4 directly, and Shopify's connection to it is flaky enough that you'll waste more time debugging the data than you'll save from the insights.

Finance. Revenue and refunds. Useful if you're doing basic bookkeeping. Otherwise it's a view of what you already know from Shopify's transaction ledger. Skip this.

Custom. The only report worth building. Shopify lets you create custom views on your data. Most operators never find this because it's buried in the UI. This is where the real work happens - if you know SQL and understand database structure well enough to join your own tables. For most stores, this is too high a bar.

What you're really missing.

By $500K annual revenue, you've hit the wall. The questions that matter are no longer answerable inside Shopify's native reports. You need to know: Which customers are profitable? Which marketing channels are actually driving conversions, not just clicks? Which product changes move the needle on retention? Is my repeat rate declining because I'm acquiring worse customers, or because the product got worse?

Shopify's reports can't answer any of these. They're designed to tell you what happened. They're not designed to tell you why it happened, or what to do about it. The moment you need causation instead of correlation, you've outgrown Shopify's native tooling.

The cost is real. If you're misallocating marketing budget because attribution is broken, you're probably wasting $10K–$40K per month by $1M revenue. That's $120K–$480K a year that could have gone to the channels actually driving customers. Most stores never discover the problem because they assume Shopify's dashboard is telling them the truth.

What to build instead.

The answer isn't to replace Shopify reporting. It's to build on top of it. Wire your data to a warehouse on day one. Connect your ad platforms, email list, SMS gateway, organic channels - every data source that touches the customer journey. Then build a reporting layer that joins all of it.

Sentinel is the layer we built to solve this problem. It sits on top of your Shopify data and watches for the patterns that matter: cohort decay, attribution drift, margin collapse, channel shift. It tells you what's actually changing, not just what the numbers are.

The alternative is hiring someone to manually reconcile your data every week. Most operators try this first. It breaks the moment your channels change or your volume scales. By then you're usually down $100K–$300K in efficiency losses, and you're finally ready to build the system right.

When to fix this.

If you're under $100K annual revenue: stay in Shopify. You don't have the volume for these patterns to matter. Focus on product and positioning instead.

At $100K–$500K: pay attention to attribution. You're probably running 2–4 marketing channels. Wire them to a spreadsheet once a week. Track which channels drive actual customers, not just clicks. This is a manual process but it'll save you from making a major budget allocation mistake.

Past $500K, especially past $1M: the manual approach breaks. The cost of bad data exceeds the cost of fixing it. Build a Sentinel-style intelligence layer. It pays for itself in the first month by catching one major budget mistake or one cohort problem early.