What Shopify Analytics Actually Tells You - And What It Doesn't.

Most Shopify operators never look past the Overview dashboard. Revenue is up. Sessions are up. Add-to-cart conversion is up. The numbers move in the right direction, so the store gets left alone. This is how stores break.

Shopify's native analytics does one job extremely well: it tells you what happened. The Overview tab shows you total revenue, AOV, conversion rate, and a seven-day trend. The Sales report breaks down revenue by day. The Customers report tallies your repeat buyers. The Acquisition report names your traffic sources. All of these are real. All of them are incomplete. And by the time you realize what's missing, you've usually left $20,000 to $200,000 on the table.

We built Sentinel specifically because operators we worked with kept hitting the same wall: the moment they crossed $1M in annual revenue, Shopify's native analytics stopped answering the questions that mattered. This post is the map of that wall. It names the four specific gaps, explains why they exist, and walks through the intelligence layer that operators use to plug them without rebuilding their entire stack.

What ShopifyQL surfaces — and what it hides.

Before you decide what to build on top of Shopify, get specific about what its data layer (ShopifyQL, the query language behind every native report) actually contains. We tested this directly against a live Shopify store on 2026-03-09 while building Sentinel. The result is a small, boring, useful inventory.

The eight datasets ShopifyQL exposes: sales, orders, products, customers, sessions, payment_attempts, web_performance, and benchmark. The shopifyqlQuery field was added to the GraphQL Admin API in version 2025-10, and as of January 2026 custom apps must use OAuth. So if you read older guides, the access path is now genuinely different.

What ShopifyQL does not contain at all: bounce rate, exit rate, page-level analytics, and reliable session duration. Session duration is in the schema but reads as “0 seconds” in production — not a bug you can fix from your end, just a metric Shopify never wired in. Pre-built admin reports (the Behavior section, the polished marketing dashboards) have no API endpoint at all. They live in the admin UI and stop there.

That distinction is the whole story. Anything you can ask FROM sessions or FROM sales against ShopifyQL is real, queryable, and yours. Anything that lives in the Behavior reports or behind a “page view” is locked to the admin UI or simply not tracked — you have to wire GA4 (or a custom Web Pixel) to see it.

What Shopify's UI shows on top of that.

Now the dashboards. Shopify gives you eight reporting tabs. Four are genuinely useful — and now you can read each one against the data layer above.

Overview. Hits revenue, sessions, conversion rate, average order value, and a rolling seven-day view. Backed by ShopifyQL sales + sessions. Good for trending, useless for diagnosis.

Sessions. Breaks down traffic by source, device, location, and landing page. Backed by the sessions dataset, which has dimensions for landing_page_path, utm_source, referrer_source, device_type, country, and a human_or_bot_session flag. Strong for traffic source. Silent on what happened on the page.

Sales. Revenue and conversion rate by day, with a breakdown by product, collection, and sales channel. Time-series is clean. Grain stops at order line.

Customers. Total customers, repeat rate, and cohort-level revenue aggregation. This is where the first crack appears. Repeat rate tells you what fraction of your customers bought twice. It does not tell you whether those repeat customers are more valuable than new ones, whether they're accelerating or slowing down, or whether you're acquiring the kind of customer who even wants to buy again.

Marketing. Ties into your Shopify app connections — Facebook, TikTok, Google — and attempts to show you return on ad spend. The connection is real. The attribution is not. More on that in a moment.

The inventory, finance, and acquisition reports exist. Don't spend time in them. They're shadows of what you need. The real work happens in the four above — and the real blindness starts the moment you need to ask a follow-up question that ShopifyQL has no field for.

The four walls you hit past $1M.

The pattern is consistent enough that we've given it a name. At around $800K–$1.2M revenue, typically six to twelve months after hitting their first seven-figure milestone, operators ask the same four questions in sequence. Shopify's native analytics cannot answer any of them.

Attribution gap. My marketing dashboard shows $120K spend and $380K revenue from ads. That's a 3.2x ROAS. But I'm running Google, Meta, TikTok, and email. The Shopify marketing dashboard can't see the full funnel. Which channel is actually driving the second and third touch, not just the last click?

Cohort blindness. I have 4,200 repeat customers. But are they profitable? New customers from January cohort are coming back 44% of the time. The June cohort came back 22% of the time. Why? And which cohort should I be optimizing the product to retain?

Cross-channel blindness. I can see what Shopify is telling me about ad ROAS. I have no idea what my server-side data shows. My email platform logs 45% open rate. My SMS gateway shows 8% click rate. My organic social has no tracking attached. Are these channels working or am I throwing money away?

Data lag. It's Tuesday morning. I want to know if Monday's campaigns worked so I can adjust today's spend. Shopify reports typically land by Tuesday evening. By then, I've already burned another $5K against the same targeting if the Monday numbers were bad. Even good tools have a 24-hour lag built in.

These four questions are not complicated. They're just not answerable inside Shopify. And they cost more than most stores can afford to lose.

Attribution: why Shopify sees the last click, not the journey.

Here's the structural issue. Shopify tracks the final conversion - the customer who clicked the ad and bought - and credits that touch. It sees the last click because that's the only touch that happens inside Shopify. Everything else is outside.

Reality is messier. A customer typical journey contains five to eight touches across channels. They see your TikTok ad on Sunday, don't convert. They land in your email list Tuesday. Wednesday they get a nurture email and click through to product page. Thursday they don't buy. Friday they search your brand on Google, click your ad, and buy. Shopify credits 100% of that $120 transaction to the Google campaign. It has no way to know that TikTok, email, and organic search all shaped the decision.

The math compounds. If attribution is off on 30% of your orders, and each misattributed order shifts spend decisions by 15% on average, you're probably overspending on one channel by $15,000–$40,000 a month and underspending on another by the same amount. Over a year, that's $180K–$480K of budget that never hit the right lever.

This is why seven-figure operators stop using the Shopify marketing dashboard for budget decisions around month four of running significant ad spend. They've either built their own attribution layer or hired someone to track it manually. Most choose to build, because manual tracking breaks the moment your channels change.

LTV and cohort analysis: the metric Shopify can't give you.

Shopify shows you total revenue per customer. It doesn't show you lifetime value, and the difference is the difference between a profitable business and a business that looks profitable until you do the math.

Total revenue per customer averages the 30-day buyer and the 18-month repeat customer. Revenue per customer might be $240. That's not a useful number. What you need is the curve. First-time customers from your January 2026 cohort spent an average of $87 in month one, $12 in month two, $8 in month three, and $0 in month four. LTV = $107. Your CAC is $85. Margin on $107 of revenue at 62% gross margin is $66. You're profitable by $3 per customer - assuming nothing breaks.

The June cohort spent $91 in month one, $7 in month two, $2 in month three, $0 after. LTV = $100. CAC = $120 if you're acquiring at the same rate. You're now underwater on that cohort by $15 per customer. Your product changes between January and June were supposed to improve retention. They did the opposite. Without this curve, you don't know.

Shopify's Customers report totals repeat buyer rate. It does not show cohort curves. It does not show decaying revenue over time. It does not show which changes to messaging, product, or price moved the curve up or down. Stores running at $1M+ revenue almost never have a working LTV model. Most discover they need one in month 18, after cohorts two through four have already failed.

Cross-channel silence: the channels Shopify can't see.

Shopify knows about the channels connected to its marketing dashboard. It has no sight into email, SMS, organic social, YouTube, partner networks, or any channel that doesn't have an official integration. If 40% of your revenue comes from channels Shopify can't see, then 40% of your reporting is missing.

This is the hidden problem with "organic" traffic. Your Sessions report shows traffic to the store, but it's blind to offline signals. A customer mentioned your brand on Reddit, or your product showed up on a Reddit thread three months ago. They search your brand name, land on your site, and convert. Shopify attributes it to organic search. It has no way to know that Reddit - a platform Shopify can't track - was the original signal.

The same is true for every dark-channel signal: word of mouth, wholesale partnerships, affiliate programs, TikTok influencer tags, and even your own email list if you're not using Shopify email. Shopify does not know these channels exist. You have to wire them in manually, and the moment you do, you realize how much revenue came from places you weren't measuring.

The solution is not a new tool. It's an intelligence layer.

Most operators' first instinct is to replace Shopify analytics with something else. Wrong move. The problem is not that Shopify is bad at reporting. The problem is that reporting is not enough. You need intelligence - a system that can see across channels, track cohorts, watch for drift, and answer "why" when the Overview dashboard shows something changing.

The architecture looks like this. Shopify stays as your source of truth for transaction data. GA4 becomes your source of truth for multi-touch sessions and attribution. Your ad platforms (Google Ads, Meta, TikTok) stay connected. Email, SMS, and any other channel get routed to a data warehouse on day one, not later when you "get serious." Then a layer sits on top that joins all four.

Server-side tracking is the load-bearing piece. Instead of relying on pixels and cookies - which iOS 15+ made unreliable - you send conversion events from your backend to your analytics vendor's server API. GA4 sees the full picture. The ad platforms get told which orders actually came from which clicks. Your warehouse gets timestamped records. Attribution becomes real.

The full reporting playbook walks through the integration. The short version: it takes two weeks to wire, costs nothing in tools, and solves 80% of the visibility problem on its own. The other 20% is the Sentinel-style intelligence layer that watches the numbers and tells you when something changed.

When to fix analytics vs. when to keep shipping.

Not every store needs this. Use this decision tree.

If you're under $50K/month revenue: stay in native Shopify. Your channels are probably simple. Your customer base is small enough that you know them. Fixing analytics is a distraction from the work that actually moves the needle - better marketing, better product, better brand. Do that first.

At $50K–$200K/month: the attribution problem starts to hurt. You're running more than two ad channels. You probably have an email list. You should be thinking about a real dashboard, but you don't need the full layer yet. GA4 + manual channel audits in a spreadsheet gets you to 80% of the answer.

Past $200K/month, especially past $1M annual: the cost of bad attribution exceeds the cost of fixing it. You're probably misallocating $10K–$50K per month. That's $120K–$600K a year. A proper reporting stack + intelligence layer pays for itself in month one.

How JAAX approaches this with operators.

We do two things. First, we build the reporting layer. Shopify + GA4 + your ad data + a warehouse that ties them together. This is the foundation. It ships in two weeks. Second, we build or integrate an intelligence system on top. Sentinel is what we built for our own store. We can build the same system for yours, or we can help you integrate an existing tool.

The reporting layer answers "what happened." Sentinel answers "why did it happen" and "what do I need to do about it." For a store at $1M–$10M revenue, the intelligence layer usually covers itself in the first month. A 3% improvement in ROAS on a $40K/month media spend is $14,400 in recovered efficiency. Sentinel costs a fraction of that.

The engagement usually runs two phases. Phase one is the integration and architecture sprint - two weeks, flat fee, outputs a live dashboard and an attribution model. Phase two is the ongoing intelligence layer - monthly, SLA-backed, tells you which numbers to move.

If you want early access to the tool, Sentinel is built for exactly this problem. Mid-article soft CTA: we're opening the waitlist later this month. If you want help building the reporting stack first, the AI integration page has the service details.

Tomorrow's homework.

Do this today. Open your Shopify analytics. Note down three questions you want answered: what channel should I spend more on, which customer cohort has the best LTV, and what changed yesterday that I need to know about. Try to answer each one using only Shopify's native reports. Count how many clicks it takes.

If you spent more than three clicks on any of them, or if you couldn't answer at all, that's the edge of Shopify's map. Beyond that edge is where seven-figure stores do their actual analysis. The architecture is boring. The payoff is usually between $50K and $500K recovered efficiency in the first year.