The eCommerce Analytics Audit: 10 Questions to Ask About Your Current Setup

Most eCommerce operators assume their analytics setup is working. They have GA4 installed, Shopify Analytics open in a browser tab, and a dashboard someone built two years ago. Data is flowing in. Charts are updating. That feels like enough.

It usually isn’t. Most setups are measuring the wrong things, missing 30–60% of purchase events, and producing reports nobody checks on Monday morning because they don’t help you decide what to do next. The problem isn’t that you’re ignoring analytics. It’s that you can’t tell whether your current setup is genuinely useful or just decorative.

These 10 questions will show you.

The 10 questions that reveal whether your setup is actually working

Answer them honestly for your current analytics setup right now.

1. Are your purchase events tracking accurately?

Open your analytics tool and look at your purchase count for last month. Then open your Shopify orders dashboard for the same period. Do the numbers match?

They rarely do. GA4 on Shopify misses between 30 and 60 percent of purchase events under default configuration, not because of a bug you can fix quickly, but because of how GA4 handles Shopify’s server-side order processing and multi-currency flows. If you haven’t looked at this gap before, the scale of GA4’s tracking inaccuracy on Shopify is worth understanding before you optimize anything based on that data.

If your analytics shows 120 purchases but Shopify recorded 190, you don’t have a conversion rate. You have noise. Every decision built on that number is built on a broken foundation.

Pass: Your analytics purchase count is within 5% of Shopify’s order count for the same period. Fail: There’s more than a 10% gap, or you’ve never compared these two numbers.

2. Can you see conversion rate by individual product?

Your site-wide conversion rate is 2.4%. What does that tell you? That somewhere between zero and all of your products are converting. Nothing more.

The question that actually drives decisions: which products have a 7% conversion rate, and which have 0.3%? Which ones are getting added to cart but abandoned before checkout, versus ones that rarely make it to cart at all?

Shopify Analytics shows total orders and revenue. It doesn’t break conversion rate down by SKU, by category, or by traffic source combined with product. If your highest-traffic product converts at 0.8% while a niche item converts at 9%, that tells you exactly where your media budget is going wrong. Without that breakdown, you’re guessing.

What Shopify Analytics doesn’t surface about your product performance is a long list. Product-level conversion rate is near the top.

Pass: You can filter by product and see conversion rate, not just units sold or revenue. Fail: You’re working with site-wide or category-level aggregates only.

3. Do you know which products appear most in abandoned carts?

Your overall cart abandonment rate is probably somewhere around 70%. The industry average has been in that range for years. What you probably don’t know is which specific products are in those abandoned carts.

That distinction changes everything you do about it. If your highest-priced item appears in 60% of abandoned carts, the problem might be price friction. If a specific SKU keeps showing up alongside products that do convert, it might be a trust issue on that product’s page specifically. Recovery emails sent to the whole segment convert at 5–8% on a good day. Targeted interventions for the specific product causing the abandonment can do meaningfully better.

Cart abandonment analytics broken down by product and SKU is a fundamentally different capability from knowing your overall abandonment rate. If you only have the latter, you’re treating the symptom.

Pass: You can identify the top 5 SKUs appearing most frequently in abandoned carts. Fail: You know your overall abandonment rate but not which products are driving it.

4. Can you tell which products drive repeat customers versus one-time buyers?

You have customers who order every 6 weeks. You have customers who ordered once in 2024 and never came back. What’s different about them?

A lot of the time, it comes down to which product they bought first. Certain items attract high-LTV customers who naturally repurchase across categories. Others attract bargain hunters who would never come back regardless of your retention emails. Most analytics setups can’t answer this question; they’ll show you repeat purchase rate as a single site-wide number, but they won’t show you which products in a customer’s first order predict whether they return.

If you’re spending money to acquire customers, not knowing which product acquisition is actually worth it is an expensive blind spot.

Pass: You can filter your customer base by first-purchased product and see retention or repeat purchase behavior by that segment. Fail: You know your aggregate repeat purchase rate but not which products drive it.

5. Do you know your actual retention rate at the cohort level?

“Our retention is around 30%” is one of those numbers that sounds like it means something. What it actually means depends entirely on how it was calculated, over what period, and for which customers.

A 30% retention rate across all customers since you launched your store is a completely different picture than 30% 90-day retention for customers acquired through paid social in Q4 2025. The second number tells you whether your Black Friday acquisition campaign is worth running again. The first number doesn’t tell you much at all.

eCommerce customer retention analytics at the cohort level, broken down by acquisition period, channel, or first-purchased product, is what separates operators who understand their numbers from those managing a metric they can’t actually see clearly.

Pass: You can see retention rate for specific cohorts: customers acquired in a given month, from a specific channel, or who first bought in a particular category. Fail: You have one retention number that applies to your entire customer history.

6. Can you see a customer’s full purchase history in your analytics tool?

When a customer contacts support, when you’re building a retargeting segment, or when you’re trying to understand why a cohort’s LTV is declining: can you pull up a specific customer and see every product they’ve bought, in order, over time?

This sounds basic. Most tools don’t do it cleanly. GA4’s user explorer requires a user ID you’d have to cross-reference manually. Shopify’s customer view shows orders but not the behavioral events between them. Being able to say “this customer bought X, then Y three weeks later, then stopped,” and having that visible in one place, is a capability most setups lack.

Pass: You can trace a customer’s purchase history and behavioral sequence in a single view. Fail: You’d need to cross-reference Shopify orders and your analytics tool separately to reconstruct this.

---

If you’ve answered “fail” on three or more of these questions, your setup is producing data that isn’t helping you make product decisions. That’s fixable, but only once you’ve named the gap clearly.

See what product-level analytics for Shopify actually looks like in Stormly → Start free trial

---

7. Does your team open the analytics tool without being asked?

This is the least technical question on the list and possibly the most diagnostic one. If checking analytics requires someone to remember, or requires a manager’s prompt, or is a scheduled ritual that gets skipped when things get busy, which means your setup isn’t integrated into how your team makes decisions.

A good analytics setup creates pull. People open it because it answers questions they’re already asking. A bad one creates friction: 12 menus to navigate, a custom report to build each time, or a slow export that has to run first.

The weekly analytics workflow that actually gets used looks very different from an impressive dashboard nobody checks.

Pass: At least one person on your team opens your analytics tool unprompted, at least three times per week. Fail: Checking analytics is a scheduled task, not a reflexive habit.

8. Can you get from a question to an answer in under 10 minutes?

You notice something in your sales: a category has been trending down for three weeks. You want to know whether this is across all SKUs in that category or isolated to one product. How long does it take you to find out?

If the answer is “I’d need to export to a spreadsheet,” your setup wasn’t designed for this kind of question. eCommerce moves fast enough that a 45-minute analysis process either doesn’t get done, or gets done after the window to act has closed.

The data paralysis that stalls most eCommerce teams isn’t a lack of data; it’s that the analysis process is slow enough that questions get abandoned before they get answered. Ten minutes is a reasonable bar. If your tool takes longer for common diagnostic questions, you’ll stop asking them.

Pass: Common diagnostic questions take under 10 minutes without exporting data. Fail: Non-standard questions require exports, spreadsheet work, or waiting for a data analyst.

9. Do you have automatic alerts when revenue or conversions drop?

Things go wrong in eCommerce without warning. A tracking pixel breaks silently. A product page has a checkout error on mobile that nobody reports for three days. A stockout on your top-converting SKU goes unnoticed until you check inventory. These problems can compound for days if you’re only looking at analytics when you think to.

A properly configured setup should alert you before you notice a problem manually, not for every minor fluctuation, but for statistically significant anomalies on the metrics that matter: revenue, conversion rate, cart additions, repeat purchase rate.

If your alerting system is the Shopify daily summary email, you’re finding out about problems after customers have already had a bad experience.

Pass: You have automated anomaly detection configured on at least revenue and conversion rate. Fail: You find out about drops when you happen to look, or when someone reports a problem.

10. Does your analytics tell you what to do next?

This is the hardest question, and the one most setups fail.

Data is easy. Interpretation is the bottleneck. You have 1,400 sessions this week, 38 orders, 2.7% conversion rate. Now what? Which of your 180 products should be in your next email? Which should you pull from paid search? Which category deserves attention in your next buying meeting?

The eCommerce KPIs that actually drive decisions aren’t the same as the 27 metrics on the default Shopify Analytics screen. The difference is whether each number tells you something actionable or just describes what already happened.

A setup that passes this question doesn’t just show charts. It surfaces which products need attention, which are outperforming their baseline, and what changed that’s worth investigating.

Pass: After your weekly analytics review, you can identify 1–2 specific product-level actions without additional analysis. Fail: You leave your analytics review with a general sense of how things went, but no specific next steps.

---

What a passing setup actually looks like

Realistically, most setups built on GA4 and Shopify Analytics alone will fail 4–7 of these questions. That’s not unusual; it’s the default. GA4 wasn’t designed for product-level eCommerce decisions. It was designed for session reporting and marketing attribution.

Passing all 10 requires:

• Accurate purchase tracking (not what GA4 gives you out of the box on Shopify)
• Product-level analytics, not just session or order-level aggregates
• Customer-level purchase history you can query without exporting
• Cohort analysis broken down by product, channel, or acquisition period
• Anomaly detection that runs automatically, not when you remember to look
• A UI fast enough that people use it without being prompted

Stormly is built specifically for this combination. Not as a replacement for your marketing attribution tool, but as the product analytics layer that sits between your Shopify data and the catalog decisions you make every week.

Within your first session, you can connect your store, run the product-level cart abandonment report, and see which SKUs appear most in abandoned carts. That’s the kind of question a passing setup answers in under 5 minutes.

What to do if you failed the audit

Don’t try to fix everything at once. The order matters.

Fix tracking accuracy first. If your purchase event data is wrong, everything built on top of it is wrong. Check the gap between GA4 purchase counts and Shopify order counts. Decide whether to repair the integration or switch to a tool that tracks Shopify accurately by default.

Get product-level visibility second. You need to filter by SKU and see conversion rate, cart abandonment, and repeat purchase behavior at the product level. If your current tool can’t do this without a spreadsheet export, this is the gap costing you the most decisions.

Build a workflow that sticks third. A good analytics setup that nobody uses is still a failing setup. The eCommerce analytics workflow your whole team will actually use is worth thinking about before you invest in new tooling, because the tool matters less than whether people open it.

Run your first product-level analytics audit in Stormly → Free trial