How to Use Product Analytics to Optimize Your eCommerce Catalog (Stop Guessing Which Products to Keep)

You have 400 products in your Shopify store. Your analytics dashboard shows you the top 10 by revenue. You promote those, reorder them, put them in your email campaigns. The other 390? You handle them by feel: restocking what sold last season, dropping what didn’t move, keeping the new arrivals you like.

That’s not a catalog strategy. It’s pattern-matching with incomplete data.

The problem isn’t that you’re lazy. It’s that your tools don’t give you the right view. Revenue rank tells you what sold. It doesn’t tell you what converted, what got abandoned, what brought customers back, or what’s quietly killing your store’s profitability one return at a time.

Why Revenue Rank Is the Wrong Lens for Catalog Decisions

Take two products from the same Shopify store:

• Product A: 4,200 sales last quarter, €38 average price, consistently in your top 5 by revenue
• Product B: 340 sales last quarter, €38 average price, barely makes the top 50

By revenue rank, Product A is your star. Product B is a footnote.

Now add conversion rate and repeat purchase data:

• Product A: 1.8% CVR, 11% repeat purchase rate, 64% return rate
• Product B: 9.4% CVR, 58% repeat purchase rate, 7% return rate

Product A has a return problem. It generates a lot of gross revenue, then destroys a lot of margin in refunds and processing costs. It converts poorly. And almost none of its buyers come back.

Product B converts at 9.4%, meaning roughly 1 in 10 visitors who view it buy it. More than half of its buyers purchase again within 90 days. It has almost no returns.

If you knew this, would you still be promoting Product A in every email campaign? Would you still be putting it in your paid ads? Would you keep ordering it in the same quantities?

This is what eCommerce product analytics is actually for. Not charts. Not dashboards. Decisions like this one, made systematically across your whole catalog.

The Four Catalog Metrics That Actually Matter

When you look at catalog management through a product analytics lens, four metrics do most of the work:

1. Product-level conversion rate (CVR)

Not overall store CVR. Per-SKU CVR: the percentage of sessions that include a product page view and result in a purchase of that product. A 2.8% overall store CVR is meaningless if product A converts at 0.4% and product B converts at 9.1%.

Stormly’s product performance table shows this broken out by SKU, so you can sort by CVR and immediately see which products are pulling down your overall conversion rate versus which are carrying it.

2. Cart abandonment rate by SKU

Most merchants track cart abandonment as a store-level metric. But abandonment is almost never uniform. In a typical 300-product store, 15–20% of products have abandonment rates more than 2x the store average. Those are products with a problem: wrong price, confusing variants, misleading photos, slow delivery estimate on the product page.

The cart abandonment analytics by product view in Stormly flags exactly these outliers. If product C has a 71% abandonment rate against a 34% category average, that’s not a funnel problem. That’s a product problem, and a recovery email won’t fix it.

3. Repeat purchase rate by product

This is the most underused catalog metric. Which products bring customers back?

It’s almost never the obvious answer. In most stores, 1–2 products in the top 20 by revenue are generating 30–40% of all repeat purchases. Pull those out of your acquisition strategy and your LTV numbers drop fast.

The repeat purchase analytics view gives you this breakdown by SKU. Once you know which products are building your loyal customer base, you promote them differently, as loyalty anchors, not just revenue drivers.

4. Customer LTV by first-purchase product

Some products attract one-time buyers. Others attract repeat buyers. This is the LTV split, and it’s one of the most powerful levers in catalog management.

Customers who first bought from category X have 2.8x higher 90-day LTV than customers who first bought from category Y. If you’re spending the same acquisition budget driving traffic to both categories equally, you’re leaving a lot of margin on the table.

The customer lifetime value analytics by product view shows this first-purchase cohort breakdown: which entry products predict high LTV customers, and which ones bring in people who never return.

---

Try this in your own store: see how your catalog products rank by CVR, repeat rate, and LTV cohort in Stormly → Free trial

---

Four Catalog Decisions Product Analytics Makes Obvious

Once you have these four metrics, the decisions get much clearer.

Decision 1: Which products to promote more aggressively

These are your high-CVR, high-repeat-rate products. They convert well from organic and paid traffic. They build your loyal customer base. They’re the ones that should anchor your email campaigns, appear prominently in your paid ads, and get featured placement in category pages.

In most stores, this list is smaller than you’d expect, often 8–12 products out of hundreds. But concentrating promotion energy on them has an outsized impact.

Decision 2: Which products to troubleshoot or reposition

High traffic, low CVR, high abandonment. These products have an audience but something is broken in the conversion path. It might be the price point, the product photos, the description, the variants setup, or the shipping cost showing up too late.

Stormly’s funnel view broken down by product category helps isolate where the drop-off happens. Is it at the product page view stage? The add-to-cart stage? The checkout stage? Each drop-off point points to a different fix. This is what product-level funnel analytics makes actionable that store-level funnel data simply can’t.

Decision 3: Which products to retire or deprioritize

Low CVR, low repeat rate, low LTV contribution, high return rate. These products are consuming catalog space, SKU management overhead, and potentially misleading your store’s overall performance metrics.

The hard part isn’t the decision. It’s identifying these products before they’ve been in your catalog for three years and you have sunk cost bias about them. Catalog analytics makes this a data exercise, not a gut-feel one.

Decision 4: Whether new arrivals are tracking toward success or stalling

A new product launch has a predictable trajectory. By day 7, you can see early CVR signals. By day 14, the add-to-cart rate is stabilizing. By day 30, you know whether repeat purchase rate is developing or whether this product is a one-time-buy category.

Stormly’s new product launch analytics view surfaces these day-over-day benchmarks automatically. You’re not waiting until the end of quarter to decide whether to reorder. You know by day 30 whether this product belongs in your catalog long-term or not.

What Stormly’s Catalog Performance View Actually Shows

The core view in Stormly for catalog management is the product performance table. It’s not a revenue report. It looks more like this:

Product	CVR	Cart Abandon Rate	Repeat Purchase (90d)	LTV Cohort Index
Merino Beanie (Navy)	9.4%	28%	58%	2.3x
Wool Scarf Set	7.1%	31%	49%	1.9x
Premium Gloves	1.8%	71%	11%	0.6x
Cotton Cap	3.2%	44%	18%	0.8x

The Premium Gloves row is the one that jumps out. High abandonment, low repeat rate, below-average LTV cohort. This product either has a positioning problem or a product quality issue, and neither will be fixed by putting it in more emails.

The Merino Beanie row tells a different story. Customers who find it convert well, come back, and have higher downstream LTV than the average buyer. This is a product worth investing in for acquisition.

Stormly flags the outliers automatically using its anomaly detection layer. You don’t need to sort through 400 rows manually. The products that are significantly above or below category benchmark surface in the AI insights feed. This is the same anomaly detection capability described in the eCommerce revenue anomaly detection guide, applied at the product level.

A Simple Framework for Your First Catalog Analysis

You don’t need to analyze all 400 products at once. Start with this four-step process:

Step 1: Sort your top 50 revenue products by CVR

You’re looking for the gap between your revenue rank and your CVR rank. Products that rank high by revenue but low by CVR are candidates for troubleshooting. Products that rank low by revenue but high by CVR are candidates for promotion.

Step 2: Pull the abandonment rate for your top 20 revenue products

Find the three with the highest abandonment rate. These are your most urgent conversion problems. They’re already getting traffic and adding to carts, but something is blocking the purchase.

Step 3: Find your repeat-purchase anchors

Sort by 90-day repeat purchase rate. The top 5 products by this metric are your loyalty drivers. Are they in your acquisition campaigns? They probably should be.

Step 4: Check your last 5 new arrivals at day 30

For each product launched in the last 30–90 days, pull CVR, abandonment rate, and early repeat purchase signals. The ones tracking above category benchmark deserve investment. The ones tracking below need a decision: troubleshoot or deprioritize.

This process takes about 20 minutes in Stormly. It replaces a quarter of gut-feel catalog decisions with data-backed ones.

The eCommerce analytics workflow for teams guide covers how to build this kind of analysis into a weekly rhythm, so catalog decisions happen consistently rather than once a quarter when someone notices a problem.

The Catalog Decisions Most Stores Never Make

The reason most eCommerce stores manage their catalog by feel isn’t because the data doesn’t exist. It’s because session-level analytics (GA4, Shopify Analytics, most standard tools) doesn’t break behavior down to the product level in a useful way.

Shopify Analytics shows revenue by product. GA4 shows sessions and page views. Neither shows you CVR by SKU, abandonment rate by product, or LTV cohort by first-purchase category. These metrics require product analytics tools that are built specifically for eCommerce data structures, not adapted from SaaS product tracking tools.

That’s the gap. And it’s why stores running 300+ SKUs on the same default analytics stack as a 30-SKU startup are flying blind on catalog decisions that compound over time.

Getting clear on which products are actually working, and which ones are consuming catalog resources while contributing little, is one of the highest-leverage improvements most stores can make. You don’t need to rebuild your entire analytics setup. You need the right view on the data you already have.

Stop managing your eCommerce catalog by gut feel. See how your products actually perform in Stormly → Free trial