by Trivas.aiMost SKUs don't fail because they're bad products. They fail because nobody looked past the revenue column.
Product performance analysis is the process of evaluating how individual SKUs perform across revenue, margin, returns, and ad spend, not just whether units moved. That distinction matters more than it sounds. A product can sell well and still be quietly losing you money every time someone buys it.
Search the term and you'll find a dozen guides covering the same ground: pull your top sellers, sort by revenue, done. That's sales reporting wearing a performance analysis costume. It skips margin entirely, ignores returns, and treats every dollar of revenue as equally good, which it isn't.
This piece goes further. You'll get the full metric list that actually defines performance, a step-by-step framework you can run every month, a breakdown of where this data lives (and why it's scattered across three or four platforms by default), a manual-vs-automated time comparison, and an FAQ section for quick reference. It's written for DTC founders and growth leads running Shopify, Amazon, or both, who need SKU-level clarity before they plan next quarter's catalog, not another dashboard that just restates units sold. If that's your team, the kind of work covered here usually lands with whoever owns reporting, often a data analyst or the founder filling that role by default.
Revenue per SKU and units sold per SKU look similar but tell different stories. A SKU moving 500 units a month at a thin margin can dominate your units-sold report while a 50-unit SKU with triple the margin quietly outperforms it on actual profit. If you're only tracking volume, you'll miss this every time.
That's why contribution margin per product (price minus COGS, fulfillment, and allocated ad spend) should outrank raw revenue in any serious review. It's the metric that tells you what a SKU actually puts in the bank, not just what it rang up at checkout.
Sell-through rate matters more than people give it credit for, especially for inventory-heavy brands. A SKU with great revenue but a 20% sell-through rate is tying up cash in a warehouse somewhere.
Return and refund rate by SKU is the one that erodes "top performer" status without anyone noticing. A product with strong initial sales and a 25% return rate isn't a winner, it's a slow leak. And ad-attributed revenue per SKU separates products that are actually earning their ad spend from ones riding organic halo off a bestseller's ad budget. If you need precise definitions for any of these, the data dictionary is a decent reference point before you start pulling numbers.
For most brands, the data sits in four places: Shopify order exports, Amazon Seller Central reports, Meta and Google ads platforms, and GA4 funnel data. None of them were built to talk to each other.
The first break point is ad attribution. Meta and Google report performance by campaign and ad ID, not by the SKU IDs sitting in your storefront. Matching the two requires manual UTM mapping or product-ID reconciliation, and it's the exact spot where most spreadsheet-based analysis quietly falls apart.
Brands selling on both Shopify and Amazon have it worse. You end up with two separate inventory systems, two separate sales ledgers, and zero shared view of how a SKU performs across both. Amazon's reporting structure doesn't map cleanly onto Shopify's, so "performance" ends up meaning two different things depending on which tab you have open.
This fragmentation, not laziness or lack of spreadsheet skill, is the real reason searches for product performance analysis increasingly point toward dashboard and tooling solutions instead of templates. A template can't reconcile two inventory systems for you.
Step 1: Pull a 90-day baseline of units, revenue, and refunds per SKU from each sales channel separately. Don't blend them yet.
Step 2: Layer in COGS and fulfillment cost per SKU to calculate contribution margin. Gross revenue alone will mislead you here, every time.
Step 3: Match ad spend to SKU using UTM parameters or product-ID mapping, not channel-level totals. "We spent $10K on Meta this month" tells you nothing about which products that spend actually supported.
Step 4: Rank products into tiers, margin leaders, volume leaders, and underperformers, instead of one flat revenue-sorted list. A single ranking hides too much.
Step 5: Flag any SKU where return rate or ad cost is climbing faster than revenue over the same window. That's an early warning sign, not a footnote.
Step 6: Set a review cadence. Monthly if your catalog has frequent SKU turnover, quarterly if it's stable. Either way, this needs to be a standing process, not a one-off report you run once and forget.
Ranking products by gross revenue alone is the most common one, and it hides margin-negative bestsellers in plain sight. A SKU at the top of your revenue report can be losing money on every unit once fulfillment and ad spend are factored in.
Ignoring late returns is another. Returns processed after your reporting window closes understate true refund rate, making a product look healthier than it is for months.
Last-click attribution is a third problem. Crediting all ad spend to the final SKU in a conversion path ignores assisted conversions entirely, which inflates some products and starves others of credit they earned.
Comparing Amazon and Shopify performance on mismatched time windows or currency bases is an easy trap too, especially for brands selling internationally. And treating one viral spike, a TikTok moment, a single big influencer post, as a stable signal instead of checking trend across multiple periods will have you reordering inventory for demand that already evaporated.
Running this manually means exporting from Shopify, Amazon, and whatever ad platforms you're on, reconciling SKU IDs by hand, and rebuilding margin formulas from scratch each cycle. For a multi-channel catalog, that's typically several hours per reporting cycle, and it has to happen again next month.
The process holds up fine under 100 SKUs. Past a few hundred, it starts breaking in predictable places: formula errors creep into the spreadsheet, COGS data goes stale because nobody updated it after a supplier price change, and manual ad attribution mapping becomes its own part-time job.
This is the problem a Redshift-backed reporting layer is built to solve, which is the approach Trivas's BI reporting is built on. Instead of rebuilding margin and attribution data by hand every cycle, it stays refreshed automatically as new orders and ad spend come in.
Worth being clear about this section's purpose: it's not a pitch disguised as information. The point is to help you weigh your own time cost against your catalog size, not to hand you a specific hours-saved number that wasn't actually measured for your business.
What is product performance analysis?
It's the process of evaluating individual SKUs across revenue, margin, return rate, and ad attribution, not just whether they sold well. The goal is understanding which products are actually profitable after real costs are factored in.
How often should you run a product performance analysis?
Monthly for catalogs with frequent SKU turnover or seasonal products. Quarterly is usually enough for a stable, slower-moving catalog.
What's the difference between product performance analysis and sales reporting?
Sales reporting tells you what sold. Performance analysis tells you what was actually profitable once COGS, fulfillment, returns, and ad spend are subtracted out. They can point in opposite directions for the same SKU.
Can you do product performance analysis in a spreadsheet?
Yes, for small catalogs. It tends to break down once you're matching SKUs across multiple channels, keeping COGS data current, or mapping ad spend by hand, which is exactly where most teams start looking for a dashboard instead.
What metrics matter most for Amazon vs Shopify product performance?
Amazon adds layers Shopify doesn't have: FBA fees, storage costs, and Buy Box share all eat into margin in ways that aren't obvious from the sales report alone. Shopify gives you more direct visibility into margin since you're not sharing the storefront with other sellers.
A one-off spreadsheet report works fine when your catalog is small. It stops working the moment SKU count grows, and most brands hit that wall faster than they expect.
The minimum viable version of this, regardless of catalog size, tracks contribution margin, sell-through rate, return rate, and ad-attributed revenue per SKU. Everything else is detail on top of those four.
If you're ready to see what a standing SKU-level view looks like without rebuilding it by hand every month, it's worth a look at how Trivas's BI reporting handles multi-channel analysis across Shopify and Amazon in one place. And if you want more on the channel side specifically, our Shopify and Amazon resources go deeper into the setup for each. Either way, subscribe to future posts here if you'd rather get this kind of breakdown in your inbox than go looking for it next quarter.
explore more insights
3 min read
3 min read
3 min read
.d53b12e5.png&w=3840&q=75 "Trivas.ai")