Ecommerce Performance Analysis: What to Track

Ecommerce performance analysis is the habit of checking which products, channels and campaigns are actually driving results, so you can do more of what works. The trap is tracking what looks busy, such as visits and orders, instead of what pays.

Start with outcomes, not activity

• Sales trend by channel and by product, over the periods that match how you buy and sell
• Net profit and margin, not just revenue
• Order volume, so you can see demand shift as it happens
• Refund rate, by product and by channel
• Advertising cost against the sales it produced

Read the numbers together

One metric rarely tells you anything. A product with rising sales and a rising refund rate may be losing money faster. A channel with falling orders and a falling ad budget may be fine. The useful questions come from pairing metrics:

• Sales up, margin down: is a discount or an ad campaign eating the gain?
• Refunds up on one product: is the cause stock quality, a listing that over-promises, or a carrier problem?
• One channel growing fast: does its margin hold at higher volume?

Compare channels side by side

If you sell on several marketplaces, looking at each one's own dashboard hides the comparison. A single view with every channel next to each other shows which earns its keep. Our /performance page shows how this works across channels.

Find what to scale

Performance analysis should end with a decision. Look for products with strong margin and room to grow, marketplaces where you sell well but list little, and listings that sell despite weak content. Those are the cheapest wins.

How often to review

• Daily: orders, sales and any alerts
• Weekly: channel and product profit, refund rate, ad performance
• Monthly: trends, the cost base, and which products to retire or push

Vanity metrics versus useful ones

• Vanity: visits, impressions and total orders on their own
• Useful: profit per order, refund rate by product, ad spend against profit, and channel margin
• The test: would this number change a decision you make this week?

An example

A seller sees orders up 20% month on month and calls it a good month. Looking closer, refunds rose from 4% to 9% on one product and a promoted listing consumed most of the extra margin. Orders were up, performance was not. Pairing metrics is what exposes that.

What is the single most useful metric?

Net profit per SKU by channel. It combines price, cost, fees, advertising and returns into one number you can act on.

How many metrics should I track?

Fewer than you think. Five to eight that you review regularly beat thirty that nobody reads.

For a ready-made weekly view, read /insights/ecommerce-performance-reports-what-to-include.