E-Commerce Metrics That Matter: The 15 KPIs to Track, With Formulas and 2026 Benchmarks

Every e-commerce platform ships with a dashboard full of numbers, and most of them move every day. The skill is not finding metrics. It is knowing which few answer the question you are asking, how each is calculated, and what a sensible value looks like for a store like yours.

This guide groups fifteen KPIs by the question they answer. It opens the Market Metrics section of the Kurums E-Commerce pillar. If the vocabulary is new, the short versions live in E-Commerce Decoded.

Which e-commerce metrics actually matter?

Five metrics explain most of a store's performance: conversion rate, average order value, customer acquisition cost, repeat purchase rate and contribution margin. Together they describe how many visitors buy, how much they spend, what they cost to win, whether they return and whether each order makes money.

Revenue is simply their product. Sessions multiplied by conversion rate gives orders; orders multiplied by average order value gives revenue. To grow revenue you can only raise one of three things, and each has a different cost.

What is a good conversion rate, and why do benchmarks disagree?

Conversion rate is orders divided by sessions. Credible 2026 benchmarks sit between roughly 1.4% and 3%: IRP Commerce reported 1.70% for April 2026, Dynamic Yield's global benchmark was 2.66%, and Littledata put the average Shopify store at 1.4%, with the top fifth above 3.2%.

The spread is not an error. Each dataset covers different stores. IRP's panel is weighted to UK small and mid-sized merchants. Dynamic Yield's customers are larger brands with known names. Littledata measures Shopify stores of every size. Category matters even more: food and consumables convert several times better than furniture or luxury goods.

Three rules for reading your own number:

• Segment before you judge. Split by device, traffic source and new versus returning visitors. A blended 2% can hide 4% from email and 0.6% from paid social.
• Watch the steps, not only the total. Dynamic Yield's global add-to-cart benchmark is about 6%; Littledata reports average checkout completion on Shopify at 45%. If your add-to-cart rate is healthy and checkout completion is poor, the problem is the checkout, not the product page.
• Do not chase conversion with discounts alone. A sitewide 20% off raises conversion and lowers the margin on every order, including the ones you would have won anyway.

How should you read cart abandonment?

Cart abandonment is the share of started carts that do not become orders. Baymard Institute's average of 50 studies is 70.22%, and it has barely moved in a decade. Seven in ten is normal; it is a baseline to improve from, not an emergency.

Much abandonment is window-shopping. The recoverable part comes from friction. Baymard's research consistently finds unexpected extra costs at checkout, mainly shipping and taxes, as the leading stated reason, followed by forced account creation and long or complicated forms. Baymard estimates that the average large store could lift conversion by about 35% through checkout design alone.

Practical actions, in order of payoff: show delivery cost and time on the product page; offer guest checkout; add the wallets your market uses; then set up an abandoned-cart email sequence. Mobile needs separate attention, because abandonment on phones runs well above desktop in every dataset.

Why is average order value the quiet lever?

Average order value is revenue divided by orders. It matters because most per-order costs are fixed: picking, packing, postage and the cost of acquiring the click do not double when the basket does. A higher AOV drops almost straight through to contribution margin.

The ways to raise it are well known: free-shipping thresholds set slightly above the current average, bundles, quantity breaks, and relevant add-ons at the cart, a tactic explained in Upsell / cross-sell, Decoded. Benchmarks are of limited use here because AOV is set by category and country. For orientation, IRP Commerce's panel showed an average order of about £127 in mid-2026, while ECDB estimates Amazon's average order at around US$57: two healthy businesses with very different baskets.

Read AOV alongside revenue per session (conversion rate multiplied by AOV). A change that lifts AOV and cuts conversion can leave you worse off; revenue per session tells you the net effect in one number.

How do CAC, ROAS and MER fit together?

Customer acquisition cost is total sales and marketing spend divided by the number of new customers won. ROAS is revenue attributed to ads divided by ad spend. MER, the marketing efficiency ratio, is total revenue divided by total marketing spend. CAC tells you the price of growth; ROAS and MER tell you whether the spend pays back.

They disagree with each other for a reason. Platform-reported ROAS depends on attribution: each ad network claims the sales it touched, so the sum of platform ROAS usually exceeds reality. MER ignores attribution entirely, which makes it cruder and more honest. Use platform ROAS to compare campaigns inside one channel, and MER to judge the whole budget.

The break-even ROAS is the number to know by heart. It is 1 divided by your contribution margin before advertising. If 40% of each sale is left after product cost, fees and shipping, break-even ROAS is 2.5: below that, the campaign loses money on the first order.

A longer treatment, including blended versus paid CAC, is in e-commerce unit economics. The plain-English version is CAC & LTV, Decoded.

What do repeat purchase rate and LTV tell you?

Repeat purchase rate is the share of customers who have ordered more than once. Customer lifetime value is the margin a customer generates over the whole relationship. Together they tell you whether you are building a customer base or renting one-off transactions.

Calculate LTV conservatively:

1. Use contribution margin per order, not revenue.
2. Use observed orders per customer over a fixed window, typically 12 months for a young store, not a projected "lifetime".
3. Work by cohort: customers first acquired in the same month. Blended averages flatter you, because loyal early customers mask weaker recent ones.

The classic health check is the LTV to CAC ratio, with 3 to 1 as the conventional target. It is a rule of thumb borrowed from software. In e-commerce, where cash is tied up in stock, the payback period is at least as important: how many months until a customer's cumulative contribution covers what you paid to acquire them. Under six months is comfortable for most stores; over twelve needs outside funding.

Retention tools earn their cost here. See the Kurums comparisons of loyalty software and email marketing software.

Which profitability metrics should you check every month?

Three: gross margin, contribution margin and return rate. Gross margin shows whether the product is priced correctly. Contribution margin shows whether the order is worth taking after fees, shipping and marketing. Return rate shows how much of your reported revenue is temporary.

Returns deserve more attention than they get. A returned order costs outbound shipping, return shipping, handling, and often a markdown on the item, while the acquisition cost is already spent. Rates vary hugely by category; fashion runs far higher than consumables. Report net revenue (after returns and discounts) wherever you can, and measure return rate by product, since a handful of items with sizing or description problems usually drive most of it.

The accounting side, including how cost of goods sold is built up, is explained simply in Cost of goods sold, Decoded.

What operational metrics protect the customer experience?

Four operational numbers predict reviews and repeat orders: order-to-ship time, on-time delivery rate, stock-out rate and inventory turnover. They rarely appear on marketing dashboards and they decide whether marketing spend is wasted.

• Order-to-ship time: hours from payment to carrier handover. Marketplaces rank and penalise sellers on it.
• On-time delivery rate: share of orders arriving within the promise shown at checkout.
• Stock-out rate: share of sessions or days in which a best-seller is unavailable. Advertising an out-of-stock product is the most expensive mistake in e-commerce.
• Inventory turnover: cost of goods sold divided by average inventory. Low turnover means cash asleep on shelves.

If you sell on marketplaces, add the platform's own seller scores (defect rate, late shipment rate, valid tracking rate). They are thresholds, not targets: fall below and listings are suppressed.

How do you build a KPI dashboard you will actually use?

Pick one owner and one definition for each metric, then split the list by rhythm. A weekly view of five numbers and a monthly view of the rest is enough for most stores under a few million in revenue.

Write the definitions down. "Conversion rate" calculated on sessions in one tool and on users in another will produce two numbers and one argument. Decide whether revenue includes tax and shipping, whether returns are netted off, and which tool is the source of truth for each figure.

Once the internal numbers are steady, look outward. Market size, penetration, payment habits and logistics differ sharply by country, and they set the ceiling for every metric above. That is the subject of the market metrics to check before entering a country.

Frequently asked questions

Keep reading in the E-Commerce pillar