Ecommerce KPIs: Metrics and Benchmarks

Ecommerce KPIs should show whether your store is making progress toward a business goal. Start with purchase conversion, average order value, acquisition cost, repeat buying, and profit. Choose the measures that fit your current problem, then pair each with a safeguard against an expensive improvement.

A larger basket can come with lower profit. More returning customers can coexist with fewer new buyers. The useful question is what changed together, and what you should do next.

Which ecommerce KPIs should you track?

Track the numbers that could change your next decision. A metric describes something, such as traffic or order value. It becomes a key performance indicator when you give it a goal, an owner, and a review date.

For example, a store trying to grow baskets might choose AOV as its primary KPI. Profit left after the offer and shipping becomes its safeguard. That pairing helps the team decide whether a bundle deserves another test.

Revenue remains a useful outcome, but it does not explain why results changed. Follow the customer journey from visits to purchases, then to repeat buying. Keep supporting measures close to the stage where you can act.

Some measures flag trouble before the final result arrives. Fewer checkout completions may warn of a sales problem, while monthly revenue records the outcome. Treat an early signal as a reason to investigate, rather than a reliable prediction by itself.

Check demand and purchase conversion

Purchase conversion shows how often a visit leads to a purchase. Use your report's purchase definition consistently, and keep signups separate from sales. A successful email signup and a completed order answer different questions.

Traffic tells you how much demand reached the store. Read it by source and device alongside conversion. If a new campaign brings more browsing traffic, total conversion can fall even when existing channels remain steady.

Start with the part of the journey that changed. Fewer product visits suggest a demand problem. Plenty of product visits but fewer cart additions suggest checking availability, price, images, and product information.

Cart abandonment means shoppers added products but left without buying. Checkout abandonment starts later, after shoppers begin checkout. Keep those labels separate when deciding whether to test product pages or payment and delivery information.

Show shipping costs and delivery expectations before shoppers commit. If checkout completion drops suddenly, try the purchase journey yourself and check payment errors. Broad discounts are an expensive first response to a broken checkout.

Bounce rate and time on site provide context, rather than a sales verdict. A short visit can mean a shopper found an answer quickly. Compare the page's intended purpose with what visitors did next before deciding it needs a redesign.

Email click rates show whether recipients followed the offer into your store. Ad click rates show whether an audience responded to the message. Read both beside purchases and acquisition cost; more clicks alone do not show profitable growth.

Choose a basket target that fits your products

Choose an AOV reference that resembles what you sell. Average order value means the average spending per order. It is useful for basket decisions, but it does not tell you what you kept after costs.

Use the AOV shown in your sales report, and check which sales amounts it includes. Compare like amounts across periods; tax, delivery charges, and refunds can change what the displayed basket means.

The BTN industry references below cover a 90-day period ending October 7, 2026, across hundreds of Shopify stores. The chart shows the sample for each industry. The cosmetics midpoint is $50, compared with $77 for skincare and accessories.

These differences describe the stores shown. They do not tell you that a cosmetics store should raise prices until it matches skincare. Product price, pack size, replenishment habits, and customer mix can change the sensible target.

For a bundle test, check whether shoppers add useful products rather than substitute a discounted pack for their usual purchase. Read AOV beside contribution per order and purchase conversion. A higher basket can leave less money if the offer adds shipping or discount costs.

The industry AOV reference gives more product comparisons. Use it to choose context, then judge the test against your own earlier results.

Protect revenue with a refund safeguard

Pair sales growth with the revenue you retain after refunds. Refunded revenue and returned products are different measures. A partial refund can reduce sales without a product coming back, while an exchange may involve a return without the same loss of revenue.

The BTN refunded-revenue references cover a 90-day period ending October 7, 2026, across hundreds of Shopify stores. Each chart row includes its sample. The clothing and fashion midpoint is 2.5%, compared with 0.4% for food and drink.

The chart does not show why an individual store issued refunds. Investigate the products and reasons behind a change. Fit questions, damage, delivery problems, and inaccurate descriptions call for different fixes.

Give recent sales time to show their refunds before declaring an offer successful. A promotion can look stronger immediately after checkout than it looks after customers receive their orders. Compare sales, refunds, and retained revenue over a consistent review period.

For clothing, examine size guidance and the products driving the change. For food, examine damage and delivery complaints when those reasons appear. The next action should follow the evidence from your customers, rather than the category midpoint alone.

Keep repeat buyers separate from returning buyers

Choose a retention measure that matches the question you want answered. Repeat purchase rate asks whether customers bought more than once during a stated period. Returning customer rate asks whether current buyers had already made their first purchase before that period.

These are separate KPI labels, even when a dashboard places them together. A customer who first bought earlier can return once during the current period. That customer is a returning buyer, but has not bought twice during the period.

A rising returning customer share also needs context. It could accompany stronger repeat demand, or it could reflect fewer new customers buying. Look at customer counts and revenue from each group before increasing retention discounts.

For a replenishment store, check whether customers return around the time they are likely to need more. For durable products, a short repeat window may be a poor primary goal. Relevant accessories, referrals, or customer satisfaction may better reflect the relationship.

Customer retention asks whether an earlier customer group keeps buying over time. Lifetime value asks how much that relationship is worth. Neither is interchangeable with the share of today's buyers who are returning.

Our repeat purchase benchmarks explain longer purchase windows. Choose the window before setting your goal, and keep it consistent when reviewing progress.

Judge acquisition against profit and customer value

Judge acquisition by what it costs and what the resulting customers contribute. Customer acquisition cost, or CAC, is the spending associated with gaining a new customer. Agree whether your report includes ads alone or also creative, agency fees, and other acquisition expenses.

Read acquisition spending alongside the new customers gained during the same period. Keep returning buyers out of that customer count. If the report only shows advertising cost, label it clearly before comparing it with a broader acquisition budget.

A low acquisition cost can still be unattractive if customers place small orders and do not return. A higher cost may be affordable when contribution is strong and repeat purchases arrive soon enough. Your cash position matters alongside the eventual customer value.

Return on ad spend, or ROAS, describes revenue credited to advertising relative to its spend. Attribution rules affect which sales receive that credit. Compare campaigns using consistent reporting windows, and do not add platform revenue claims as though they were separate sales.

Marketing efficiency ratio, or MER, compares overall revenue with the marketing spending you have chosen to include. It provides a broader view, but returning customer revenue can make acquisition look stronger. Check new customer demand alongside it.

Revenue shows sales activity. Gross margin shows what remains after product costs, before other expenses. Contribution shows what remains after the variable costs you include, such as delivery, payment fees, and advertising.

Net profit goes further by including the remaining business expenses. Keep those cost labels visible, especially during promotions. A good revenue result does not automatically mean the offer improved profit or cash available for inventory.

Customer lifetime value, also called LTV or CLV, describes value over the customer relationship. Check whether your report shows revenue or profit, and whether it includes expected future purchases. Money a customer might spend later is different from money already received.

Use customer groups with similar time to buy again when comparing acquisition quality. A recently acquired customer has had less opportunity to return than an older customer. Avoid treating that difference as proof that the new campaign brought worse buyers.

Add service and stock checks when they change a decision

Add an operating measure when it explains a customer or profit problem. Customer satisfaction scores describe responses to a service or purchase survey. Recommendation scores ask how likely customers are to recommend the business; they answer a different question.

Response time measures the wait for an initial support reply. Resolution time measures how long the issue remains open. A faster first reply helps little when customers still wait for a missing order to be resolved.

For inventory, watch product availability when conversion falls. Stockouts can limit sales even if demand remains strong. Inventory turnover describes how quickly stock sells and is replaced, while delivery time shows how long customers wait to receive it.

Choose the operating check connected to your current problem. If refund complaints mention damage, review packing and delivery. If popular products are unavailable, review replenishment before paying for more traffic to those pages.

Review a small dashboard and assign the next action

Keep the dashboard small enough that a review ends with a decision. Show the current result, a useful earlier comparison, the goal, and the person responsible. Label the date range and metric definition so the team reads the same question.

A practical starting routine is to review traffic, purchases, ad spending, and operating problems weekly. Use a monthly review for customer value, repeat buying, and the outcome of completed tests. During a promotion or disruption, check the measures you can act on more often.

Match comparison periods to your selling pattern. A holiday launch and an ordinary week are different situations. Note promotions, stock shortages, and changes in channel mix before attributing a movement to your latest test.

Start a test with a written decision: keep the offer if its primary KPI improves without weakening the safeguard. For a bundle, that could mean a larger basket with acceptable contribution and conversion. Suppose a bundle raises basket spending but leaves less contribution after delivery. Pause the offer and revise its discount or contents before expanding it. Record the result and the next action instead of adding more dashboard tiles.

The ecommerce KPI dashboard guide covers assembling the view. By the Numbers supports shared and personal dashboards with cards you can pin, as described in its Shopify ecommerce reporting.

Use the ecommerce benchmark hub when you need more peer context. Your own trend tells you whether a change helped; a relevant benchmark helps you decide which question to investigate next.

How we measured these benchmarks

These benchmarks use anonymized, aggregated data from Shopify stores in the By the Numbers network. No individual store is identifiable from any figure. Stores that use an analytics app are not a random sample of all Shopify stores, so read the figures as a reference point, not as an average for the whole market.

Related guides

More on this topic from the By the Numbers team:

• Best Shopify Analytics Plan For Store Growth
• 7 Common Shopify Analytics Mistakes to Avoid
• Why You Need Custom Reports to Scale
• How To Use Shopify Data Without Feeling Overwhelmed

By the Numbers builds this into the dashboard. See Shopify ecommerce reporting.