Ecommerce metrics measure how an online store attracts buyers, converts visits into orders, keeps customers, and earns money after costs. The useful ones help you decide what to investigate or change.
A customer, an order, and a purchasing session are different units. Check which one your report counts before comparing dashboards.
The 22 metric and report topics below fall into four groups:
- Sales and conversion: revenue, conversion rate, average order value, revenue per visitor, cart abandonment, traffic sources, and bounce rate.
- Profit and acquisition: CAC, CLV, CLV:CAC, gross and net margin, product margin, and ROAS.
- Customers: retention, churn, repeat purchase rate, purchase frequency, and top customers.
- Inventory and products: inventory turnover, stock-out rate, return rate, and top-selling products.
What makes an ecommerce metric a KPI?
A KPI is a metric selected to measure progress toward a business goal. Conversion rate can be a KPI; reaching 2% by December is its target. Pick the measurement, define the desired result, and name the decision it will inform.
For the broader process of turning reports into decisions, read ecommerce analytics.
Which ecommerce metrics should you track first?
Start with the question your business needs to answer, rather than its age. A new store buying ads needs acquisition costs and contribution alongside sales. An established store with falling orders may need to inspect traffic and checkout before adding another retention report.
| Decision | Start with | Check alongside it |
|---|---|---|
| Are visits becoming purchases? | Purchasing-session conversion rate | Traffic source and tracking coverage |
| Can first orders cover acquisition? | CAC and first-order contribution | Discounts and fulfilment costs |
| Are customers buying again? | Repeat purchase or cohort repurchase | Purchase window and customer history |
| Which products deserve investment? | Net sales and product contribution | Returns and stock availability |
Keep the first dashboard small enough that someone will review it and act.
Which ecommerce metrics explain sales and conversion?
Start with the amount sold and the activity counted. People, sessions, and orders produce different rates.
Illustrative Putler Sales view with historical demo data. Sales documentation.
1. Revenue
Revenue measures sales before the expenses used to calculate profit. For the example below, net merchandise sales equal gross merchandise sales minus discounts and refunds; tax and shipping collected are excluded.
Check gross sales and net sales before comparing reports. A payment deposit can also include fees or settlement timing that this sales measure doesn’t describe.
2. Conversion rate
A purchasing-session conversion rate measures the share of storefront sessions containing a purchase.
Formula: Purchasing sessions / relevant storefront sessions × 100.
Use a matching source and period for both counts. Dividing backend orders by GA4 users produces a different measure; multiple orders in one session also differ from one purchasing session. Specify the rate before judging it.
3. Average order value (AOV)
AOV describes the average value of an order, using the report’s stated revenue basis.
Example basis: (Gross merchandise sales − discounts) / orders, before post-order refunds.
Shopify’s sales definitions specify adjustment exclusions. If you deduct refunds instead, label the result net revenue per order and explain why it differs. Check order value alongside contribution before encouraging bigger baskets.
4. Revenue per visitor (RPV)
RPV measures revenue per unique visitor within a stated period.
Formula: Defined revenue / unique visitors.
Revenue per session uses sessions instead. Keep these labels separate, particularly across stores where the same person can visit both. Our worked example provides storefront sessions, so it calculates revenue per session; a unique visitor count cannot be inferred from its buyers.
5. Cart abandonment rate
Cart abandonment measures the share of a defined cart population that does not result in a purchase.
Session-based version: Sessions with a cart addition but no purchase / sessions with a cart addition × 100.
A cart-based cohort or checkout-abandonment report uses different inputs. Check the measurement window and where shoppers stop before selecting a recovery tactic; some visitors are comparing options or saving items for later.
6. Traffic sources
Traffic-source reporting groups visits by origin, such as organic search, paid ads, email, or referrals. It is a breakdown rather than one KPI.
Compare conversion and order value within a consistent channel definition. A session-acquisition report and an ad platform’s attributed-sales report can credit activity differently. Check their scope before shifting spend because one channel appears to win.
7. Bounce rate
GA4 bounce rate is the share of sessions that were not engaged.
Formula: Non-engaged sessions / sessions × 100.
An engaged session lasts longer than 10 seconds, has a key event, or has at least two views, under Google’s documented definition. A single-page session can therefore be engaged. Check event settings and traffic mix before interpreting a change.
How should you track profit and acquisition?
Check what sales leave after costs before increasing spending. Revenue-based returns alone cannot establish profitability.
8. Customer acquisition cost (CAC)
CAC measures the defined cost of acquiring a new customer.
Formula: Acquisition costs / new customers acquired.
State whether costs include only ads or also agency work, creative, and staff time. Align the measurement period and attribution approach. The fictional example includes acquisition ads and labour; dividing ad spend by new customers would understate its chosen CAC definition.
9. Customer lifetime value (CLV)
CLV estimates customer value across a relationship. An observed customer-revenue total and a forecast of future contribution are different readings.
Simple revenue model: AOV × purchases per year × relationship length in years.
Match the time units and state assumptions. A month of purchases doesn’t establish customer lifespan. Use customer lifetime value methods that fit the data you actually have.
10. CLV-to-CAC ratio
This ratio compares the chosen customer-value measure with acquisition cost.
Formula: CLV / CAC.
A revenue-based ratio omits the cost of supplying those sales. Check contribution and the time needed to recover acquisition spending before scaling. A universal 3:1 target cannot account for every store’s margin, repeat-purchase pattern, or cash constraints; a high ratio alone doesn’t prove underinvestment.
11. Gross and net profit margin
Gross margin measures the share of net sales left after cost of goods sold (COGS).
Formula: (Net sales − COGS) / net sales × 100.
Contribution deducts additional named variable costs, such as payment fees and fulfilment. Net profit deducts the remaining business expenses and applicable taxes.
Net profit margin formula: Net profit / net sales × 100.
Specify which costs a report includes; the sample’s contribution after marketing is still not net profit.
12. Product margin
Product gross margin applies the same cost check to a product or SKU.
Formula: (Product net sales − product COGS) / product net sales × 100.
Shopify’s profit reports depend on recorded product costs for the relevant measures. Include fees, fulfilment, and other variable costs when calling a result product contribution. Review returns before giving a high-revenue product more exposure.
13. Return on ad spend (ROAS)
ROAS compares attributed revenue with advertising spend.
Formula: Revenue credited to ads / ad spend.
Name the attribution model, window, and revenue treatment. ROAS excludes costs outside its denominator, so it isn’t a profit calculation or proof the ads caused every credited purchase. Compare the resulting contribution with acquisition costs before increasing the budget.
Which customer metrics show repeat buying?
Keep full-history customer labels separate from buying activity within a selected period.
Illustrative Putler Customers view with historical demo data. Customer-report documentation.
14. Customer retention rate
Retention measures whether a defined starting customer population remains active under an explicit rule.
Non-subscription example: Prior-cohort customers who buy again in the window / eligible prior-cohort customers × 100.
Call this cohort repurchase when that is what you’re measuring. An active-customer formula needs compatible starting and ending populations. Choose a window that suits the buying cycle; see customer retention metrics for deeper treatment.
15. Churn rate
Churn measures customers lost from a defined population during a period.
Formula: Customers classified as lost / customers in the starting population × 100.
A cancelled subscription supplies different evidence from a shopper who hasn’t reordered yet. For non-subscription stores, state the inactivity rule. Churn and retention add to 100% only when their populations and definitions are complementary.
16. Repeat purchase rate
A same-period repeat purchase rate measures the share of active buyers who place at least two orders during that period.
Formula: Buyers with two or more period orders / active period buyers × 100.
This differs from an order share and from a history-based returning-customer classification. Shopify’s customer reports explain how wider order history can affect customer labels. Keep the rule visible when comparing tools.
17. Purchase frequency
Purchase frequency measures how often an active customer orders within the specified period.
Formula: Orders / unique active customers.
Use consistent rules for cancelled, refunded, and zero-value orders. A monthly frequency cannot be inserted unchanged into an annual CLV model. Compare suitable customer groups and buying cycles before setting reorder timing; furniture and replenishable goods rarely share a useful universal frequency target.
18. Top customers
A top-customer report ranks buyers by a chosen measure, such as net spend, frequency, or recency. RFM combines recency, frequency, and monetary value to describe purchasing patterns.
Choose a ranking that supports the action. A high spender with a recent refund may need support before another offer. Check the actual revenue distribution instead of assuming that 20% of customers always produce 80% of sales.
Which inventory and product metrics protect availability and margin?
Read sales rankings alongside stock, costs, and returns before changing purchasing quantities.
19. Inventory turnover ratio
Inventory turnover measures how often the cost value of inventory is sold through over a period.
Formula: Period COGS / average inventory at cost.
Keep both inputs on a compatible cost basis. Annual and monthly turnover figures aren’t directly comparable without adjustment. Review seasonality and lead times before changing purchasing quantities; a universal target can leave one category overstocked and another unavailable.
20. Stock-out rate
Stock-out rate describes unavailability, but the unit needs to be explicit.
One-SKU time measure: Unavailable tracked days / total tracked days × 100.
Across SKUs, a SKU-day measure is another option. Neither automatically measures the share of demand lost. Check availability for the products shoppers wanted, alongside lead time and carrying cost, before adding a fixed percentage of buffer stock.
21. Product return rate
A unit-based product return rate measures returned items against an eligible sold or shipped population.
Formula: Returned units / eligible units × 100.
Match the cohort and allow its return window to mature. A goodwill refund without a returned item affects revenue but doesn’t create a physical return. NRF’s 19.3% online-sales estimate for 2025 came from large-US-merchant research; it isn’t a 2026 unit-return target.
22. Top-selling products
Top-selling product reports rank sales by units or defined revenue. The result identifies demand, not necessarily profit.
Compare each leader with product contribution, return patterns, and availability before changing placement or stock. Product relationships can also reveal frequently bought products worth considering together. Check the economics of any bundle rather than assuming a larger order leaves more money.
What do these numbers show in one two-store example?
This fictional dataset contains six paid orders across two stores during September 2026. Its four active buyers include two merchant-verified customer matches across stores. The example shows calculation choices; it is not a real merchant result, a benchmark, or a software export.
| Order | Store / buyer | Net merchandise revenue | Contribution before marketing |
|---|---|---|---|
| O1 | A / C1 | $100 | $50 |
| O2 | A / C2 | $80 | $40 |
| O3 | A / C3 | $60 | $28 |
| O4 | B / C1 | $80 | $34 |
| O5 | B / C2 | $120 | $60 |
| O6 | B / C4 | $50 | $22 |
Gross merchandise sales of $550, less $50 discounts and a $10 goodwill refund, leave $490 net merchandise revenue. Deducting $200 COGS, $16 payment fees, and $40 fulfilment leaves $234 contribution before marketing. Taxes and shipping collected are excluded; the refund involved no physical return or cost recovery.
C1 and C2 bought before September; C3 and C4 are new buyers. The figures answer different questions:
- Order value: $500 before post-order refunds / 6 orders = $83.33; $490 after that refund / 6 = $81.67 net revenue per order.
- Repeat buyers: 2 buyers with two September orders / 4 active buyers = 50%. Returning customers placed 4 of 6 orders, an order share of 66.7%.
- Acquisition: ($80 acquisition ads + $20 acquisition labour) / 2 new buyers = $50 CAC.
Those new buyers generated $110 revenue and $50 contribution before acquisition. Crediting their orders to the $80 ad spend gives 1.375x ROAS, yet their contribution falls $50 short of the fully loaded acquisition cost. Later purchases are unknown, so this doesn’t establish lifetime unprofitability.
Another $120 went to retention ads. Total contribution after the $220 marketing cost is $14, before fixed costs and taxes.
The stores recorded 600 and 400 storefront sessions. Net revenue per storefront session is $490 / 1,000 = $0.49. These summed sessions are not unique visitors, so this is not RPV.
Download the fictional sample data and assumptions.
Which ecommerce benchmarks are useful?
Useful benchmarks disclose their population, period, and measurement rule. Compare a similar business and tracking setup before judging performance against them. When those details differ, your own consistently measured baseline is often a better starting point than a general poor, average, or good label.
| Reference | Observed figure | Scope and limitation |
|---|---|---|
| Littledata conversion benchmark | 1.4% median; top 20% above 2.7% | 433 Shopify stores using Littledata and GA4; 20 July-17 October 2026 |
| Baymard cart abandonment | 70.22% | Pooled across 50 studies from different periods; not one fresh 2026 cohort |
Littledata’s current sample, generated 21 October 2026, uses purchasing sessions divided by sessions and requires at least 1,000 sessions and 90 purchases per store. Our six-order example is far below that purchase minimum; don’t use the table to grade it.
Baymard’s research separately reports extra costs as a reason cited by 40% of its relevant survey group after excluding browsing/not-ready respondents. That isn’t a claim that displaying shipping costs will recover 40% of abandoned carts.
Why can ecommerce reports disagree?
Reports can count different activity, dates, and revenue. Check their definitions before diagnosing performance.
Google Analytics metric settings. Source: Google Analytics Help.
Check these five things before comparing:
- Unit: sessions, unique visitors, orders, customers, or items.
- Population: storefront activity, all sales channels, new buyers, or a prior cohort.
- Amount: treatment of discounts, refunds, tax, shipping, and costs.
- Timing: timezone, order date, refund date, and attribution window.
- Collection: consent, tracking coverage, bot filtering, and measurement changes.
Shopify documents a session-measurement rollout on 21-23 September 2026. Annotate changes like this before interpreting a new conversion rate as a campaign improvement.
For a simple fictional illustration, six purchasing sessions divided by 1,000 sessions is 0.6%; the same six divided by 600 is 1%. The purchase count hasn’t increased. This is denominator arithmetic, not a measured effect of Shopify’s update.
How should you build a useful ecommerce metrics dashboard?
Choose sources for the decision: store reports for orders and customers, web analytics for visits, advertising reports for attributed results, and reliable cost and inventory records.
Match the source to the question
Native Shopify reports already include customer cohorts, RFM, and cost-based profit views where the required inputs exist. A separate tool should solve an actual coverage or workflow gap.
Putler can help combine sales and customer analysis across supported connected sources. Its transaction reporting provides a place to inspect the underlying records; Time Machine also documents revenue and customer scenarios based on model inputs.
That doesn’t establish a COGS-based product-profit engine or remove the need for financial checks. In our example, advertising, product costs, and fulfilment are supplied external inputs. Confirm the data you need before choosing another tool.
Set a useful review cadence
Check daily for operational exceptions, such as missing orders or unavailable bestsellers. Review conversion and acquisition when the available volume supports a comparison, and use longer windows for repeat buying, CLV, and inventory patterns.
Name an owner and next action for each selected KPI. For layout guidance, use the ecommerce dashboard guide.
Where should you start?
Choose one decision this week, such as whether to increase acquisition spend. Write down the metric’s amount, unit, population, and period, then check the costs or customer history that could change the answer.
If connected-store reporting is the missing piece, explore Putler and check its supported integrations against your sources.
FAQ
What are ecommerce metrics?
Ecommerce metrics measure store performance across sales, acquisition, customer behavior, and operations. Examples include conversion rate, AOV, CAC, repeat purchase rate, and inventory turnover. A useful definition states what is counted, which population is included, and the period, so the result can support a decision.
What is the difference between ecommerce metrics and KPIs?
A metric is a measurement. A KPI is a metric selected to track progress toward a business goal. Conversion rate might be the KPI for improving checkout performance; reaching a specified rate by a specified date is the target. Keep the calculation consistent when checking progress.
Which ecommerce metrics should I track first?
Start with the decision and the evidence it requires. Conversion and order value help explain sales; acquisition cost and contribution help assess whether first orders cover spending. Repeat buying, returns, and availability add context. Store age alone doesn’t justify delaying cost checks or tracking every available measure.
What is a good ecommerce conversion rate?
A useful comparison matches platform, industry, audience, tracking, and period. Littledata’s checked July-October 2026 sample of 433 Shopify stores had a 1.4% median purchasing-session rate, with its top 20% above 2.7%. Those figures describe that sample; they aren’t universal targets or a reliable score for a tiny new store.
What is a good CLV-to-CAC ratio?
There isn’t one safe target for every store. State whether lifetime value represents revenue or contribution, how it is estimated, and which acquisition costs are included. Then check the expected payback period and cash needs. A revenue ratio can look attractive while the costs of serving those purchases consume the difference.
What is the average ecommerce return rate?
The answer depends on the year, product category, population, and definition. NRF and Happy Returns estimated returns at 19.3% of online sales for 2025, based on large-US-merchant research. That figure isn’t a 2026 actual or automatically comparable with your returned-units rate. Allow the relevant return window to mature.
Why do ecommerce dashboards show different numbers?
Dashboards can use different populations, timezones, attribution windows, and revenue treatments. They may also collect different activity because of consent or tracking coverage. Compare the underlying definitions and records before deciding one report is wrong; an order count, purchaser count, and purchasing-session count need not match.