Most dashboards fail in one of two ways. Either they show too little, such as revenue and orders, and hide the problems that cost money, or too much, with dozens of charts that no one reads. A good store dashboard sits between them: about ten numbers, defined clearly, compared with the past and tied to decisions.
This guide proposes ten, explains how to define each, where the data comes from and what to do when it moves. The figures that depend on cost and profit connect to ecommerce profit tracking and customer lifetime value, and the reporting method to sales reporting. Definitions differ between platforms, so write down yours.
The ten numbers
| # | KPI | Definition | Why it matters |
|---|---|---|---|
| 1 | Net revenue | Sales after discounts, refunds and returns, excluding tax and shipping charged | The honest top line |
| 2 | Orders | Count of paid orders in the period | Volume, and the base for many ratios |
| 3 | Average order value | Net revenue divided by orders | Lever for profit; watch with margin |
| 4 | Conversion rate | Orders divided by sessions | Effectiveness of the site and traffic quality |
| 5 | Gross margin | (Net revenue minus cost of goods, shipping and payment fees) divided by net revenue | What is left to pay for everything else |
| 6 | Customer acquisition cost | Marketing spend divided by new customers | What it costs to win a customer |
| 7 | Repeat purchase rate | Share of customers with more than one order in a window | Loyalty and lifetime value |
| 8 | Refund and return rate | Refunded orders or value divided by orders or value | Product, listing and fulfilment quality |
| 9 | Fulfilment time | Average and 90th percentile time from order to dispatch | Customer experience and cancellations |
| 10 | Support contacts per 100 orders | Tickets divided by orders, times 100 | Operational friction |
Notice that three of the ten are about money kept, not money taken: margin, acquisition cost and refunds. A store can grow revenue while losing money, and a revenue-only dashboard will applaud right up to the cash crunch.
Define each metric in writing
Most arguments about numbers are really arguments about definitions. Does revenue include shipping? Is a session counted per visit or per day? Are returns subtracted in the month of the sale or the month of the refund? Does a customer who orders under two emails count once? Decide, write it down and apply it everywhere. A short metrics dictionary prevents the situation where the platform, the ad dashboard and the accountant each report a different figure.
| Item | Decision to record |
|---|---|
| Revenue | Gross or net; tax, shipping, gift cards and discounts |
| Date basis | Order date or payment date; time zone |
| Customer identity | How to match repeat customers |
| Sessions | Source (analytics tool) and bot filtering |
| Cost of goods | Which cost method; handling of bundles |
| Attribution window | How marketing spend maps to new customers |
| Exclusions | Test orders, wholesale, internal orders |
1 to 3: revenue, orders and order value
These are the familiar trio. Watch orders and average order value together: revenue can rise because orders rise, because value rises or because of a promotion that boosted both for a week. A fall in orders hidden by a rise in value, or the reverse, tells different stories. Compare with the previous period and the same period last year to account for seasonality. If you sell on several channels, show them separately as well as combined.
4: conversion rate
Conversion is orders divided by sessions, and it moves for reasons that have nothing to do with your site: traffic mix, promotions, device mix and bot traffic. Break it down by device and source before drawing conclusions. A mobile conversion rate that is well below desktop usually points to checkout or page-speed issues; a drop after a theme or plugin update calls for a technical check. Use conversion as a prompt to investigate, not as a score.
5: gross margin
Margin tells you what you really earn. Include the cost of the goods, shipping you pay, payment fees and ideally returns allowances. Track it overall and by product and channel. A product with strong sales and thin margin may deserve a price rise or a cost review. Link it with payout reconciliation so that the numbers match the bank.
6: customer acquisition cost
Total marketing spend divided by new customers, ideally excluding returning customers who would have bought anyway. Compare with first-order margin and with lifetime value to see how quickly a customer pays back. If acquisition cost rises while conversion and order value are flat, the channels are saturating, and it is time to test elsewhere or improve retention.
7: repeat purchase rate
The share of customers who buy again within a fixed window, such as 90 or 180 days, is the simplest retention measure. Track by cohort, as in the CLV guide. A stable or improving repeat rate suggests that products and service satisfy; a falling one points to quality, expectations or a changing mix of new customers.
8: refund and return rate
Measure by value and by count, overall and by product. A single product with a high rate is a listing or quality problem, a rise across many suggests a carrier or packaging issue. Split reasons: wrong item, damaged, not as described, changed mind. See returns automation for how to capture reasons properly.
9 and 10: fulfilment time and support contacts
These are your operational health checks. Average dispatch time hides late orders, so also show the 90th percentile or the share dispatched within the promised time. Support contacts per hundred orders reveal friction; a jump after a change points to its cause. Split by topic: order status, product questions, returns and payments. Both numbers link to automation: better messages reduce contacts; better routing reduces fulfilment time. See where is my order and order status update messages.
Layout and rhythm
- One page. Ten tiles with the value, the change versus the previous period and a small trend line.
- Colour for meaning. Use colour only to flag out-of-range values.
- Weekly review for the whole set; daily glance at orders, fulfilment and tickets.
- Monthly deep dive by channel, product and cohort.
- Named owner for each tile and a note on the action when it moves.
Where the data comes from
| KPI group | Sources |
|---|---|
| Sales, orders, AOV, refunds | Store platform reports or an export |
| Conversion and sessions | Web analytics |
| Margin and costs | Platform costs, supplier invoices, payment and carrier statements, accounting |
| Acquisition cost | Ad platforms and email costs, with new customer counts |
| Repeat rate | Customer and order data |
| Fulfilment and tickets | Shipping software, helpdesk |
Pull everything into one place automatically: a spreadsheet fed by scheduled imports, a reporting tool or a small database. Automation matters here because a dashboard updated by hand soon goes stale. Our sales reporting and finance, risk and reporting work builds such dashboards with the definitions written in.
Who should see which numbers
A dashboard is for decisions, and different people make different ones. Give the owner the whole page of ten numbers each week. Give the marketing lead acquisition cost, conversion by source and email revenue, daily if they want it. Give operations the fulfilment time, late orders and stock cover. Give support the contacts per hundred orders and the top reasons. Keep each view short and focused, with the same definitions underneath. When everyone sees the same numbers, meetings shorten, because the argument about whose figure is right disappears. Agree a regular review rhythm and write down the decisions made after each one, so that the dashboard becomes a record of what the team learned, not just a picture of the past.
A worked example
A store reviews its dashboard and sees revenue up 12 percent on last year, but margin down from 41 to 36 percent, refund rate up from 5 to 8 percent and support contacts per 100 orders up from 9 to 14. The pattern points to a new product line with sizing problems: returns and tickets rose together, and the lower margin comes from free return shipping. The store adds a size guide, adjusts the product titles and offers an exchange instead of a refund. Within two months the refund rate falls to 6 percent and margin recovers to 39 percent. Revenue alone would have shown only the growth. The figures are illustrative.
Traps
- Vanity metrics. Sessions and followers do not pay bills.
- Mixed definitions between tools.
- No comparison. A number alone says little.
- Too many tiles. If everything is highlighted, nothing is.
- Averages that hide the tail in fulfilment times.
- No decision attached. Each tile should answer: what would we do if this moved?
Sources and further reading
Figures and rules in this guide come from the primary pages below. Platform limits and prices change, so open the page before you build on a number.
- Shopify Help Center: reports and analytics, what Shopify reports and how it defines them
- Google Analytics Help: ecommerce reports, ecommerce measurement in GA4
- Stripe Docs: reports, financial reports for payments