An online shop produces a great many numbers, but a few metrics are enough to tell whether the business is heading in the right direction. The problem is that each one, read on its own, can mislead. A rising conversion rate can hide smaller orders; growing revenue can come with more returns. It pays to know them well and always read them as a group.
Conversion rate
This is the ratio of orders to visits, or of orders to visitors: the choice of denominator must be stated, because it changes the result. Per visit, it measures how often a pass through the site turns into a purchase; per visitor, it measures how many people become customers in a period.
It is sensitive to the kind of traffic. A campaign that brings many browsers lowers average conversion even if sales go up. That is why it should always be read by channel and by device type, never only as an overall number.
Average order value
This is revenue divided by the number of orders. It shows how much each customer spends per purchase on average. It rises with suggestions of complementary products, free shipping thresholds and multipacks.
Beware of the average: a few very large orders can shift it a lot. Adding the median, or looking at the distribution by value band, avoids misreadings.
Revenue per visit
This is perhaps the most compact metric: revenue divided by visits, or equivalently conversion rate multiplied by average order value. It combines both effects and solves the problem of reading them in isolation.
| Period | Visits | Orders | Conversion | Avg order value | Revenue per visit |
|---|---|---|---|---|---|
| Month A | 20,000 | 400 | 2.0% | €60 | €1.20 |
| Month B | 20,000 | 460 | 2.3% | €50 | €1.15 |
In month B conversion went up, but each visit is worth less. Without revenue per visit, month B would look better.
Repeat purchase and lifetime value
A customer who comes back is worth more than one who buys once. The repeat purchase rate measures the share of customers who place at least a second order within a set period, for example twelve months. Customer lifetime value estimates the total revenue generated by a customer over the whole relationship.
Both require linking orders to the same customer, which is usually done by the shop's management system rather than the analytics tool. To estimate them honestly, use cohorts: customers acquired in the same month, followed over time.
Returns and cancellations
Revenue recorded at the moment of purchase is not real revenue. Returns, refunds and cancelled orders must be subtracted, and it is worth measuring their rate by product category. A product with many sales and many returns may be a description or sizing problem, not a success.
Margin
Revenue says how much comes in, not how much stays. Where possible, link orders to margin by product or category. A campaign that mostly sells discounted items can increase revenue while reducing profit. Shipping costs and payment fees matter too, and often vary a lot from one order to the next.
Cart abandonment
The share of carts created that never become orders is a useful indicator, but it needs breaking down. Many people use the cart as a wish list or to find out the shipping cost. The most valuable information is where abandonment happens: at the shipping cost calculation, at the registration request, at the choice of payment.
A practical way to read it is to measure each checkout step separately as a small funnel: cart, shipping details, shipping method, payment, confirmation. If the biggest loss happens when shipping costs appear, showing them on the product page reduces the final surprise. If it happens at the request to create an account, guest checkout is the most obvious hypothesis to test.
A consistent reference
Definitions should be written down and shared: whether revenue includes VAT, whether it includes shipping, whether cancelled orders are excluded, which time zone is used. Analytics and the management system rarely match to the cent; when the gap exceeds a few percentage points, it is worth investigating the tracking.
In short
- Always read conversion and average order value together.
- Revenue per visit combines both effects in a single number.
- Repeat purchase and lifetime value are best measured with cohorts and order system data.
- Subtract returns and cancellations and, where possible, think in terms of margin.
- Write shared definitions and regularly reconcile analytics with the order system.
