Open the analytics in almost any ecommerce tool and you get the same set of numbers: conversion rate, average order value, repeat purchase rate, customer lifetime value and a cohort chart. They are good numbers. They were designed for a store that sells to thousands of people, most of whom buy once.
A wholesale store is a different business. It might have forty accounts. Most of them order every few weeks, on terms, and ten of them carry half the revenue. Point consumer analytics at that store and the charts still render. They just answer questions nobody running the business is asking, and they stay quiet about the one that matters: which account is about to stop buying.
This post goes through the standard metrics one at a time, shows what each one does to a forty-account wholesale business, and then covers the three things worth measuring instead.
The standard metrics, one by one
Conversion rate
Conversion rate measures how many anonymous visitors become buyers. A trade customer logs in, places the order they place every month, and leaves. Plenty of wholesale orders never start with a website visit at all: they arrive by email or phone and someone on your team keys them in.
On a trade store, conversion rate mostly measures how often your existing customers check prices. It goes up and down for reasons that have nothing to do with sales.
Average order value
In a consumer store, average order value is a lever you pull with bundles, free shipping thresholds and upsells. In wholesale, order size is set by the customer's own stock levels and shelf space.
A falling average can mean an account now orders more often in smaller drops, which is fine. It can also mean the account has moved half its range to another supplier, which is not. The store-wide average cannot tell those two apart, and it is pushed around by your three largest accounts.
Repeat purchase rate
For a consumer brand, the share of customers who come back is the whole question. In wholesale, nearly everyone comes back. That is what an account is.
A repeat purchase rate of 95 percent looks healthy and tells you nothing. The account that stopped ordering in August is still counted as a repeat customer, because it did repeat, right up until it stopped.
Customer lifetime value
Lifetime value is a forecast built from averages across a large population. With forty customers there is no population. The average is set by two or three accounts, and the forecast for any single account is a guess presented as a number.
You do not need a model to tell you what a trade customer is worth. You know what they bought last year, and you know them by name.
Cohort retention
Cohort charts group customers by the month they first bought and track how many are still active. They work when each cohort holds hundreds of people.
A wholesale store might have won two new accounts in March. If one of them goes quiet, that cohort drops from 100 percent to 50 percent, and the chart cannot tell you whether that is a lost customer or a seasonal pause.
What these metrics have in common
They describe customers in aggregate, and they treat every customer as roughly equal. In wholesale both assumptions fail. Your customers are few enough to know individually, and unequal enough that losing one can matter more than everything the charts show.
A simple example: a store turning over $2 million a year across forty accounts, whose largest customer buys $200,000 a year. That one account is 10 percent of revenue. Losing it is a bad quarter. In a cohort chart it is one dot.
What to measure instead
1. The gap since each account's last order, against its own rhythm
The most useful number in wholesale is how long it has been since an account last ordered, compared with how often that account normally orders.
Twenty days without an order means nothing for an account that orders every two months. It means a great deal for one that orders every week. A store-wide "days since last order" average hides this completely. It only works account by account.
This is the earliest warning you will get. Customers rarely announce that they are moving to another supplier. Their orders get further apart, and then they stop.
2. Revenue concentration
Know what share of your revenue comes from your top five and top ten accounts. This is not a performance metric. It is a risk map.
It tells you which gaps in ordering need a phone call this week and which can wait. A quiet week from an account worth 12 percent of revenue is a different event from a quiet week from one worth 0.5 percent.
3. Revenue tied to accounts that are slipping
Put the first two together. Of the accounts whose ordering has slowed or stopped, how much did they buy over the last twelve months?
That figure answers the question consumer analytics never ask: how much revenue is quietly at risk right now, and from whom.
Doing it by hand
You can build all three from a WooCommerce order export. Group the orders by customer. Work out each account's usual interval between orders. Compare it with the time since their last order. Then sort by what each account bought over the past year.
The first time, that takes an afternoon in a spreadsheet. The trouble is that it is out of date the next morning, and the accounts you most need to catch are the ones that change while nobody is looking.
It also helps to look at ordering and payment side by side. An account that has gone quiet and an account that has stopped paying are often the same account, and the second signal usually shows up in your accounts receivable before anyone notices the first. If your on-account orders are sitting in "processing" with no record of what has been paid, this post explains why WooCommerce leaves them there.
Where Customer Dashboard fits
Kascado's free Customer Dashboard does a simpler version of this inside WooCommerce. It finds your B2B customers from their roles, including any you add in its settings, and sorts them into active, at risk and dormant by how long it has been since each one last ordered, using thresholds you can set. It shows the revenue at risk from the customers who have gone quiet, and lets you filter the customer list by status with each account's revenue alongside, so the accounts worth a call this week are easy to find.
The short version
Consumer analytics are not wrong. They are built for a business with thousands of interchangeable customers. A wholesale store has a few dozen customers who are not interchangeable at all.
Measure each account against its own ordering rhythm, know which accounts carry the revenue, and keep a running figure for the revenue tied to accounts that are slipping. Those three numbers will tell you more about the health of a wholesale business than every chart on a standard analytics dashboard.