Revenue and customers

One revenue number, and the ability to explain it

Revenue by division, customer and period, on one basis, with the concentration and the trend visible behind it.

What you will be able to see

The numbers you can finally trust

Each one exists in your systems today. None of them is currently something you would bet a decision on.

Revenue by division: On one definition, with unattributed revenue shown, not hidden.
Customer concentration: How much of the result depends on how few accounts.
Trend: Daily and monthly, against the comparable prior period.
Customer detail: Balance, history and behavior in one place per account.

What changes

Nothing here is a new metric. It is the one you already care about, made trustworthy.

Why you cannot trust it today

Revenue is computed at the wrong grain
Work at document header level instead of from the underlying postings and you fold in things that are not revenue. You also lose the ability to split it by the dimensions you manage by.
Attribution has gaps
A good share of revenue often cannot be assigned to a division at all. We show that share instead of spreading it around to make the chart look tidy.
Everyone has their own extract
Several spreadsheets, each defensible, none reconciling. Meetings are spent agreeing whose number to use.

What we do about it

Model from the postings
Where the line-level detail exists, we build revenue from it instead of from document headers. Most attribution problems disappear at that point.
Show the gap
Unattributed revenue is displayed as its own figure. A number that hides its own uncertainty is worse than one that admits it.
Publish one definition
The measure is modeled once and read everywhere, so the meeting starts from an agreed figure.

We check in week one. Where only header-level data exists, we tell you what that limits, instead of showing you a breakdown the data cannot support.
Yes. The warehouse is a normal one and can be read by whatever you already use.

Other measures