Ask three people in the same business how many leads came in last month and you'll often get three answers. None of them is wrong. Each one is using a different definition, usually without knowing it.
Where the differences come from
- Which records count. Does a duplicate count? A test record? A lead from an existing customer?
- Which date. The day it was created, the day it was qualified, or the day it closed? A sale booked on the 31st and paid on the 2nd lives in two different months.
- Which system. The CRM, the ad platform and analytics each count their own version. Ad platforms count conversions they claim; your CRM counts records that exist.
- Which statuses. A cancelled job, a refunded order or a withdrawn application may or may not be subtracted.
What a good definition contains
| Name | The plain word your team uses, such as "Booked job" |
|---|---|
| Counts | Exactly which records are included |
| Excludes | What is removed, and why |
| Dated by | The single date that places it in a period |
| Source | The system of record, and how other systems join to it |
| Owner | Who agreed the rule and who can change it |
Write it once, apply it everywhere
A definition only helps if every report uses it. That means the dashboard, the weekly snapshot, the monthly review and the board pack all compute the number from the same rule, in the same place. When someone changes the rule, it changes everywhere, and the change is logged.
Reconcile before you rely on it
Once a rule is written, compare its result with the report your team already trusts. Where they differ, explain why in one sentence. Usually the explanation is a definition choice, and the team agrees which version wins. Only then put the number in front of leadership.
This is the core of how we work. Read how it works.