Two analysts pull win rate from the same CRM and return different numbers. Both are correct. One counted deals, one counted dollars, and neither knew the other had made a choice. A metric definitions document exists to remove that choice from the analyst and put it in one place everyone reads.
What is a sales metric definitions document?
It is a single reference that records how every reported metric is calculated, from which fields, with which filters, and by whose authority. The document is not documentation of a dashboard. It is the specification the dashboard has to conform to.Most revenue teams discover they need one during a board prep cycle, when a number in the deck fails to match a number in the CRM and three days disappear into reconciliation. The reconciliation itself is usually straightforward. What takes the time is establishing which of the two numbers is the correct one, and that argument has no resolution mechanism without a document.
The practical test is whether a new RevOps hire can reproduce last quarter's board numbers on their second week without asking anyone a question.
Which fields does each definition need?
Seven fields per metric, and the document breaks down if any of them is optional. Anything less produces a definition that sounds precise and is still ambiguous in practice.| Field | What it captures | Example entry for win rate |
|---|---|---|
| Name | The exact label used on every dashboard | Win rate (dollar-weighted, closed basis) |
| Plain-language definition | What it measures in one sentence | Share of closed dollar value that resulted in a won deal |
| Formula | The precise calculation | Closed-won amount divided by closed-won plus closed-lost amount |
| Source fields | System and field names | CRM opportunity: amount, stage, close date |
| Filters and exclusions | What is deliberately left out | Excludes renewals, excludes deals under $1000, excludes test records |
| Period basis | Which date drives inclusion | Close date within the reporting period |
| Owner and cadence | Who approves changes and how often it refreshes | Sales ops lead, weekly |
Which metrics cause the most disagreement?
Five metrics generate most of the friction, and each one has multiple defensible calculations. Win rate has four common variants: count-based or dollar-based, and cohort-based on creation date or period-based on close date. All four are legitimate. Pick one for the company scorecard, document the others as named variants, and never let an unlabeled win rate appear on a slide.Pipeline created depends entirely on where you draw the line. Counting at stage one measures marketing throughput. Counting at qualification measures pipeline that has a chance of closing. Teams that count at stage one and forecast off the result end up with coverage ratios built on records that never had a buyer attached.
ARR needs an explicit treatment of mixed contract terms, multi-year deals, and professional services. Sales cycle length needs a start event, because created date and first-meeting date can differ by weeks. Pipeline coverage needs a stated numerator, because open pipeline in the period, qualified pipeline in the period, and total open pipeline produce very different ratios against the same quota.
How should retention and forecast metrics be documented?
Write out the full waterfall rather than a single formula, because retention is a reconciliation and not a ratio. ORM structures the monthly movement as beginning ARR, churned customer ARR, churned product ARR, product decrease ARR, new customer ARR, new product ARR, increased product ARR, and ending ARR, with beginning ARR always equal to the prior month's ending ARR.Documenting the waterfall rows is what makes gross and net retention auditable. Anyone can then trace net revenue retention back to specific movement rows rather than to a formula nobody can decompose.
Forecast accuracy needs three decisions written down: which submitted forecast is scored, at what point in the period it was captured, and whether the comparison runs against closed-won bookings or recognized revenue. Reaching roughly 90% accuracy on new and expansion business is achievable with sustained manual effort, but the number only means something if the measurement rules are fixed before the period starts.How do you get finance and sales to agree?
Assign each definition to the team that has to defend it outside the company. Finance owns revenue, ARR, and retention definitions because those numbers appear in board reporting and diligence. Sales leadership owns stage definitions and quota structure. RevOps owns everything downstream and owns the change process itself.That split resolves most standoffs before they start. When sales wants to count bookings at signature and finance counts at start date, the answer is that both definitions live in the document under different names, and the company scorecard uses the finance one.
Run one working session to draft, then circulate for a two-week comment window with a hard close. Definitions documents that stay open for comment indefinitely never ship.
What about the argument that your data is too messy for this?
Messy data does not prevent reliable measurement, and inconsistent data does. Nearly every revenue team believes their CRM data is uniquely bad and that this is why they cannot run the business the way they want. The belief is close to universal, which is the clue that it is not the real constraint.Garbage in does not have to mean garbage out. As long as the data is consistently garbage, meaning the same field is wrong in the same way across periods, patterns hold and predictions work. What breaks measurement is a field that meant one thing in Q1 and something else in Q3 because a manager changed the picklist without telling anyone.
A definitions document is the enforcement mechanism for that consistency. It is worth more than a cleanup project, because cleanup fixes the past and definitions protect the future.
How do you keep definitions from drifting?
Require a change request with a named approver, log every change, and audit the dashboards against the document quarterly. Drift almost never happens in the document. It happens in a dashboard, when someone adjusts a filter to make a chart look right for a specific meeting and the adjustment stays.The quarterly audit is a mechanical exercise. Pull each metric from the live dashboard, recompute it from the documented formula, and flag mismatches. Expect to find mismatches every quarter, even in a team that is running the process properly. Finding none usually means the audit was not run against the live dashboard.
Version the document and keep the old versions readable. When someone asks why a number in last year's board deck does not match today's calculation, the answer should be a diff rather than a guess.
Frequently Asked Questions
What is a sales metric definitions document?
A single reference that records the formula, source fields, filters, owner, and refresh cadence for every metric a revenue team reports. It exists so two people running the same query get the same answer. Without it, most dashboard disputes turn into definition disputes that nobody can settle.
Which sales metrics cause the most disagreement?
Win rate, pipeline created, ARR, sales cycle length, and pipeline coverage. Each has several defensible calculations. Win rate alone can be measured on deal count or dollars, and on deals created in a period or deals closed in a period, which produces four different numbers from identical data.
Who should own the metric definitions document?
RevOps owns the document and the change process. Finance owns the revenue and retention definitions inside it, and sales leadership owns the pipeline stage definitions. Splitting ownership by domain keeps each definition close to the team that has to defend it externally.
How do you stop metric definitions from drifting?
Require a change request with a named approver before any formula changes, log every change with a date and reason, and version the document. Then review the whole file quarterly against what the dashboards actually calculate, because drift usually happens in the dashboard rather than in the document.
Does a definitions document fix bad CRM data?
No, but it makes bad data survivable. Consistent data produces reliable predictions even when it is incomplete. What breaks a forecast is inconsistency, where the same field means different things in different periods or across different teams, and a definitions document is what enforces the consistency.
See how ORM turns these insights into action
ORM builds custom revenue forecast models for B2B SaaS companies. Not dashboards. Prescriptive analytics that tell you what to do next.
Schedule a Demo