Insight · Data Governance
Why Dashboards Disagree — and Why It Matters
When different reports show different numbers, nobody trusts the numbers. A metric dictionary fixes this and unlocks faster decisions.
What this explains
Why reporting trust breaks down when KPI definitions drift and how a metric dictionary restores consistency.
Who it is for
Data leaders, operations teams, and executives dealing with recurring disputes over dashboards and business metrics.
Core takeaway
Trust improves when every important metric has one definition, one calculation approach, and one accountable owner.
The situation we see all the time
Finance says ARR is $2.1M. Sales says it's $2.3M. Product says it's $1.9M. Who's right? Nobody. And until someone's right, every decision gets questioned.
The cost of metric disagreement
- Meetings about meetings: 30 min sync to figure out which number is real
- Slow decisions: Executives second-guess reports. Reviews take twice as long
- Lost trust: Analytics becomes political. Teams stop using dashboards
- Duplicate work: Everyone builds their own "trusted" version
Why it happens
No single source of truth
Each dashboard owner interprets "Active User" or "Revenue" differently. If it's not documented, different people make different choices.
Hidden calculation logic
The SQL that calculates a metric lives in a dashboard. Someone else has the same metric in a different SQL query. Silently different.
No ownership
Nobody's responsible for keeping definitions aligned. When Finance updates Revenue, does Sales know? Does Product?
The solution: A metric dictionary
A metric dictionary is a governed, centralized catalog where every metric has ONE definition, ONE calculation, and ONE owner.
✓ Definition
What is "Active User"? Logged in in last 30 days.
✓ Calculation
SQL query documented. No hidden logic.
✓ Owner
Product Analytics owns this metric. They're accountable.
✓ Source Data
Comes from analytics.user_activity. That's it.
Example: ARR (Annual Recurring Revenue)
Definition
Total monthly subscription value × 12, including active subscriptions and excluding churned accounts
Calculation
SUM(subscription_monthly_value × 12)
WHERE status = 'active' AND churn_date IS NULL
Owner
Revenue Operations (accountable for accuracy)
Source Data
crm.subscriptions table (updated daily by CRM sync)
Notes
Excludes free trials. Updated daily at 1am UTC. Last verified: Jan 2025
Now every dashboard uses the same ARR. Finance, Sales, and Product all see $2.1M. Done.
How to build it
List the pain points
What metrics do people disagree about? Start with top 10-20 (Revenue, ARR, Active Users, Churn)
Document each metric
Definition. Calculation. Owner. Source. No surprises, no ambiguity.
Enforce in your BI tool
Lock the calculation in Looker, Power BI, or your semantic layer. All dashboards use the same formula.
Set up change control
Changes go through review. Notify stakeholders before deploying. No silent updates.
Monitor and iterate
Quarterly: Review the dictionary. Add new metrics. Update calculations. Keep it current.
The payoff
- Trust restored: Dashboards show the same numbers. Disagreements end.
- Faster decisions: No more "which report should we believe?" Questions get answered faster.
- Less rework: One calculation. One formula. No duplicate effort.
- Accountability: Someone owns each metric. Clear responsibility.
Common questions
Why do dashboards disagree?
Dashboards disagree when teams use different metric definitions, hidden calculation logic, inconsistent source data, or unclear ownership over KPI changes.
What is a metric dictionary?
A metric dictionary is a governed catalog that documents each metric definition, calculation method, owner, source data, and change history.
How does a metric dictionary improve BI trust?
It improves trust by making KPI logic consistent across reports, reducing disputes, speeding decisions, and giving teams a shared source of truth.