Do you have a single source of truth for revenue?

Can you trace any number on a dashboard back to its source system in under 10 minutes?

Do you know which datasets are SOX, GDPR, HIPAA or PCI relevant?

Is PII masking enforced at source for every PII column?

Do you have automated lineage from source through your Gold layer?

Do you have an SLA for dataset freshness, measured weekly?

Are schema changes detected and flagged within 24 hours of deploy?

Do data quality checks run on every Gold dataset, with pager-grade alerting on failures?

Do you know your top 10 most-queried tables and what each one costs you per month?

Is your data warehouse spend predictable month-to-month within 10%?

Do you know which dashboards your leaders actually opened last week?

Can an executive drill from a headline number to the underlying cause in under 60 seconds?

Do you have variance vs. plan analysis on your top financial KPIs, refreshed monthly or better?

Can your team build and ship a new board in under five working days?

Is there a single calibrated KPI dictionary across the company (one revenue, one churn, one margin)?

Do your dashboards surface AI-written annotations that explain anomalies as they appear?

Can any period be compared to any other period (vs. plan, vs. LY, vs. forecast) in one click?

Have you retired any dashboards in the last 90 days?

Do you measure dashboard usage and review low-use boards monthly?

Does at least one board drive a recurring decision (a meeting, a queue, a follow-up) every week?

Do you have an AI agent currently in production (not a chatbot, an agent that does something)?

Can you measure the ROI of any deployed AI feature in dollars (saved, recovered, or avoided)?

Do AI signals come with confidence scores and audit trails the user can inspect?

Have your AI agents been reviewed and approved by your audit, risk or security team?

Do you track adoption of AI recommendations (how many fire vs. how many are acted on)?

Can you trace which data sources fed which AI decision, for any decision in the last 90 days?

Is there a kill switch and a runbook for every AI agent in production?

Do you monitor model drift (input distribution, output quality) weekly or better?

Have AI signals translated into measured P&L outcomes (saved or recovered) in the last 90 days?

Do you have a formal AI governance policy that has been read by someone other than its author?