When an AI tool tells you it "saved $200K," how do you know if that number is real?
What Are Confidence Tags?
Confidence tags are labels attached to every number in a report that tell you where the number came from and how much you should trust it. Three levels:
- [verified] — pulled directly from a system of record: admin export, invoice, billing API. The source is auditable.
- [calculated] — computed from verified inputs using a defined formula. The math is transparent and reproducible.
- [estimated] — based on assumptions derived from interviews, samples, or benchmarks. The assumptions are stated explicitly.
Why Don't Other Platforms Tag Their Numbers?
Because it's hard, and it reveals weakness. If 60% of your findings are [estimated], that's less impressive than presenting everything with equal confidence. Most consulting reports and analytics platforms present estimates as facts because it looks better. The problem: decisions based on untagged numbers are decisions based on unknown confidence levels.
How Do Confidence Tags Change Decision-Making?
When a board member sees "$47,880 in unused Copilot seats [verified]," they act on it. When they see "$15,000 in potential API optimization [estimated]," they investigate further. The tag doesn't reduce the value of the finding — it increases the trust in the entire report. One honest [estimated] tag makes every [verified] tag more credible.