Insights

Field notes on revenue data and AI governance.

Why reported revenue numbers drift from the systems that produce them, and what to fix before someone outside the building checks.

Revenue data integrity

Your forecasting tool can't see what happened after the deal closed

Forecasting tools score a snapshot of the CRM. The evidence that breaks forecasts lives in field history, and nothing you've bought reads field history.

August 8, 2026 · 2 min read

Which NRR did you tell the board?

Three teams can compute three different NRRs in good faith. The tells are what the number is computed from, when churn leaves it, and whether you can reproduce it.

August 8, 2026 · 2 min read

What a revenue data integrity assessment costs (and what you get)

Groundwork's assessment is $15,000 fixed and takes two weeks. Here is exactly what the fee buys, what remediation costs after it, and how the credit works.

August 8, 2026 · 3 min read

Revenue data integrity: what it is and how it's different from RevOps

A working definition, the three practices the discipline covers, and where it sits relative to RevOps, BI, and the forecasting-software layer.

August 8, 2026 · 4 min read

How to prepare revenue data for diligence

Diligence is a reproduction test, not a presentation review. What the team on the other side actually checks, the artifacts that pass, and when to start.

August 8, 2026 · 3 min read
Start here · AI Governance

5 Ways AI Agents Fail When Your Data Stack Wasn't Built for Them

AI agents fail on revenue data for five reasons, and none of them is the model: APIs that buckle under concurrent calls, write conflicts with no resolution rules, audit logs that can't tell agents from humans, over-broad permissions, and untested rollback. All five are testable before you deploy.

June 19, 2026 · 11 min read