01 · The problemThe CRM tracks deals. Nobody set it up to measure the business.
The pattern repeats at almost every company before it has a RevOps team. Quota lives in a spreadsheet, so attainment is settled after an argument. Renewals aren't their own opportunities, so the ARR in the board deck doesn't match the CRM and churn is rebuilt from invoices. There are forty reports and no agreed number, and the forecast is whatever was said on Monday's call.
None of that is a dashboard problem. It starts in the data model.
02 · What I builtFrom CRM fields to a Monday brief
A Salesforce org runs a simulated company: eight reps and about 200 accounts, with deals that open, move, slip, close, renew and churn every night. Everything after the simulation is real.
- CRM changes the metrics need: a quota record per rep per period; opportunity types that separate new business, expansion and renewal; contract start and end dates on won deals; a date stamp for each lifecycle stage; forecast categories mapped to stages.
- A daily load into BigQuery, plus a nightly snapshot of open opportunities, so the forecast as it was called can be compared to what closed.
- A SQL metric layer: staging tables, then metric marts for attainment, ARR and the ARR bridge, gross and net revenue retention, funnel conversion and stage timing, forecast trend and pipeline coverage.
- A seven-page dashboard suite: Overview, ARR, Churn & retention, Rep performance, Funnel, Forecast and Pipeline.
- Push reporting: a weekly revenue brief, plus alerts when coverage drops below 3x, a deal slips twice, or a renewal is at risk.
- A published metric dictionary: every number has a formula and a note on where it usually goes wrong.
03 · StackWhat it runs on
| Layer | Tool | Used for |
|---|---|---|
| CRM | Salesforce | Quota, opportunity types, contract dates, stage stamps, forecast categories |
| Warehouse | BigQuery | Daily load and nightly opportunity snapshots |
| Modeling | SQL | Staging tables and metric marts |
| Delivery | Dashboards, scheduled export, briefs | The seven-page suite and the views on this site |
04 · What changedOutcome
- Every number in the demo traces back to a written definition and a CRM field.
- Forecast accuracy is measurable, because the call is snapshotted every night instead of remembered.
- The whole chain is public: the demo on this site reads the same export as the full suite.
The company and its deals are simulated. The pipeline, metric layer and dashboards are real.