Nine finance managers, one forecast, and a version-control problem that ate the close cycle every month. This is the tool that replaced the spreadsheets, built end-to-end in Google Apps Script.
Labor forecasting for a large technology organization ran on parallel Excel models, one per finance manager, each blending Delivery Manager baselines with their own judgment, each formatted differently, each emailed around at close. Consolidation meant chasing versions, reconciling collisions, and losing hours every cycle to questions that had nothing to do with finance: which file is current, whose number is right, and what changed since last month?
Inside a corporate environment, the sanctioned toolchain was Google Workspace: no servers, no databases, no deployment pipeline. So the constraint became the architecture. Google Sheets as the data layer (source data, prior-month snapshots, SOW breakouts, and an append-only adjustments log), Apps Script as the backend (an automated Excel drop-zone that ingests, converts, snapshots, and archives each monthly cycle), and a custom web app as the front end, served through HtmlService to every FM simultaneously.
This is the production interface with the Google backend swapped for a simulated data layer. Everything else is real: edit any Sep–Jan cell (switch to Without SOWs or Only SOW Labor first), run the bulk tools, flip the heatmap, and hit save to watch the audit log capture your changes. All figures and names are fictional. NovaTech is an invented company.
Simulated dataset · Regenerates on refresh · Best experienced on desktop
The parallel spreadsheets are gone. Nine finance managers now forecast asynchronously in one workspace, against one dataset, with one audit trail, and consolidation stopped being a job. What used to be version reconciliation is now variance analysis: the tool absorbed the mechanical work so the close conversation could move to what the numbers mean, not which file they live in.