A record that holds up, and reports that build themselves.

01 · The challenge

Fast to build. Hard to trust.

Every contact with a landowner is a legal record. On a program heading toward condemnation cases, those notes and their dates can become evidence. The original setup couldn’t carry that weight.

• Tract IDs were typed by hand, so the same tract showed up different ways
• Dates were stored as text, so turnaround time couldn’t be calculated at all
• The same work was tracked in several places, and nobody could say which was right
• The main tracker had grown to 203 columns, closing in on the platform’s cell limit before the program was half done
• Twice-weekly client reports were rebuilt by hand, often after hours

02 · What we built

Three layers, each with one job.

1. RegisterOne row per tract: the record of truth
2. Project plansOne template, milestones as rows, not columns
3. ExceptionsFormulas flag what’s late, blocked, or missing
4. Role viewsPeople work in views, never in the sheets
5. Roll-upThe client report becomes an output, not an input

1. One register of record. A single tract register holds what each tract is. Reporting and the GIS integration read from it, so there is one answer to “which sheet is right?”
2. Work streams as rows. One project plan per project from a single template. Collapsing 203 columns to about 30 means a new project is a new sheet, so capacity grows with the program instead of running out.
3. A queue instead of line-by-line review. Formulas flag what’s late, overdue, blocked, or missing. Quality checks become a list a person can actually work.
4. Access by role and by milestone. A Work App with role-based dynamic views. Each of ten roles sees and enters only the milestones it owns, and access can be revoked cleanly.
5. Clean data, once. Landowner names from three trackers were consolidated into one register, and every decision is logged so the next question has an answer on record.

03 · How we proved it

Same report. No hand assembly.

Before anything replaced the old setup, the new model rebuilt the client’s roll-up deliverable for the pilot tracts, and we compared it cell by cell against the existing files. It matched, apart from a couple of known quirks in the source data.

Nothing went big-bang. The old environment keeps running, untouched, until the team is ready to retire it, and the two are reconciled before anything is switched off.

04 · Where else it works

Any program that outgrew its first build.

The same pattern fits right-of-way and permitting programs, multi-site construction, asset rollouts, and any operation where one tracker has quietly become too many columns, too many copies, and too much manual reporting.