Skip to content

AI uses for records

Spreadsheet trackers as systems of record: are they licensable?

By SourceX Editorial · Updated

Short answer

Spreadsheet trackers can be licensable when they work like a system of record: every row is dated, versions are kept, columns stay consistent and each row links to a job, order, ticket or project in another system. A tracker overwritten every week with no history is a snapshot, not a record of work, and carries little value.

Key takeaways

  • A tracker counts as a record when the team actually runs work from it and it keeps history.
  • Row-level dates, an ID column and an outcome column matter more than formatting or formulas.
  • Version history in Google Sheets, OneDrive or SharePoint can recover changes the sheet itself overwrote.
  • Trackers usually add value alongside system exports, filling gaps the systems never captured.
  • Client-owned trackers and tabs full of personal data need to be separated early.

When does a spreadsheet count as a system of record?#

A spreadsheet counts as a system of record when it is where the team actually logs work and decisions, not a report copied from somewhere else. If people would stop working without it, it is probably a record.

Most established companies have a handful of these. An engineering firm tracks RFIs and submittals in Excel before or alongside Procore. A distributor keeps carrier claims in a shared sheet. A software company runs its escalation list in Google Sheets. A manufacturer logs nonconformances in a workbook because the ERP's quality module was never set up.

The test is behavior, not the tool. A Google Sheet that has served as the dispatch board for years, with every row kept, can be a better record than a half-configured module in an expensive system.

Checklist: is your tracker licensable?#

A tracker is a good candidate when most of these statements are true. Missing items do not disqualify it, but each one adds preparation work.

Run the checklist on each tracker separately. A company's carrier claims sheet may pass while its staffing plan fails, and the answer for one says nothing about the other.

  • Each row has a created date and, where relevant, a closed or completed date.
  • Rows are added, not overwritten, and old rows stay in the sheet or an archive tab.
  • Version history is retained in Google Sheets, OneDrive or SharePoint, or dated copies exist.
  • Column meanings have stayed stable, or changes are noted with dates.
  • An ID column links each row to a job, order, ticket, project or customer record elsewhere.
  • An outcome or resolution column says how each item ended.
  • Notes columns contain real reasoning, not just status words.
  • There is one master tracker, not several diverging copies.
  • The tracker is internal, not a client deliverable or a client's own log.
  • Personal data is limited and sits in identifiable columns.
  1. 01Your work records
  2. 02Personal details removed
  3. 03You approve the release
  4. 04Licensed to AI developers
How it works with SourceX: Your work records → Personal details removed → You approve the release → Licensed to AI developers.Illustrative

Which trackers tend to carry value?#

The trackers that carry value record decisions and exceptions, because those are what core systems most often miss.

Trackers that only mirror system data, such as a weekly export pasted into a sheet for a meeting, rarely add anything. The ones worth keeping hold information that exists nowhere else: the reason a bid was declined, the engineer's answer to an RFI, the settlement agreed with a carrier.

Which trackers tend to carry value?
TrackerCommon inWhy it can matterLink it to
Bid and estimate logTrades, engineering, distributionShows which work was pursued, priced and won or lostCRM or estimating records
RFI and submittal logEngineering and architectureCaptures questions, responses and turnaroundProject numbers in Deltek or Procore
Nonconformance logManufacturingRecords defects, dispositions and causesWork orders and lots in the ERP
Carrier claims and exceptionsLogistics and 3PLShows what went wrong in transit and how it was settledShipments in the TMS or WMS
Escalation trackerSoftware and servicesLists hard cases and who resolved themTickets in Zendesk or Intercom, issues in Jira
Resource and staffing planConsulting and engineeringShows how people were allocated against demandProjects and time entries

Common tracker problems and how to fix them#

Most tracker problems come from editing the sheet as a live board rather than keeping it as a log. Several can be repaired after the fact, as long as the fix is documented.

Keep a short log of every repair: what was changed, from which version and by whom. That log becomes part of the provenance record if the tracker is ever included in a package.

Common tracker problems and how to fix them
ProblemFix
Rows deleted or overwritten when items closeRecover prior states from file version history, and from now on move closed rows to an archive tab
Several copies emailed around and edited separatelyPick the master, merge unique rows from the copies and record which copy each came from
Columns renamed or repurposed mid-yearWrite a short column history with dates and map old columns to new ones
Dates typed in different formatsNormalize into a new column and keep the original text
Hidden tabs with personal or client detailsInventory every tab, then exclude or separate the sensitive ones
Status values only, no outcomeAdd an outcome column going forward; for history, infer outcomes only from linked records and flag them

Illustrative: an engineering firm's RFI logs#

Illustrative: a fictional civil engineering firm adopted Procore for newer projects, but older projects tracked RFIs in Excel workbooks stored in SharePoint project folders. Each workbook had its own column layout, and project numbers matched the firm's Deltek records.

The managing principal asked the operations lead whether the older logs were worth keeping in usable form. The team collected the workbooks, mapped their columns to Procore's RFI fields, used SharePoint version history to recover responses that had been typed over, and linked each log to its Deltek project.

Logs for projects where the client owned the deliverables were set aside for a rights review. The rest now form one consistent RFI history spanning both eras, which the firm also uses to train new project engineers.

How trackers fit with system exports#

Trackers fit best as companions to system exports. A carrier claims sheet means more next to the TMS shipments it refers to, and an escalation tracker means more next to the tickets and engineering issues it lists. On their own, trackers often lack the surrounding detail that makes each row understandable.

SourceX treats trackers as systems in their own right during the Supply step of the SourceX five-step transaction: they appear in the inventory with location, date coverage and linking columns. If a tracker is included, the SourceX Evidence Packet records which file versions were used. In the SourceX Enterprise Data Value Framework, a well-kept tracker can add human-generated signal and domain expertise, while poor data cleanliness raises preparation cost.

How to make trackers record-grade from today#

Making a tracker record-grade takes a few small rules rather than a new system, and the rules cost almost nothing to adopt.

  • Add created and closed date columns and make them required.
  • Add an ID column that matches the system the work belongs to.
  • Never delete rows; move closed items to an archive tab.
  • Freeze column names and keep a dated note whenever one must change.
  • Keep the master in a location with version history turned on.
  • Add an outcome column and a short reason column.

Frequently asked questions

Is a Google Sheet with edit history better than Excel files on a shared drive?

Often, because cloud version history keeps prior states automatically. Excel files stored in OneDrive or SharePoint can also keep versions, depending on library settings. Files on an old network share usually keep only what someone saved as a separate copy, so check before assuming history exists.

Do we need to clean our trackers before a fit check?

No. A fit check uses metadata only: which trackers exist, what they cover, how many years they span and which systems they link to. Cleanup, if any, comes later and only for trackers that are actually included.

What about trackers that list client names?

Client names are usually replaced with organization tokens during preparation. The larger question is whether the tracker or its contents belong to the client under your contract. Trackers that are client deliverables are reviewed during the rights step and are often excluded.

Can a tracker replace a missing system export?

Partly. A tracker can show what happened when the system history is gone, but it usually holds less detail. Treat it as the best surviving record, describe its limits clearly and avoid presenting it as a full substitute for the lost system data.

Do formulas, macros and conditional formatting matter?

Not much. What matters is the values people entered and when they entered them. Formulas that pull from other files can break when those files move, so capture computed values alongside the formulas before archiving, and note any macro that changed data automatically, since its edits are not human decisions.

Related resources

See if your company qualifies

A short company assessment. No data uploads are needed.

See if your company qualifies