Power Accelerate

It Was Never Meant to Be This

It started as a calculation. Then someone added a macro, then a user form, then a rates card, then an approval step. Nobody decided to build a business system in Excel. It happened one helpful addition at a time, and now a department depends on it.

The Logic Is Undocumented

The rules live in formulas nested six deep, in VBA behind a user form, and in the head of whoever wrote it. There is no specification. Reading the workbook is the only way to know what it actually does, and almost nobody does that before quoting a rebuild.

Governance Cannot See It

No version control, no audit trail, no access model beyond whoever has the file share. Numbers that reach the board are computed in a file that anyone can edit and nobody can diff. Audit findings against spreadsheets are one of the most common triggers for a modernization budget.

One Person Deep

Every one of these workbooks has exactly one person who understands it, and that person is usually close to moving on. The risk is not that the file breaks. It is that the file keeps working and nobody can change it.

Archive

For workbooks nobody edits any more but whose numbers must survive. Rows and structure are extracted to a queryable SQLite archive that outlives Excel, the file share and any licence.

Routed here when: no populated data sheets, no formulas and no VBA - nothing left that a rebuild would carry across.

Keep in Excel

Sometimes the honest answer is to leave it alone. A genuine ad-hoc model with no shared users and no governance exposure does not need a platform. The tool will say so rather than manufacture a migration.

Routed here when: plenty of formula logic but very little stored data, no VBA and no structured tables - a calculation model, not an application.

Lift to Dataverse or Canvas

For structured, shared workbooks with light logic. The sheets are already tables in all but name. They map onto Dataverse with a model-driven app, or onto a canvas app, without being reinvented.

Routed here when: eight or more populated data sheets wired together by cross-sheet lookups reads as relational, so Dataverse. List-shaped tracker workbooks fall through to Canvas on SharePoint.

Rebuild as a Code App

For the workbooks that are genuinely applications. Heavy VBA, user forms, deep formula chains, external integrations. AI rebuilds them as a Power Apps Code App over Dataverse, with the logic carried across rather than guessed at.

Routed here when: more than 1,000 lines of VBA, or three or more user forms, or a database-backed connection with VBA behind it.

Rows to SQLite

Every sheet becomes a table, every row a record, in one portable SQLite database per workbook. Queryable with plain SQL, readable years from now without Excel and without a licence.

Formulas as Data

The full formula inventory exported with sheet, cell, occurrence count, depth and complexity, so the rules can be reasoned about as a set rather than hunted cell by cell.

VBA Source

One readable .bas file per module, extracted straight from the workbook package. Standard modules, class modules, ThisWorkbook and sheet code-behind, and user forms with their control layouts.

The Dependency Graph

Cross-references, sheet overlap and recompute order exported as structured data, so the rebuild knows what depends on what before it starts writing anything.

Risk & Quality Signals

Data-quality findings, validation rules, the input surface, and a scan of the VBA for embedded credentials and connection strings. Anything found is reported and redacted from the module bodies before they are bundled, so the secrets do not travel with the handoff.

AI Rebuild Bundle

One zip carrying the canonical model, the schema, sample data, the VBA source and a written rebuild instruction. Hand it to public AI tooling such as GitHub Copilot and it has everything it needs.