Moving a consultancy off spreadsheets without losing a quarter

Almost every consultancy under thirty people runs the same way: a folder structure on a shared drive, a scheduling spreadsheet, a Word template, an accounting package that does not know about any of it, and one person who knows where everything is. It works, which is exactly why replacing it is risky. The failure mode is not that the new system is worse. It is that the migration stalls halfway and the firm ends up running two systems at once for a year.

Sequence by what breaks, not by what is easy

The temptation is to start with whatever imports most cleanly. Start instead with whatever currently costs you the most when it goes wrong, because that is where the firm will feel the benefit early enough to keep going.

For most testing firms that order is: reporting first, then scheduling, then proposals, then invoicing. Reporting is the largest recurring time sink and the place where inconsistency is most visible to clients. Scheduling is where double-booking costs real money. Proposals and invoicing are painful but rarely urgent.

Phase one, reporting only

Move nothing except report production. Load your existing template, run one live engagement through it end to end, and deliver from the new system. Keep the spreadsheet for everything else, untouched.

This is deliberately a small commitment. If it does not work you have lost a week on one engagement, not a quarter across the firm. If it does, everyone has now seen the thing they cared about most get better, which is what buys you the rest of the migration.

• Pick a real engagement, not a test one. Test data produces test-quality conclusions.
• Do not clean up the template first. You want to know how it behaves as it is.
• Have the person who normally writes the report do it, not the person who is enthusiastic about the new system.
• Compare the output against what you would have shipped, line by line, once.

Phase two, the knowledge base

Before scheduling, take the findings you have written most often and turn them into reusable templates. Twenty covers most firms’ recurring work. This is the highest-leverage hour in the whole migration and it is routinely skipped because it feels like admin rather than progress.

Take them from reports you have already delivered rather than writing them fresh. They are already in your house voice and already survived a client reading them.

Phase three, everything with a date on it

Now move scheduling, and move it wholesale rather than in parallel. Two calendars is worse than either calendar. Pick a date, move the forward book, and delete the spreadsheet the same day so nobody can quietly keep using it.

The single most reliable predictor of a stalled migration is a system left running "just in case". Nothing forces adoption like the old thing being gone.

Phase four, the commercial side

Proposals and invoicing last, because they touch your accounting and because getting them wrong is visible to clients in an uncomfortable way. By this point the engagements exist in the system, so a proposal can create one and an invoice can reference one, which is most of the benefit.

What to carry across, and what to leave

Leave the historical archive where it is. Bulk-importing four years of completed engagements is a large amount of work that produces a system full of records nobody opens. Keep the old drive read-only and move history only if someone actually asks for it twice.

Roughly how long it takes

A five to fifteen person firm that follows this order and does not try to do it all at once typically has phase one done inside a week, the knowledge base in an afternoon, scheduling within a fortnight, and the commercial side over the following month. The firms that take a year are, almost without exception, the ones that tried to do all four at the same time during a busy quarter.