Migrating your practice off spreadsheets: a 6-phase plan

Migrations fail for the same reason almost every time: trying to move everything at once, during the busiest week of the month. This plan spreads the work across six phases and deliberately leaves most of the old data behind.

Before starting, the decision that saves the most time: you are not going to migrate your whole history. Clients who have not been in for two years do not need to be inside the new tool; they need to be retained and accessible, which is not the same thing. Separating those two ideas cuts the work to a fraction.

Phase 1 · Inventory and scope

Open everything holding client data and list it: spreadsheets, PDF folders, the notebook, email, your phone. Then sort every client into three groups.

In a typical practice this split usually leaves 20-30% of clients in the active group. That is the real size of your migration.

Phase 2 · Field mapping

Put your sheet next to the new tool’s record and decide, column by column, where each value goes. It is the dullest phase and the one that prevents 90% of later problems. Watch the cases that always cause trouble:

• Dates: unify the format before exporting. Mixing dd/mm/yyyy with mm/dd/yyyy silently corrupts whole months.
• Decimals: comma versus point depending on system locale. A weight of 70.5 can become 705.
• Combined fields: first and last name in one cell must be split first.
• Free text with line breaks: these break CSVs unless quoted.
• Notes columns mixing several things: decide whether they go to observations or the clinical record.
• Values that only exist inside a formula: export the result, not the formula.

Phase 3 · Templates and recipes first

Before loading a single client, rebuild what you will reuse: plan templates, your standard instructions, your recipes. It feels backwards — clients seem more urgent — but it determines whether the tool saves you time from week one or costs you time for a month.

A good starting point: the five plans you have duplicated most over the last year, and the ten instructions you repeat at nearly every visit.

Phase 4 · Load and sample-check

Load five test clients first, not three hundred. Review those five thoroughly and only then continue. What to check in the sample:

1. Accents and special characters display correctly.
2. Dates of birth match and the calculated age is right.
3. Weights and measurements keep their decimals.
4. No field has shifted by one column.
5. Long notes are not truncated.
6. No duplicate clients were created.

Phase 5 · One month in parallel

For four weeks: every new client goes only into the new tool, and follow-ups are handled wherever they already live. Do not duplicate work by recording in both; the point of running parallel is not having everything in both places, it is that a failure in the new tool does not leave you without a practice.

Use that month to migrate naturally: every dormant client who reappears is set up in the new tool at their appointment, with fresh data they confirm themselves.

Phase 6 · Cut-off and archive

Set a date and hold it. From that day the spreadsheet becomes read-only. And before closing it:

• Export the full history in a stable format readable without the original tool.
• Store it encrypted, with at least two copies in different locations.
• Note where it is and who has access, and add it to your record of processing.
• Set an annual review date to confirm the archive is still readable.
• Remove shared access to the old sheet: it still contains health data.

Before calling the migration done

• You can export your data out of the new tool whenever you want.
• The active client count matches what you had.
• Templates work and genuinely save time in a real consultation.
• The archived history opens correctly from a different machine.
• Nobody is still working in the old sheet "just for one thing".