Migrating off spreadsheets: a one-weekend playbook for travel agencies

The reason most travel agencies are still running on spreadsheets isn't that they don't know better. It's that the perceived migration cost is too high. "We'll do it in Q3, when we have time." There is no Q3.

This post is the playbook for shipping it over a weekend. Not a perfect migration — a good-enough migration that lets you stop opening the sheet on Monday morning.

The Friday inventory

You almost certainly have these tabs (or files):

• A Tours tab with rows like Russia 12-day Imperial / from $7,200 / Moscow + St. Petersburg
• A Departures or Calendar tab with dates and seats remaining
• A Bookings or Reservations tab with names, dates, deposits paid
• A Guests or Contacts tab — usually the messiest, often duplicated against bookings
• An Inquiries or Leads tab, if you're disciplined; an inbox folder, if you're not

Print these. Get them on one wall.

The cut: what to migrate, what to leave

Migrate (Friday evening, 90 minutes):

• All active tours. Every tour you actually sell today. Skip retired ones — archive the sheet, you can dig it out if you ever need it.
• Departures with future dates. Anything in the past is history; it can stay in the spreadsheet as an archive.
• Active bookings (departure date in the future, or final payment outstanding).
• Guests attached to active bookings. The rest can be imported later as a backfill.

Leave behind (for now):

• Historical bookings older than 12 months
• Guest records with no booking history
• Old quotes that never converted
• The custom formulas in the cells. Whatever they computed, your new system should compute server-side or you write a one-shot script.

The temptation is to migrate everything. Don't. The point of the weekend is to stop using the sheet on Monday, not to make the new system a museum of the old one.

Saturday: the import

Most travel CRMs accept CSV. The format that works without a fight:

tour_name, departure_date, capacity, base_price, currency
"Imperial Russia 12-day", 2026-06-14, 8, 7200, USD

For bookings:

tour_name, departure_date, guest_email, guest_name, status, deposit_paid
"Imperial Russia 12-day", 2026-06-14, [email protected], Anna Müller, confirmed, 1500

Two CSVs is enough to get an agency operating. You'll backfill notes, custom touches, and dietary preferences in the first week of real use — that data was probably never in the sheet anyway. It was in someone's head or in an email thread.

Sunday: the cutover

The cutover isn't technical, it's behavioral. Three things have to happen:

1. The sheet becomes read-only. Lock it. If someone needs to edit it, they have to ask why. The friction is the point.
2. Everyone's email forwards into the CRM inbox. Not "should." Configure the rule before Sunday evening.
3. The Monday morning routine starts in the new tool. Print the new dashboard URL and tape it over the spreadsheet bookmark.

If the AI-drafted reply feature is set up, the very first inquiry on Monday morning is the tutorial. Watch the AI write its draft. Edit it. Send it. That's the new ritual.

What goes wrong

The two failure modes:

The team keeps the spreadsheet open in another tab "just in case." Six weeks later they're still double-entering everything. Fix: delete the tab from everyone's bookmarks. Painful, correct.

The CRM doesn't have a field that the spreadsheet had. Usually a custom one — deposit paid via Wise (currency) on (date). Make a Notes field on the booking; document that it goes there. Don't let one missing field block the migration.

What you get back

Two things, neither of which shows up in the dashboard immediately:

• The team stops asking each other "do you have the latest sheet?" That sentence is a tax. It evaporates the day after migration.
• You can finally answer "how much revenue is in the pipeline?" in one click. The spreadsheet always lied about this because nobody updated it consistently. The CRM is honest because it's the system of record.

A weekend is enough. The reason it doesn't get done in a weekend is almost never technical. It's that nobody wrote the playbook down.