Why spreadsheets fail at scale
Spreadsheets don't fail suddenly — they fail slowly, then all at once. The failure modes are predictable:
✕
Version proliferation
After 6 months you have asset-register-v2-FINAL-updated-march-REAL.xlsx on three different laptops. Nobody knows which one is current.
✕
No mobile access
Technicians can't update a shared spreadsheet from the workshop floor. Work orders get transcribed hours later — if at all.
✕
PM schedules drift
A 12-month PPM schedule in a spreadsheet relies on someone manually checking due dates. One distracted week and three inspections are overdue.
✕
No audit trail
When an inspector asks "who signed off this inspection and when?" — the spreadsheet says "John" with no timestamp, no evidence, no defensibility.
✕
Reporting is a side job
Every metrics request requires manual copy-paste into a separate sheet. The manager who produces the report spends half a day on it every quarter.
If 2–3 of the above describe your current situation, you're ready to migrate. The good news: it doesn't take months.
Before you start: the non-negotiables
Don't start the 72 hours until you have these three things locked:
1
Executive sign-off
A migration that doesn't have management support will stall the moment a technician complains. You need one sponsor who can say "yes, we're doing this" when people push back.
2
A chosen CMMS
The 72-hour plan assumes you've already evaluated and selected a system. If you haven't, read our CMMS Buyer's Guide first and run a 30-day POV. Trying to migrate and evaluate at the same time is a common failure mode.
3
A clean data snapshot
Before the weekend starts, get a complete export of every spreadsheet that contains asset, PM schedule, or maintenance history data. Treat these as your source of truth. Don't let anyone edit them during the migration window.
Friday Evening · 3–4 hrs
Data audit & export
Your goal for Friday evening is a complete, clean inventory of everything that needs to move. No cleaning yet — just cataloguing.
Checklist: Friday Evening
Find every spreadsheet that contains asset or maintenance data (include SharePoint, Google Drive, email attachments)
Export each as CSV — name them clearly: assets-site-A-2026-04.csv
Count total rows: how many assets, how many active PMs, how many open work orders
Identify your top 10 asset types by count — these go in first
Flag any fields that don't have a clear CMMS equivalent (custom fields, notes, attachments)
Note any PMs due in the next 30 days — these need to be in the system by Monday morning
Assign one person to own the import for each site
Set a Slack/Teams channel for the migration weekend
Rule: don't clean data on Friday
Export first, clean Saturday. Trying to clean and export at the same time means you finish neither properly. Get everything exported to a shared folder and stop.
Saturday · 6–8 hrs
Clean & import
Saturday is the hardest day. Data cleaning is unglamorous but it determines whether your CMMS is trusted from day one or becomes just another spreadsheet.
Data cleaning rules
Standardise asset naming
Pick one format: [Type]-[Site]-[Number]. AHU-SiteA-001. Don't allow free text names — they'll create chaos in search and reporting.
Merge duplicate assets
If the same boiler appears in three spreadsheets with slightly different names, it becomes one asset in the CMMS. Mark duplicates before you import.
Set a "good enough" threshold for history
Don't try to import all historical maintenance records. Import the last 12–24 months of work orders for critical assets only. The rest can stay in the archive spreadsheet.
Flag missing serial numbers
Leave the field blank rather than guessing. You can update it when a tech next visits the asset. A wrong serial number is worse than no serial number.
Map PM frequencies explicitly
Every PM needs a defined frequency. "Quarterly" is fine. "When needed" is not. Force a decision on every PM before it goes in.
Import order matters
1
Sites / locations first — Everything else is parented to a location.
2
Asset types / categories — Define your taxonomy before you import assets.
3
Assets — Start with the site you know best. Fix the import template, then run other sites.
4
PM schedules — Attach to assets after assets are confirmed correct.
5
Users / technicians — Add after the asset structure is validated.
6
Open work orders — Import in-progress work last. These are the most likely to have dirty data.
Sunday · 4–5 hrs
Configure & test
Sunday is about validation and confidence-building. By Sunday evening, every person who uses the system on Monday should have logged in, seen their assets, and closed at least one test work order.
System checks
All assets imported and visible
PM schedules generating correctly
Notifications firing to the right people
Mobile app works offline (disconnect Wi-Fi and test)
Role-based access confirmed (techs can't edit what they shouldn't)
Team prep
Send login credentials to all users
Run a 20-min Zoom walkthrough for each team
Create a "cheat sheet" PDF: how to open, complete, and close a work order
Nominate one internal champion per site
Set up a support channel (Slack/Teams) for Monday questions
Sunday evening gate: go/no-go check
Before you go to sleep Sunday, answer: can every technician close a work order on their phone without asking for help? If yes: go. If no: identify the specific blocker and fix it tonight. Don't launch with known friction on Monday morning.
Monday · Go live
Go live — hard cutover only
The single biggest mistake in CMMS migrations: running the old spreadsheet and the new system in parallel “for a few weeks.” Don't. Parallel running guarantees data drift, confusion, and eventual abandonment of the new system.
Monday morning means: the spreadsheet is locked (read-only or archived). All new work orders go in the CMMS. No exceptions.
Monday morning message to send your team
“As of today, all maintenance work orders, PM completions, and asset updates go through [CMMS name]. The spreadsheet is now read-only. If you have a question or hit a problem, message [champion name] or drop it in [#channel]. We'll have someone available all day.”
First week targets
100% of new work orders created in the CMMS
All PMs due in the next 7 days visible and scheduled
Daily check-in: how many work orders closed via mobile?
End of week: run your first PM completion rate report
Identify any asset types still missing — add them in week 1
Handling resistance
Every migration has at least one person who really doesn't want to change. Here's how to handle the most common objections:
"I don't have time to learn a new system."
Show them: closing a work order on mobile takes 30 seconds. The spreadsheet took 3 minutes plus the effort of finding the right file. This isn't more work — it's less work once the muscle memory is there.
"What if the system goes down?"
Most CMMS platforms have better uptime than a shared Excel file on a laptop. Check the vendor's uptime history. Also: good CMMS apps have offline modes so you can still work without connectivity.
"We'll lose all our history."
You're not deleting the spreadsheet — you're archiving it. It's still there if anyone needs to look something up. And you're migrating the last 12–24 months of critical history anyway.
"The old way worked fine."
This is usually not about the software — it's about autonomy. Involve that person in the configuration. Make them a site champion. Give them ownership of a small piece of the rollout.
Common pitfalls
1
Trying to import everything
Fix: Import critical assets and active PMs. The rest can be added over the next 30 days by the team as they encounter the assets.
2
Parallel running
Fix: Hard cutover only. See above. Parallel running kills migrations.
3
No internal champion
Fix: One person per site owns the system. They're the first call when something goes wrong, not the CMMS vendor.
4
Over-configuring before go-live
Fix: You don't need 47 custom fields before Monday. Get the core structure right — assets, PMs, users — and add complexity after the team is using it.
5
Not training on mobile specifically
Fix: Desktop demos and mobile use are completely different. Make sure technicians have practiced on the actual device they'll use on the job.