Projects to leave Excel usually die at one of two extremes: in the six-month analysis that starts with "let's document every process first," or on the Monday morning that starts with "everyone is on the new system as of today." Both fail for the same reason — they treat migration as a switch-flipping moment.
The migrations that work follow a formula so boring it hurts: one flow, real work, a small team, four weeks. That's the playbook.
Why are we still on Excel? (The honest answer)
Not because Excel is bad — because it looks free and bends to anything. The problem is the invisible invoice that starts running once a team passes three people: the _final_v3 (2).xlsx family of the same file, formulas only one person understands, copy-paste errors, and decision history that lives nowhere. We've itemized that invoice in Gloyd vs Spreadsheets; no need to repeat it here.
What actually postpones the move is perceived risk: "What if the new system slows us down?" This entire playbook exists to shrink that risk.
Week 1: Pick one flow, move the data
Don't migrate everything — pick one flow. The best candidate is usually the RFQ→quote→order chain: the pain is daily and the output is measurable.
Then the data: do not launch a master-data perfection project. Import your active suppliers, customers, and recurring products from CSV/Excel — you live in Excel, which means the file already exists. Whatever's missing gets created the first day it's needed; that's not a defect of the method, it is the method.
The team: one owner, two users. A pilot without an owner is nobody's pilot.
Week 2: Run a real RFQ round on the platform
Don't poke at test data — run a real request through the system. Send one RFQ to three to five suppliers, collect responses through the portal, compare on screen, convert the winner to a PO in one click.
The goal of this round is to produce two numbers: how many days did this take the old way, and how many hours now? How many copy-pastes then, how many now? Those two numbers are your internal sales deck from week three onward.
A note on the supplier side: a supplier arriving via an invitation link faces a learning curve the size of a web form. If one or two still reply by email in round one, nothing is lost — you enter their quote, and they meet the portal in round two.
Week 3: Bring the other side in
Once the flow sits comfortably inside, it's time for self-service: open portal access for customers, so quote approvals and order tracking move from your phone to their screen. If you sell, make your recurring catalog items visible in the portal store — priced items take orders, unpriced ones collect requests.
This is also the week to watch for "Excel relapses": who opens the old file, at which step? That spot is either a training gap or a genuine friction in the flow — both fixable, but only if visible.
Week 4: Don't ban Excel — retire it
The most important rule of the whole migration: don't forbid the old file. Bans produce secret usage. Instead, flip the file to read-only archive status and declare the platform the only home for new work. Excel remains a superb analysis tool — every list exports to Excel anyway. The difference is that the source of truth is now the record, not the file.
By the end of week four you have: real deals that ran on the platform, a measured speed difference, a few suppliers used to the portal — and the question "which flow is next?"
The three classic mistakes
- Big-bang: moving every department on the same day. Scaling without a pilot just scales the risk.
- Perfect-data obsession: "every product card must be complete first." Complete data is the output of a used system, not its precondition.
- Endless parallel running: keeping both systems alive indefinitely is double work. The parallel period is a date, not a mood — our recommendation is the four weeks above.
You don't need a budget window to start; the trial opens without a card. Four weeks later you either hold a measured win — or your old files, with nothing lost. That's the beauty of the playbook.
