Six spreadsheets. Six staff. Three years of trading. And on any given Monday, three different people could give you three different revenue numbers for the prior week — and all three would be right, by their own definitions.
This is a real business. An e-commerce retailer selling primarily through Shopify, with some marketplace channels alongside it. Not a start-up in chaos. A functioning business with processes, a small team, and genuine trading history. Just no single source of truth.
The spreadsheet situation
The six spreadsheets had grown organically, as they always do. Each one had started as a solution to a specific problem:
- A weekly sales tracker
- A monthly product performance sheet
- A customer acquisition log
- A refund tracker
- An inventory forecast
- A "summary" sheet that pulled from all of them
None of them agreed. The summary sheet pulled from the others, but the others had different cut-off times, different definitions of what counted as a "sale," and different people maintaining them with different levels of rigour. Over time, the summary sheet had become the most unreliable document in the business — because it faithfully reflected all the inconsistencies of everything feeding into it.
The trigger
The breaking point came in a team meeting. The owner and the ops manager spent twenty minutes arguing about whether the business had grown or declined in Q4. Both had data in front of them. Both were right, according to their own spreadsheet. Neither could prove the other wrong.
The meeting ended without a decision. Not because the data was unavailable — but because there was too much of it, in too many places, with too little consistency.
Two people. Both with data. Neither able to prove the other wrong. That's not a data problem. That's a system problem.
That afternoon, they decided to do something about it.
What they did
No consultants. No new software subscriptions. No data migration project.
They opened Shopify, went to the standard orders export, and downloaded a CSV. It took about thirty seconds. They pulled the same kind of export from Xero. They sent both files to The BI Playbook for a Review.
Five business days later, the Review came back: a written report covering revenue trends, customer cohorts, product performance, refund rates, average order value, and repeat purchase rates — alongside the specific definitions to apply going forward and the decisions the data was pointing at. No field mapping project. No setup work. No consultants billing by the hour.
All from the same data that had been sitting in Shopify and Xero the entire time.
The outcome
The owner retired four of the six spreadsheets within the week. Not because anyone told them to — because the Review was answering the questions those spreadsheets had been trying to answer, with consistent definitions and a far smaller maintenance burden.
The remaining two spreadsheets stayed. One covered inventory forecasting, one tracked supplier terms — neither of which were in the transaction data. Those were legitimate, purpose-built tools. They kept them.
Weekly review meetings went from 90 minutes to 25. Not because there was less to discuss — because the first 65 minutes had previously been spent establishing what the numbers actually were before anyone could discuss what to do about them. That problem went away.
What actually changed
Here's the part worth understanding: the underlying data was exactly the same. Every transaction, every customer, every refund — it all existed in Shopify and Xero before the CSV was ever exported. Nothing new was created.
What changed was having a single shared view of the business, built on consistent definitions everyone agreed on. No manual entry. No formula drift. No version control arguments. One set of numbers, one shared reality.
The Q4 argument that prompted all of this? That kind of conversation doesn't happen any more — because everyone's looking at the same numbers before the meeting starts.
That's what BI actually does. Not magic. Not complexity. Just clarity — and the decisions that follow from it.