For most of my career, finding out which customers and products actually made money was a project. You assembled a team, pulled a year or two of transactions out of the system, argued about allocations, and waited weeks for a picture that was already going stale by the time it arrived. Many companies never started, and I understood why.
That barrier is gone.
What changed
For The 80/20 Business Process I ran the entire diagnostic inside an ordinary spreadsheet, driven by its built-in AI assistant. I started from a single tab of raw sales transactions, the kind any sales or financial system can export. In about two hours, the spreadsheet produced the 80/20 tables for customers and products, the Pareto charts, the color-coded customer-product matrix, the quadrant and quartile analyses, the complexity allocation, a product-level true profitability model, and the whale curve that ties it all back to operating income.
Two hours to see, with numbers, exactly where a business makes its money and where it quietly gives it back.
The tool doesn't matter much. The same sequence runs in Excel with Copilot, in Sheets with Gemini, or with Claude or ChatGPT connected to your data. What matters is the thinking behind the prompts and what you decide once the picture is on the screen.
Three views of the same customers
One small feature of that exercise turned out to be the most persuasive. I asked for a single switch at the top of the matrix that flips every cell between sales, margin and quantity, while the ranking and the quadrant borders stay anchored to margin.
Switch to sales and the core block dominates the page. Switch to quantity and the tail suddenly looks very busy. Same customers, same products, three very different stories. It is the fastest way I know to show a skeptical team that revenue and effort do not live in the same place.
The part the machine can't do
I want to be blunt about the consequence. When the diagnosis costs an afternoon, the only remaining excuse for not knowing your true profitability is the decision not to look.
And looking is still the easy part. The spreadsheet will show you, undeniably, that a customer you would have sworn was core is a net drag, or that a product you had written off is quietly funding half the P&L. It will not reprice that customer, retire that product, or have the conversation with the sales team. A beautifully formatted whale curve that nobody acts on is just a nicer way of not deciding.
The arithmetic is now nearly free. Judgment and resolve are what's left, and they were always the scarce resource.