THE WATERLINECompanion · Gabor Szorad

Chapter 3 · Descend: find and map

Analytics tools worth pointing at your own numbers

A specialised analytics tool, not a general chatbot, and a sharp question about your own data.

Most cases announce themselves as pain. A task someone hates, a groan you can hear across the office. A smaller set never does, because there is no task attached to them at all. They hide in the data.

A retailer I worked with was convinced they were running near break even. In fact they were giving away around €50,000 a quarter in margin, by habitually discounting items that were already selling through at full price. Nobody felt that as a pain, because nobody was doing a hard job badly. There was just money on the floor that no one had inspected.

The question that found it.
Which items sold through at least 75% of last season's stock without ever being discounted? The judgment came first. The tool scaled it.

Data analysis

getdot.ai

Start with your question

  1. Point it at your own numbers, sales or margin or stock, and ask one question you already half suspect the answer to.
  2. Read the answer as a sceptic. Where you have real expertise you can tell when it is wrong.
  3. Put the finding into money. If you can say where the value lands, it is a case. Price it in the calculator.
Before you connect anything.
Decide which data may go into an outside tool. The one page rule is on the data policy page.
The Waterline companion. Everything here works in Claude, ChatGPT, Copilot and Gemini. Nothing to install. Last reviewed 5 November 2026.