What you'll learn
Turn plain-language business questions into working Excel formulas — XLOOKUP, SUMIFS, dynamic arrays, LAMBDA — without memorising syntax.
Clean a hostile 419-row export with seven planted data problems into an audited table, with reconciliation and a reusable monthly runbook.
Build four retail deliverables: a management report, a cleaning operation, a printable One-Pager and a weekly Trading Pack.
Analyse sell-through, inventory aging, promo effectiveness and revenue per square meter to drive markdown, reorder and range decisions.
Verify every AI-generated result with a five-check discipline, so the numbers you present survive challenge from finance and management.
Explain, audit and rewrite inherited formulas nobody understands, spotting fixed ranges and performance traps before they break.
Build a prompt library and monthly refresh workflow that turn one-off answers into repeatable reporting assets your team can share.

Requirements
Microsoft 365 Excel (Windows, Mac or web) and a Claude Pro, Max, Team or Enterprise plan. The add-in does not run on Office 2019 or free accounts.
Comfortable with basic Excel: entering data, simple formulas such as SUM, and moving between sheets. No advanced functions required.
No coding, no VBA and no statistics background needed. Every dataset and file used in the course is provided for download.

Description
This course contains the use of artificial intelligence.

You know the feeling. A question arrives — which stores missed target, what is our promo dependency by category, which SKUs are tying up cash — and before you can answer it, you have to translate it into syntax. Which function. What argument order. Whether the field you need even lives in the sheet you are working in.

That translation layer is what this course removes.

You describe the outcome in plain language. Claude writes the formula, cleans the data, assembles the report. And then you verify it — because the number carries your name, not the tool's.

WHAT MAKES THIS COURSE DIFFERENT

Most AI-in-Excel courses stop at generic examples: a sales table, a customer list, a budget. This one goes into a specific industry and uses its actual vocabulary and its actual decisions.

Everything runs on one realistic retail dataset — 5,000 transactions, 119 SKUs, 10 stores, 12 months of history, plus a deliberately hostile 419-row export with seven planted data problems. You will find every one of them on camera. Clean textbook data teaches you nothing about the file that actually lands in your inbox.

WHAT YOU WILL BUILD

Four deliverables you can put in front of a director on the day you finish

- The Friday Report — a four-block management summary built in one conversation, then re-scoped in a single sentence

- A cleaning operation — a hostile export turned into an audited table with a reconciliation, a duplicate log and a reusable monthly runbook

- The One-Pager — KPIs, chart, variance heat strip and written commentary on one printed page, cloneable per category

- The Monday Trading Pack — momentum, sell-through extremes, aging alerts and evidence-linked actions, regenerated weekly with one prompt