Spreadsheet Analysis for Decision Makers
Turn a messy export into an answer you can defend in a meeting, without learning to code.
Beginner with Yasmin Halloran
Design reporting that changes a decision instead of decorating a wall.
Almost every organization has a dashboard that was celebrated at launch and abandoned within a quarter. It happens for the same reasons each time: it was built for a stakeholder rather than a decision, it shows what was easy to measure, and nobody agreed what anyone should do when a number moves.
This course starts from the decision and works backwards. You identify the choice a person makes repeatedly, the information that would change it, and the smallest view that delivers that. Then you build it, watch real users try to read it, and fix what they misread.
You should already be comfortable getting data into a shape you can work with. If you are not, take Spreadsheet Analysis for Decision Makers first. The tool you build in is up to you: students have used spreadsheets, business intelligence platforms and a plain web page, and the design principles are the same.
Applications for the Q4 2026 cohort close on Friday 18 September 2026. We reply to every application within two business days.
Analysis and reporting for people who are not analysts
We work with genuinely messy data, including the duplicate rows and the inconsistent dates, because clean sample files teach the wrong instincts.
Each module has one live session and one piece of practice applied to your own work.
Finding the choice somebody makes every week, and the information that would genuinely change it.
Definitions people agree on, metrics that resist gaming, and knowing what each number will do when the business changes.
Matching the shape of the question to the shape of the mark, and the four chart types that cover most real reporting.
Designing the path an eye takes through a screen so the headline arrives before the detail.
Making a number mean something by putting it next to what it should be, and agreeing what counts as a problem before it happens.
Deciding what deserves an interruption, who receives it, and how to retire an alert everybody has learned to ignore.
Watching five people read your dashboard without help, recording where they hesitate, and fixing what they misread.
Getting a dashboard into a routine somebody already has, and agreeing who owns it in six months.
By the last session of Dashboards People Actually Use you will have practised each of these on your own work, not on a case study.
Getting an honest answer out of the numbers you already have, and reporting it so it counts.
Turn a messy export into an answer you can defend in a meeting, without learning to code.
Beginner with Yasmin Halloran