Q3 Retail Turnaround
A 240-store chain lost 11% of revenue last quarter, and nobody can say which stores or why. You get three years of transactions, a store master with duplicate IDs, and one question from the CEO.
- Time needed
- 3 weeks
- Level
- Open to all
- Data
- 1.2M rows
The brief
The brief is deliberately underspecified, the way it arrives on the job. Revenue is down 11% across 240 stores, and the CEO wants to know which stores drove it and what to do about it. The store master has duplicate IDs from a 2024 migration, and promo flags are missing for two regions. There's no answer sheet: you're judged on the decisions you make and how honestly you document them.
What you submit
- A five-slide deck answering the CEO's question, written for a non-technical reader
- The notebook or query file that produced your numbers, runnable end to end
- A short data-quality note: what you dropped, what you imputed, and why
How to participate
- Register free: create your account and tell us a little about yourself.
- Download the dataset pack and the one-page brief from your challenge dashboard.
- Work solo or in a pair. Submit from your dashboard whenever you are ready.
- Mentors score blind, then publish the leaderboard and written feedback for every entry.
How it’s judged
- Correct numbers
- A recommendation the CEO could act on
- Honesty about what the data can't tell you
- Anyone can re-run your work and get the same answer
Good to know: You don't need to be a PrimrIQ learner. You can work in SQL, Python or Excel. If you're newer to this, try "Clean the unclean sales master" first: it's the same kind of data, in 4 days.
What you walk away with
Everyone who submits gets written mentor feedback. Top 3 win a full flagship programme seat.