Excel
Copilot
ClaudeAnalyze
Advanced·45 min
Spot trends and anomalies with the Analyst agent
Hand a spreadsheet to Copilot's Analyst agent and have it reason through the data step by step to find what's unusual.
Updated Sep 30, 2026
Scenario
One region is reporting impossible margins.
Mission
Find the 5 transactions responsible for the data skew.
Start Files
| TransactionID | Product | Region | Revenue | Cost | MarginPct |
|---|---|---|---|---|---|
| TX-1014 | Premium Support Contract | West | 4739 | 4170 | 12 |
| TX-1009 | Premium Support Contract | West | 3939 | 3309 | 16 |
| TX-1040 | Standard License | West | 1673 | 1221 | 27 |
| TX-1038 | Standard License | West | 4409 | 3483 | 21 |
| TX-1057 | Training Bundle | West | 2200 | 1452 | 34 |
Preview — first 5 rows of west-region-transactions.csv
Your task
- Sort the region's transactions by margin percentage.
- Flag any transaction where the margin falls outside a plausible range for that product line.
- Isolate the 5 rows that account for most of the region's reported margin swing.
Prompt idea:
This region's transactions show margins far outside our usual range for this product line. Given the sample rows below, what data-entry patterns (decimal errors, duplicated discounts, wrong unit price) would produce this kind of skew, and how would you flag them in Excel?
Check Your Result
Which product line do all 5 anomalous transactions belong to?
Need a hint?
Sort by MarginPct descending — the anomalies cluster at the very top.
Reveal Solution
The solution files contain the completed challenge. Review the approach to see if you missed any edge cases.
Related Workflows
Analyze
Advanced·45 min
Analyze customer profitability
Merge cost and revenue data to find hidden margin drains.
ChatGPTUpdated Sep 30, 2026
WORK WITH LIMINALS
Make AI useful for your team.
Get practical help with AI consulting or ask about training courses for your team.