Module 08: Copilot in Excel Advanced + Analyst
Lesson 08.2: Finding Trends, Outliers, and Patterns
Lesson Promise
Use Copilot to look for trends, outliers, and patterns while avoiding false certainty.
Real-World Scenario
A customer success team wants to know whether churn risk is rising by segment, region, or product tier.
Core Concept
Trends show direction over time, outliers show unusual records, and patterns show relationships worth investigating. They are not automatically proof of cause.
Ask Copilot to show the columns used, the time window, the comparison group, and the reason a value is considered unusual.
For text-heavy data such as survey comments or support notes, ask for themes and sentiment carefully, then spot-check raw examples before using the summary.
Step-By-Step Workflow
- Define the time period and comparison group.
- Ask for trends first, then outliers, then possible drivers.
- Request a table with finding, evidence, confidence, and verification method.
- Spot-check outliers manually.
- Ask what alternative explanations could exist.
- Turn strong findings into follow-up questions rather than instant decisions.
Prompt Lab
Bad Prompt
Find patterns.
Better Prompt
Find trends and outliers in churn risk by segment and region. Show columns used and explain why each outlier matters.
Expert Prompt
Analyze this customer health dataset for trends, outliers, and patterns. Compare churn risk by segment, region, product tier, and month. For each finding, provide evidence, possible explanation, alternative explanation, confidence level, chart recommendation, and manual verification step. Do not claim causation without evidence.
Hands-On Exercise
Ask for three trend findings and three outliers, then manually verify one of each.
Deliverable
A trend-and-outlier brief with evidence, possible explanations, and verification notes.
Executive Analysis Review Checklist
Common Mistakes
- Turning one spreadsheet question into a confident executive recommendation without checking the data.
- Asking for forecasts without naming assumptions, time horizon, inputs, and uncertainty.
- Using charts or PivotTables that look polished but do not answer the decision question.
- Treating Analyst, Python, file upload, or advanced editing features as universally available.
- Sharing analysis that does not separate observed facts, assumptions, hypotheses, and recommendations.
Quiz / Checkpoint
Why should trend and outlier findings include alternative explanations?
Because patterns can be real without proving cause; alternatives keep the analysis honest.
Official Sources To Verify
- Get started with Analyst in Microsoft 365 Copilot
- Get started with Copilot in Excel
- Frequently asked questions about Copilot in Excel
- Get insights about numerical data with Copilot in Excel
- Get insights from text-based data with Copilot in Excel
- Create PivotTables with Copilot in Excel
- Create charts with Copilot in Excel
- Edit with Copilot in Excel
- File formats supported by Microsoft 365 Copilot
- Copilot Chat licensing explanation
Want your AI product explained to a large AI-native audience?
Kingy AI helps AI companies turn complex products into clear, useful YouTube videos that drive awareness, product understanding, demos, clicks, and search visibility.
