Lesson and practical work
Professional outcomes
- Turn an ambiguous stakeholder concern into testable analytical questions.
- Separate metric movement from the drivers that may explain it.
- Recognize where data quality, definitions, and communication affect the answer.
- Describe analytical work as a decision process rather than a list of tools.
Need a refresher?The analyst workflow in one viewView concise foundation
An analyst connects a decision to evidence. The practical sequence is: clarify the decision, define the population and period, identify reliable data, calculate and validate metrics, investigate drivers, communicate limitations, and recommend a proportionate next action.
SQL, Excel, Python, and BI tools support that sequence. They are not the sequence themselves.
Orders increased, but monthly sales fell
“We received more orders this month, but sales fell. Tell me what happened.”
The request leaves at least three definitions unresolved: whether sales means gross, net, or invoiced revenue; whether the month is complete; and whether order count includes cancellations. Begin with a plan: reconcile the period, confirm definitions, decompose revenue into volume and average order value, then test price, discount, product mix, returns, and regional effects.
Decompose the result before searching for a story
| Metric | Prior | Current | Change |
|---|---|---|---|
| Completed orders | 10,000 | 11,400 | +14% |
| Net revenue | $1.20M | $1.15M | -4.2% |
| Average order value | $120 | $101 | -15.8% |
| Gross margin | 38% | 32% | -6 pp |
Volume improved, but each order generated less revenue and margin. This directs the investigation toward discounting, lower-priced product mix, returns, or channel mix. It does not prove which driver caused the movement.
What the result means—and does not mean
A defensible interim statement is: “The revenue decline is associated with a material fall in average order value, while margin compression suggests that discounting or product mix may also be involved.” You cannot conclude that a promotion caused the decline without promotion, pricing, and product-level evidence.
Analyst's checklist
- Compare complete and equivalent reporting periods.
- Confirm whether revenue is gross, net of returns, or invoiced.
- Exclude or separately report cancelled and refunded orders.
- Reconcile total revenue before and after segmentation.
- Check whether currency conversion or reporting rules changed.
Common mistakes and failure modes
- Opening a tool before clarifying the decision.
- Reporting percentage change across incomplete periods.
- Assuming association between promotions and margin establishes causation.
- Presenting many segmented charts without identifying material drivers.
Challenge yourself
A second extract shows online orders up 30%, store orders down 8%, and online average order value 25% below store average.
- What decomposition would separate channel mix from within-channel performance?
- Which metric would you reconcile first?
- What would you avoid telling the director at this stage?
Show hint
Write revenue as order volume × average order value for each channel. Compare channel shares and each channel’s own performance.
Show analytical guidance
A shift toward a lower-value channel can reduce total average order value even if neither channel deteriorates. Use a mix-versus-rate decomposition before attributing the change to customer behavior.
How this may appear in an interview
Walk me through how you would investigate a revenue decline when order volume increased.
What a strong candidate should communicate
Start with definitions and period completeness, validate totals, decompose revenue into volume and value, segment along plausible drivers, and communicate both the strongest evidence and remaining uncertainty. A strong answer names validation checks and the stakeholder decision—not only tools.
What if the decline disappears after excluding returns posted this month for orders placed last month?
Where this skill becomes evidence
Use this structure in a sales-performance project: business question, definitions, validation, driver decomposition, material finding, limitations, and recommended follow-up. That evidence is stronger than a dashboard screenshot.
Test your analytical judgment
Answer before opening the guidance. More than one defensible approach may exist.
Why can order count rise while revenue falls? Give two plausible mechanisms.
Show guidance
Which checks would you complete before comparing this month with last month?
Show guidance
Draft a two-sentence interim finding that does not overclaim causation.
Show guidance
Move from reviewing expert reasoning to performing the analysis.
Guided practice can add the synthetic dataset, multi-step assignment, validation checks, solution comparison, project workflow, and progress tracking.
- Work with the data
- Make and defend assumptions
- Validate the result
- Build portfolio evidence