Skip to content

Week 02 Assignment — Tools & Business Analytics

A capstone-style assignment covering Week 02: Power BI / Tableau dashboards, business analytics frameworks, KPI design, ML concepts, and data storytelling. The focus shifts from analysis to communication and business impact.

Time budget: 8-10 hours Submission: A published dashboard (Power BI or Tableau) + a 5-slide presentation + a short recorded walkthrough (optional but recommended) Weighting: 60% of the course grade

The Brief

You're a data analyst at FitLife, a subscription fitness app. The leadership team is meeting to plan next quarter and they've asked you to build a dashboard and present the state of the business.

"We need to know: are we growing? Where's the money coming from? What's our biggest risk? And what should we do about it? Build us something we can look at every week, and walk us through what it says."

You're presenting to the CEO, CFO, and Head of Product. Each cares about different things. Your dashboard and presentation must serve all three.

Dataset

Use a subscription or e-commerce dataset with: customers, subscriptions/orders, dates, revenue, and ideally a churn or status field. Suggested:

  • Telco Customer Churn (if focusing on retention)
  • Subscription/SaaS sample data (Maven Analytics has several)
  • Your Week 01 dataset, extended — continue the same analysis with a dashboard focus

Required Deliverables

Part 1 — Dashboard (40 points)

Build an interactive dashboard in Power BI Desktop or Tableau.

  • [ ] KPI header row: 4-5 key metrics with trend indicators (revenue, active users, churn rate, ARPU, growth rate)
  • [ ] Revenue analysis: trend over time with MoM/YoY comparison
  • [ ] Segmentation: revenue or users broken down by a meaningful dimension (plan tier, region, acquisition channel)
  • [ ] A retention or cohort view: show how well you retain customers
  • [ ] At least 2 interactive elements: slicers, filters, or drill-through
  • [ ] Clean design: consistent colours, clear titles, no chart junk
  • [ ] Calculated measures: at least 3 custom DAX measures (Power BI) or calculated fields (Tableau)

Part 2 — Business Analytics (20 points)

Demonstrate at least TWO of these frameworks applied to the data:

  • [ ] KPI framework: define 3-5 KPIs with targets and explain why each matters
  • [ ] Funnel analysis: map a conversion funnel and identify the biggest drop-off
  • [ ] Cohort analysis: retention by signup cohort
  • [ ] RFM segmentation: Recency, Frequency, Monetary customer segments

Part 3 — ML Concept Application (10 points)

You don't need to build a production model, but demonstrate analytical maturity:

  • [ ] Identify one prediction problem in the data (e.g., "which customers will churn?")
  • [ ] Explain whether it's regression, classification, or clustering and why
  • [ ] Describe what features you'd use and how you'd evaluate the model
  • [ ] (Optional, bonus) Build a simple logistic regression or decision tree with scikit-learn

Part 4 — Data Storytelling Presentation (30 points)

A 5-slide presentation for the leadership team:

  1. Situation — where is the business now? (1-2 key numbers)
  2. Finding 1 — the most important insight, with a supporting chart
  3. Finding 2 — the biggest risk or opportunity
  4. Recommendation — what should they do, with expected impact
  5. Next steps — what you'd analyse next, what data you'd need

Requirements: - [ ] One key message per slide - [ ] Every chart supports the narrative (no decoration) - [ ] Tailored to the audience (executives — strategic, not technical) - [ ] A clear, specific recommendation tied to a finding

Grading Rubric

Criterion Weight What "excellent" looks like
Dashboard design 25% Clean, intuitive, answers the leadership questions in 5 seconds
Technical execution 15% Correct calculations, working interactivity, proper data model
Business framework 20% Frameworks applied correctly and meaningfully
Insight quality 20% Goes beyond the obvious, finds the real story
Presentation 20% Clear narrative, audience-appropriate, actionable recommendation

Tips

Design for the audience, not for yourself

The CEO doesn't care about your DAX skills. They care about whether the business is healthy. Build the dashboard they'd actually use, not the one that shows off your technical ability.

One message per slide

The most common presentation mistake is cramming. If a slide has three messages, it has zero. Split it.

Lead with the headline

Don't build up to your insight — state it first, then support it. "Churn is concentrated in our first 90 days — and it's costing us £40k/month" then show the chart.

Submission Checklist

  • [ ] Dashboard published (Power BI Service link or Tableau Public link) or .pbix/.twbx file submitted
  • [ ] All KPIs calculate correctly
  • [ ] Interactive elements work
  • [ ] 5-slide presentation (PDF or PPTX)
  • [ ] One clear, specific recommendation
  • [ ] (Optional) Recorded 3-5 minute walkthrough

← Week 01 Assignment · Next: Capstone Projects →