Capstone 6 — Social Media Analytics Dashboard¶
A self-directed capstone analysing social media performance — engagement, reach, content effectiveness, and audience growth. The most modern and creative capstone, ideal for analysts targeting marketing, growth, or content roles.
The Brief¶
A brand posts across social platforms but doesn't know what's working. They want a dashboard that shows engagement trends, identifies the best-performing content, reveals the best times to post, and recommends a content strategy.
Tools¶
Python (Pandas) · SQL · Power BI/Tableau
Difficulty¶
Intermediate — engagement metrics and time-based analysis; the challenge is defining good metrics from noisy data.
What You Must Deliver¶
- A clean dataset of posts with engagement metrics
- Engagement rate analysis (not just raw likes)
- Content type and timing analysis
- Audience growth trend
- A dashboard
- A content strategy recommendation
Suggested Datasets¶
- Social media post datasets on Kaggle (Instagram, Twitter/X, etc.)
- A platform's exported analytics (your own, if you have an account)
- A synthetic dataset (posts, timestamps, likes, comments, shares, reach)
Key Metrics to Define¶
| Metric | Formula | Why it matters |
|---|---|---|
| Engagement Rate | (likes + comments + shares) / reach | Normalises for audience size |
| Reach | Unique accounts that saw the post | Distribution |
| Engagement per post type | Group by content type | What format works |
| Best time to post | Engagement by hour/day | Timing strategy |
| Follower growth rate | (new − lost) / starting followers | Audience health |
Raw likes are a vanity metric
A post with 1,000 likes from a 1M-follower account (0.1% engagement) underperforms a post with 100 likes from a 1k account (10% engagement). Always normalise engagement by reach or followers — this is the analyst's value-add. See Interview-Preparation/kpi-metrics.
Business Questions¶
- What is the average engagement rate, and how does it trend?
- Which content types (image, video, carousel, text) perform best?
- What is the best time/day to post?
- Is the audience growing, and how engaged is it?
- Which specific posts were the biggest wins and flops, and why?
Requirements Checklist¶
- [ ] Engagement rate calculated (normalised, not raw counts)
- [ ] Content type performance comparison
- [ ] Timing analysis (best hour/day to post)
- [ ] Audience growth trend
- [ ] Dashboard with engagement trends and content breakdown
- [ ] A data-driven content strategy recommendation
- [ ] Portfolio README
Evaluation Criteria¶
| Criterion | Weight |
|---|---|
| Metric definition (normalised engagement) | 25% |
| Content & timing analysis | 20% |
| Audience growth analysis | 15% |
| Dashboard design | 20% |
| Strategy recommendation | 20% |
Stretch Goals¶
- Sentiment analysis on comments (NLP)
- Correlation between posting frequency and growth
- A/B test design for content experiments
- Predict engagement for a planned post based on type/timing
You've reached the end of the course
Six capstones, seven guided projects, two weeks of foundations. Pick one capstone (or a guided project), polish it to portfolio quality, and use it to land your first analytics role. The skills are real — now go prove them.