Real data with Real analysis. No made-up numbers
Most digital marketing students practice analytics on hypothetical data or screenshots from tutorials. I wanted to work with real e-commerce data to track traffic patterns, conversion behaviour, revenue, and produce an insights report that could actually sit in front of a marketing team.
I used Google's own Merchandise Store demo property. A real e-commerce website where Google sells branded products with months of live data publicly accessible for analysis. The result: a client-ready Looker Studio dashboard and a four-finding analyst-style insights report.
Built to be interactive, not just pretty
Live Looker Studio dashboard — 803K views, $444K revenue, sessions over time, traffic sources breakdown
The dashboard was built to answer the questions a marketing team would actually ask instead of just displaying numbers. Six components: KPI scorecards (Users, Sessions, Engagement Rate, Revenue), sessions trend over 90 days, traffic source breakdown, top pages table by views and engagement time, and an e-commerce funnel showing Add to Carts → Checkouts → Purchases.
Why Looker Studio over a static PDF: A live dashboard with a shareable link means anyone can explore the data themselves. The date range control lets them adjust the period. That interactivity is the difference between a portfolio piece and a portfolio piece that gets clicked.
4 analyst-style insights with real recommendations
Homepage is losing visitors before they shop
88K homepage views, 7-second average engagement time, highest bounce rate on the site. 77% of homepage visitors never reach the product catalogue. The homepage attracts traffic but fails at its primary job — directing people towards the products.
New arrivals page is where high-intent customers concentrate
15K views but 75K event count. Each visitor interacts ~5 times. Lower bounce rate than homepage. These are brand loyalists checking for new drops. They're being underserved by having no dedicated traffic strategy pointed at them.
1.01% purchase conversion rate vs 2–4% industry benchmark
At 78K active users, moving from 1.01% to 2% would double purchase volume without acquiring a single new visitor. Combined with 0.6% Day-1 retention — there's no re-engagement mechanism. No cart abandonment email. No retargeting.
High-engagement apparel audience isn't converting but stationery impulse buyers are
Men's apparel is the most browsed category with returning visitors. But the top-selling product is a sticker sheet. Two completely different buyer types sharing one store, each requiring a different conversion strategy. The store is optimised for neither.
What was built
What analytics actually taught me
The biggest shift this project created in how I think: numbers don't tell you what's wrong. They tell you where to look. A 1.01% conversion rate is a number. "The homepage has a 7-second engagement time and a 77% drop-off to the product catalogue" is a diagnosis. The analysis work is translating one into the other.
The Finding 4 insight — apparel browsers vs stationery impulse buyers — wasn't in any metric on its own. It came from connecting the engagement data (apparel), the revenue data (stationery), and understanding human purchase psychology. That connection is the actual skill.