Data & Product Analyst
I find the insight behind the metric — and connect it to what the business should do next.
I’m a data and product analyst with a background spanning industrial process data, business intelligence, and product analytics. I’ve worked with datasets ranging from DCS/SCADA sensor streams at one of Africa’s largest fertilizer plants to e-commerce transactional data to Google Analytics sessions in BigQuery.
What I do consistently: take a business question, go into the data, find where the real problem is (which is rarely where people assume it is), and produce a clear recommendation backed by evidence.
Based in Lagos, Nigeria. Open to remote roles globally.
Query & Analysis: SQL (BigQuery, PostgreSQL, MySQL), Python (pandas, numpy, matplotlib, seaborn) Product Analytics: Funnel analysis · Cohort analysis · Retention analysis · User segmentation Visualisation & BI: Power BI · matplotlib · seaborn · Streamlit Tools: Google BigQuery · Git/GitHub · Jupyter Notebook · Notion
Python pandas matplotlib Jupyter
The question: Do most customers return after their first purchase — and if so, how quickly, and what makes the difference between those who stay and those who don’t?
What I found: 72.4% of customers are repeat buyers — nearly 3× the e-commerce industry benchmark. This is not a casual consumer base. It’s dominated by wholesale buyers and resellers who purchase in volume, return unsold inventory as part of their normal cycle, and are the most loyal segment in the dataset.
The most counterintuitive finding: customers with the most returns and cancellations retain 96.1%—the highest of any segment. Returns are a loyalty signal here, not a churn signal.
First-session behavior predicts long-term loyalty with measurable precision: repeat buyers spent 32% more and explored 19% more product categories in their very first session.
The recommendation: This is not a retention problem — it’s an acquisition targeting problem. The product works. The business needs to attract more customers who look like the 72.4% who stay.