SaaS Growth Analytics and Experimentation
Problem
Subscription data becomes difficult to use when revenue, churn, acquisition, and customer-value measures are calculated separately. I used a synthetic practice dataset to build one repeatable reporting workflow and examine how metric definitions affect management decisions.
What I built
- Cleaned and standardized source data in Google Sheets
- Loaded structured data into BigQuery and resolved schema and null-value issues
- Defined New MRR, Churn MRR, Ending MRR, logo churn, CAC, and LTV
- Created scorecards, trend views, and filters in Looker Studio
- Documented calculation logic so the dashboard could be reviewed rather than treated as a black box
Results surfaced
- Ending MRR increased from $77,800 in January to $176,200 in December, a $98,400 or 126.5% increase across the synthetic period
- Monthly views made revenue growth, churn, acquisition cost, and customer value comparable in one reporting system
- Experiment checks combined conversion significance with confidence intervals, effect size, and sample-ratio mismatch review
These results describe the supplied synthetic dataset. They do not represent verified company performance.
Analytical judgment
The main judgment was that a polished chart is not enough. Retention and revenue metrics require consistent cohort, period, and denominator definitions. Experiment results also require a sample-integrity check before a conversion difference can be trusted.
Tools
Google Sheets, BigQuery, Looker Studio, data cleaning, KPI definition, dashboard design, statistical testing, and business interpretation.