System architecture
The system runs in four layers.
1. Data

A BigQuery warehouse uses native Search Console bulk export, preserving history beyond the 16-month limit. Google Trends, clickstream, and internal commercial data (taxonomy, margin, revenue) are integrated and normalized on a common time axis.
2. Modeling

The model calculates category demand trajectories and identifies seasonal start dates. It also distinguishes between narrow and broad demand windows using a dedicated concentration metric.
3. Intelligence

Automated SERP analysis runs on the top-10 results per category, evaluates competitive positioning, identifies visibility gaps, and produces per-category briefs. Margin and revenue data feed into the composite Opportunity Score.
4. Output

The final result is an easy-to-use dashboard. It lets you sort categories by demand trends, urgency, and how hard it is to compete. The system keeps learning and updating as new data comes in.

