Branch Footfall Forecaster
Forecasts hourly branch walk-ins so managers can roster staff.
MAPE (daily, per branch)
8.7%
MAPE (hourly, per branch)
17.2%
Peak-hour hit rate
0.86
Average queue time change (pilot)
-19%
About this model
Forecasts hourly walk-in volumes by branch and service type (teller, customer service, safe deposit) for the next four weeks, so branch managers can roster staff and open counters where queues will form. Accounts for payday, school holidays, festive seasons and state holidays.
Intended use
Branch staffing and counter planning; network reviews of branch format. Uses queue-ticket counts only, with no customer identifiers.
Training data lineage
Hierarchical forecast (branch by service by hour) trained on branch-footfall-daily queue-ticket data in BigQuery, with branch attributes from branch-atm-network. Runs weekly in Vertex AI Pipelines; forecasts land back in BigQuery for the branch operations dashboard.
Limitations & bias notes
Branches that changed format (e.g. to self-service) in the last three months use the pre-change pattern until retraining. Sudden events such as system outages or new product launches are not anticipated.
Ownership and sensitivity
Who approves access
No approval needed. Self-service: subscribe and use, logged for chargeback.
In Safe to try
Open to all staff. Aggregates, policy text, public series and synthetic data.
Self-service: subscribe and use, logged for chargeback.
- Aggregated only
- No row-level customer data; aggregates with small-cell suppression.
Try it
Live sandboxbranch=KL-014&horizon=12
fixtureFeedback
Details
- Updated
- 2026-06-29
- Latest version
- 1.3.0
- Licence
- Group Reuse
- Access
- Open to all BUs
- Framework
- statsforecast
- Language
- N/A