From 50% to 90%+ SLA: Faherty fixes BFCM CX Planning with Saras ML-Based CX Staffing

Faherty is an omnichannel apparel brand operating across eccommerce and approximately 75-78 physical retail locations nationwide. The brand runs high promotional intensity during BFCM and holiday windows, concentrating its highest order volumes, customer expectations, and CX load into a compressed multi-week period.

Faherty's CX team entered every peak season with the same reactive planning method: historical ticket averages adjusted by judgment.

• SLA compliance fell to 50-60% during the highest-volume weeks of the 2024 BFCM season.
• Ticket backlogs exceeded 1,000 open items with no early warning to trigger staffing action.
• Corrective staffing arrived too late because the team responded to volume already in the queue.
• No link between the promotional calendar, carrier performance signals, and CX staffing decisions.
• Both overstaffing and understaffing were equally likely without a forward-looking demand signal.

‍

Saras Analytics unified Faherty's order, promotional, carrier, and CX history data into a single connected layer. On top of that foundation, Saras built an ML-based forecasting model calibrated to Faherty's specific peak-season patterns, producing day-level staffing recommendations weeks ahead of demand.

• Data foundation: Unified four previously siloed data streams into a single connected layer for the first time.
• Demand forecasting: ML model produced daily forecasted CX demand volume grounded in orders, promotions, carrier risk, and historical patterns.
• Staffing recommendations: Translated demand forecasts into recommended daily headcount based on handle time and agent utilization benchmarks.
• Precision deployment: Model distinguished genuine understaffing from recoverable backlog accumulation to prevent unnecessary over-hiring.
• Recalibration loop: Weekly forecast-vs-actual reviews refined model assumptions throughout the season, reaching 2-8% variance by period end.

‍

The 2025 holiday season delivered improvement across every dimension of CX performance. Faherty held service levels during the highest-volume weeks in its calendar and entered 2026 planning with a reusable forecasting infrastructure that improves in accuracy with each additional season of data.

• SLA compliance reached 90%+ throughout peak weeks, up from 50-60% the prior year.
• Peak backlog dropped 85%, from over 1,000 open tickets to approximately 150.
• Monthly forecast variance held within 2-8% of actual CX demand under live operating conditions.
• Staffing plans confirmed and resourced weeks before peak windows instead of days after SLA breaches.
• Planning infrastructure is reusable across future peak seasons, compounding in accuracy over time.

‍