Toffee Inc.: Demand Planning for Chocolate Bars Custom Case Solution & Analysis

1. Evidence Brief: Case Extraction

Financial Metrics

Forecast Error: Mean Absolute Percentage Error (MAPE) consistently exceeds 25 percent for core product lines.
Inventory Costs: Excess stock of slow-moving items occupies 30 percent of warehouse capacity during off-peak months.
Revenue Loss: Estimated 15 percent of potential sales lost due to stockouts during peak holiday seasons (October to December).
Margin Pressure: Expedited shipping costs to fulfill late orders reduce gross margins by 400 basis points on affected batches.

Operational Facts

Production Cycle: Fixed 4-week lead time for raw material procurement and a 2-week production window.
Product Range: Two primary SKUs (Toffee Crunch and Toffee Smooth) account for 85 percent of total volume.
Distribution: Centralized warehouse serving 12 regional distribution centers.
Data Source: Demand planning relies on a 3-month simple moving average based on lagging sales data.

Stakeholder Positions

Sanjay (Demand Planning Manager): Advocates for advanced statistical modeling but lacks the budget for specialized software.
Sales Team: Consistently overestimates demand by 10 to 20 percent to ensure high service levels and avoid stockouts.
Finance Department: Prioritizes inventory turnover and pressures operations to minimize working capital tied up in safety stock.
Production Team: Prefers long, stable production runs to maximize equipment utilization.

Information Gaps

Competitor Activity: The case provides no data on competitor pricing or promotional schedules during peak periods.
Customer Granularity: Lack of data regarding point-of-sale (POS) metrics from major retail partners.
Cost of Capital: Specific internal hurdle rates for inventory carrying costs are not defined.

2. Strategic Analysis

Core Strategic Question

How can Toffee Inc. synchronize its disconnected sales and production functions to reduce the 25 percent forecast error and eliminate the 15 percent revenue loss caused by stockouts?

Structural Analysis

Demand-Supply Integration (DSI): The current system is fragmented. Sales incentives drive over-forecasting while Finance drives under-production. This misalignment creates a bullwhip effect.
ABC Analysis: Toffee Crunch and Toffee Smooth are Class A items. Current moving average models fail to capture the 40 percent demand spike observed in the fourth quarter.
Value Chain: Inefficiency at the inbound logistics and operations stages (due to poor forecasting) forces high-cost outbound logistics (expedited shipping), eroding the value proposition.

Strategic Options

Option	Rationale	Trade-offs	Resource Requirements
Statistical Model Overhaul	Replace moving averages with Triple Exponential Smoothing to account for seasonality and trends.	Requires historical data cleaning; does not address sales team bias.	Data scientist consultant; 4 weeks of historical data processing.
Collaborative Planning (CPFR)	Integrate POS data from top 3 retailers to gain real-time demand visibility.	High dependency on external partners; requires IT integration.	EDI software; dedicated account managers for retail coordination.
Inventory Buffer Strategy	Implement dynamic safety stock levels based on service level targets (98 percent for Class A).	Increases working capital requirements in the short term.	Revised inventory policy; Finance department approval for higher capital allocation.

Preliminary Recommendation

Toffee Inc. must implement a formal Sales and Operations Planning (S and OP) process combined with Triple Exponential Smoothing. This addresses both the technical failure of the moving average model and the behavioral failure of misaligned departmental incentives. Fixing the model alone will not stop the sales team from inflating numbers; fixing the process alone will not account for seasonality.

3. Implementation Roadmap

Critical Path

Phase 1 (Days 1-30): Establish a cross-functional S and OP committee. Define a single version of the truth for demand data, removing the separate spreadsheets used by Sales and Finance.
Phase 2 (Days 31-60): Transition to a Holt-Winters forecasting model. Run this model in parallel with the current moving average for one production cycle to validate accuracy.
Phase 3 (Days 61-90): Link the validated forecast directly to procurement schedules. Adjust safety stock levels specifically for the October-December peak based on the 98 percent service level target.

Key Constraints

Data Quality: Historical sales data contains noise from previous stockouts, which may skew future projections if not manually adjusted.
Organizational Silos: The Sales Team may resist a unified forecast if their performance bonuses remain tied to service levels without accountability for forecast accuracy.
Lead Time Rigidity: The 4-week procurement window limits the ability to react to sudden market shifts, making the accuracy of the initial forecast the primary determinant of success.

Risk-Adjusted Implementation Strategy

The plan assumes a 10 percent margin of error in the new model. To mitigate this, Toffee Inc. should maintain a 5 percent strategic reserve of finished goods at the central warehouse. If the new model over-predicts, the S and OP committee will trigger a production slowdown in the following month to prevent warehouse congestion. If it under-predicts, the reserve serves as the first line of defense before resorting to expedited shipping.

4. Executive Review and BLUF

BLUF

Toffee Inc. must immediately transition from reactive moving-average forecasting to a formal Sales and Operations Planning (S and OP) framework. The current 25 percent forecast error is a self-inflicted wound caused by departmental silos and inadequate statistical tools. By implementing Triple Exponential Smoothing and a unified demand signal, the company can capture the 15 percent in lost holiday revenue and reduce expedited shipping costs. Success requires re-aligning sales incentives with forecast accuracy rather than just volume. Execution must begin within 30 days to stabilize the supply chain before the next peak season.

Dangerous Assumption

The analysis assumes that the 15 percent lost sales are entirely due to stockouts. It fails to account for the possibility that a portion of this demand is transitory or brand-switchable, meaning customers may not return even if availability improves. If brand loyalty is lower than assumed, the ROI on high safety stock levels will diminish.

Unaddressed Risks

Supplier Failure: The plan relies on a fixed 4-week procurement window. Any disruption at the raw material level (e.g., cocoa price spikes or labor strikes) renders the improved forecast moot. (Probability: Medium; Consequence: High).
IT Integration Lag: The transition to advanced modeling depends on clean data. If the existing ERP system cannot export granular data, the 90-day timeline is unrealistic. (Probability: High; Consequence: Medium).

Unconsidered Alternative

The team did not evaluate a SKU rationalization strategy. By eliminating the bottom 15 percent of low-volume, high-complexity SKUs, Toffee Inc. could simplify its production environment, increase the capacity for Toffee Crunch and Toffee Smooth, and naturally reduce the forecasting burden without investing in new software.