Rolling Demand Signal Variance Tracker

When a manufacturing network faced chaotic demand swings, I built a dynamic dashboard to visualize and track week-over-week forecast volatility. This tool helped uncover root causes, empower proactive planning, and align executive decisions—transforming how the company manages supply chain uncertainty across three major production plants.

Situation

A global manufacturing company with three major production plants was struggling to meet fluctuating customer demand, which had remained unpredictable for an extended period. Leadership believed the issue stemmed primarily from a lack of production capacity, leading them to consider expensive investments in additional machinery and labor.

However, planning teams suspected deeper, more systemic problems within the forecasting process itself. They observed that forecasts were swinging by as much as 300% week-over-week, creating chaos in operations and planning efforts. This volatility was particularly concerning, as it included major changes within a 4-week frozen window. This is a critical time frame during which orders are confirmed and capacity should remain stable. Previously, the planning team had no effective way to visualize or measure these erratic changes over time, leaving them reactive and scrambling to catch up the week of or before production.

Procurement, scheduling, and materials management departments found themselves constantly adjusting purchase orders, production plans, and inventory targets—often too late to prevent either overstock or stockouts. These ripple effects impacted customer delivery performance, internal confidence in planning systems, and resource utilization across all three facilities.

Task

My goal was to design a standardized tool that could:

Track changes in weekly demand over time to identify trends, gaps, and volatility
Allow stakeholders across plants to drill down the information to identify root causes
Quantify how demand fluctuations were impacting procurement, planning, and operations during both tactical and frozen horizons
Provide leadership with simple visual cues and reporting summaries aligned to their existing monthly review cadence

In short, the task was to create a single source of truth that would serve the needs of daily planners, site leaders, and corporate stakeholders alike—bridging the gaps between demand signals, operational capacity, and executive decisions.

Action

To solve the issue, I built an interactive 18-week rolling forecast variance tracker using Excel and Power BI. This dashboard became a powerful hybrid tool that allowed both high-level summary and granular root-cause investigation.

I first structured a matrix that compared weekly forecast versions side by side, showing how projected demand shifted over time. Each row represented a forecast version (e.g., 18 weeks out, 17 weeks out…), and each column represented a specific future week. This visual approach made it easy to spot when demand for a given item spiked suddenly.

I integrated conditional formatting thresholds to highlight critical shifts in demand: yellow flags for week-over-week changes >10%, and red flags for deltas >25%. This allowed planners and procurement teams to immediately identify items or production areas at risk.

To enhance interactivity, I introduced dynamic slicers that allowed users to filter by:

Business aggregate (e.g., customer segment)
Production plant
Product-level identifier

I also enabled toggles between different forecast types: Priority 1 vs. Priority 2, so teams could simulate what loading would look like with or without certain customers included in the mix. This was crucial for testing capacity assumptions and visualizing the impact of new customer orders or lost business.

A 4-week bucket comparison feature was created specifically for executive stakeholders. It showed month-over-month shifts, summarizing which aggregates or facilities saw the largest swings. In one instance, this revealed a 29% increase in forecasted units in just a single week, helping build a case for better upstream controls.

Result

The tool quickly became a critical resource across planning, procurement, and leadership teams. Some of the key outcomes include:

Root Cause Discovery: Uncovered systemic errors in the demand pipeline, where backorders were loading into the current week instead of being spread across the forecast horizon. This prompted IT remediation and omittance from trend calculations.
Proactive Capacity Planning: Site leaders began adjusting tooling and mold availability sooner, reallocating resources between plants based on early detection of overloads.
Cross-Functional Visibility: Procurement and materials teams gained early insight into upcoming demand shifts, allowing them to smooth out supplier schedules.
Executive Alignment: The monthly summary page allowed corporate leadership to visualize long-term trends and evaluate planning effectiveness at a glance.
Improved Accountability: Because the dashboard retained weekly snapshots, teams could now track when forecasts changed, by how much, and who owned the signal. This transparency improved planning discipline and encouraged more deliberate forecasting.

What started as a tactical tool became a scalable, data-backed foundation for improving supply chain responsiveness across the organization. It is now being considered for deployment across additional business units as a best-practice forecasting interface.

Key Skills

Operational Horizon & Tactical Planning Constraints
Forecast Analysis & Demand Planning
Excel Automation & Dynamic Modeling
Root Cause Problem Solving
Data Visualization & Dashboard Design
Cross-Functional Communication
Scenario Comparison & Tiered Planning
Executive Reporting & Stakeholder Alignment

Disclaimer: The company name and proprietary business information have been intentionally omitted from this case study. All details are presented strictly for academic, professional, and educational purposes only. The solutions, data representations, and outcomes described are shared to illustrate process improvement methodologies, technical skills, and strategic problem-solving, not to disclose or represent any confidential or proprietary information.