Demand Forecasting in The Time of COVID-19: Navigating Demand Volatility Using Advanced Analytics


Demand Forecasting in The Time of COVID-19 | Case Study
  • CASE STUDIES
  • January 11th, 2021
  •   10695 Views

Overview


Shock events deeply impact and fundamentally change organizations across dimensions such as customer engagement, operations and supply chain, and finance, racing them towards a new normal.
Demand forecasting is the key to a seamless supply chain performance. Yet, forecasting demand accurately has always been challenging for businesses, even more now due to the uncertainty brought along due to COVID-19.
It is imperative for businesses to focus on building a data-driven demand forecasting model to create a resilient supply chain and minimize revenue loss through optimized demand sensing.
We developed a self-serve simulation tool to help one of the largest medical device manufacturers efficiently monitor and mitigate underlying risk owing to the uncertainty caused from the COVID-19 situation.

The Problem


Since March 2020, the manufacturer’s business processes, responsible for demand forecasting of multiple product lines, began to falter significantly as they were not designed to predict the COVID-19 impact.

COVID-19 Impact on Demand Forecast:

  • Inability to estimate impact on supply chain using existing demand forecasting construct
  • Significant amount of variability in the current process

Key problem areas:

  • Identifying parts of the portfolio that have been affected due to COVID-19
  • Estimating the degree of impact at a product level
  • Identifying the recovery period for COVID affected products
  • Forecasting future COVID related fatalities and recovered cases for each country
  • Tracking and comparing demand sensing forecast vs baseline forecast with the help of key metrics

The Mu Sigma Approach


Understanding the client’s current state and the desired future state helped us bridge the gap by answering the right questions. Defining the inputs and outputs required to answer key questions led to an improved understanding of the desired outcome:

Current State:

  • Inability to estimate the impact of COVID on supply chain using the traditional demand forecasting construct
  • Significant amount of uncertainty in the current processes due to COVID

Gap:

  • Lack of a method to accurately forecast demand by accounting for COVID 19 factors and recovery trends

Desired State:

  • Creating a post-pandemic sustainable process for various regions the client supplies to
  • Providing directional insight into length of COVID impact across regions
  • Accurate capture of the effect of COVID – 19 on demand signals
  • Establishing a good baseline for providing reliable forecasts

The Solution


To achieve this, we designed a solution across different phases to help the client improve inventory and production planning during the lockdown and the recovery phases to create a resilient supply chain.

Phase 1:

Identification of recovery shape across regions to effectively capture recent trend in data using ML models

  • Key Factors included: lockdown strength, lockdown length, number of active covid-19 cases, death rate, social distancing score, reproductivity index, population density, rate of testing, government policies

Phase 2:

Developing a COVID – 19 Forecast Simulator


We developed a self-serve simulation tool to help the client mitigate the underlying risk owing to the COVID-19 situation. The right mix of statistical modelling, machine learning and business inputs was used to enable accurate demand forecasting for clients, using an advanced demand sensing framework.

Intelligence Layer:

  • The overall process was integrated into a simulation tool, according to the current technical construct.
  • Data procurement, QC framework and Data preparation for modelling
  • Incorporation of relevant COVID features which could potentially affect demand signals
  • Modulating the features to feed them as an input for the model
  • Algorithm to forecast the demand numbers at requisite levels
  • Data warehouse encompassing relevant data for the visualization layer
UI Front End


A simplified interface was developed to assist business users in demand planning. User-interactive frontend which forms the basic encapsulation of the tool.
The user has the ability to simulate the factors and visualize potential results as needed.

Simulation Tool


A Web app was then developed which allows the users to tweak potential features and visualize its impact accordingly.

Tool Features

  • Uber-level reporting
  • Simulate numerous factors and visualize potential scenarios
  • Ability to download forecast outputs and integrate into other systems as needed

The Impact


The agile tool can be scaled across other areas/teams, bolstering future in-house centralized capability.
Mu Sigma’s structured framework has helped the client respond to the shock event and will further enable them to land on their feet post crises, strengthening their supply chain. The new prototype model/process has been developed based on a demand sensing framework that accounts for causal factors that the naïve models could not account for.
This new model has helped achieve 35 – 40% improved demand forecast accuracy in predicting product demand.
Effective governance structure, enhanced digitalization, and a continuous feedback mechanism will be vital in the successful implementation of this framework, especially during continued/expected business disruption from COVID 19 in the next 6+ months.
In this hyper-evolving VUCA world, Mu Sigma acts as a Transformation Sherpa to 140+ Fortune 500 clients to help reimagine the way they compete. Organizations out-collaborate with us through a unique engagement model consisting of an interactive ecosystem of people, processes & training, and platforms, which enables them to build capabilities in the long term while creating impact in the short term. Want to know more?
Write to us at TheSherpa@mu-sigma.com

Click here to know about how Mu Sigma helped a leading electronic manufacturer to forecast its repairs to improve its reverse logistics supply chain efficiency.