Predictive Demand and Inventory Optimisation

Synopsis

Trusted Data were called to support a regional partner of Nissan in the Greater Asian region with a network of 55 dealerships. With a business model driven on volume with relatively conservative net margins, there was significant pressure on optimising inventory levels. Having the right end-customer configuration in stock would substantially boost sales but unattractive configurations placed a substantial burden on working capital and profitability. Given the multiple vehicle configurations, the client was not dealing with a uniform product-level forecast and so the problem was ever-more complex. Trusted were engaged to navigate the client to a cost-optimised solution.

Client Overview

🔸c.19% cumulative market share for consumer vehicles and HGV’s

🔸Sourcing vehicles from Japan/Thailand Distributing vehicles to +55 dealerships 

🔸Margin was c.5% – 6% above cost, thus forecasting was critical to the client

Client Problem

“A high level approach to vehicle sales demand forecasting was not capturing localised demand, which was driving a signifcant forecast error and creating a sizeable build up of unsold vehicles”

 

Analytics Delivery

The data modelling focused around:

🔸Fitting Vehicle type & sales forecast model

🔸Combining the outcome for stock ordering

🔸Model validation on past & current data

🔸Model Implementation

🔸Training & Handover to Supply Chain

The Challenge

Forecasting completed quarterly for consumer vehicles and 6-monthly for HGV’s

Commercial contracts and subjective use of inappropriate prior information led to excessive operational costs for the client before Trusted Data were engaged.

This resulted in:

⦿  41% forecasting error on quarterly sales for consumer vehicles

⦿  Redundant inventory tying up working capital

⦿  Lost sales opportunities from incorrect vehicle-to-dealer allocation (model & volume)

Trusted Data Approach

Trusted Data tackled the most pertinent issues:

⦿  What type of vehicle variants would sell in the next 6 months?

⦿  Identify the variants by vehicle type

⦿  Where and what quantity of vehicles to move to regional depots based on dealer predicted demand?

⦿  What is the right quantity of vehicles to be available at each dealership within a given time period?

⦿  Embedding forecasting models into procurement order cycles

⦿  Tracking ordering patterns for forecasting model compliance

OUR IMPACT

%

Reduction in Quarterly Forecasting Error

Unlocked working capital resulting from improved demand forecasting