Customized Demand Planning Solution for FMCG Leader in the UK

Digital Manufacturing Case Studies

Request for Quote »

Project in Brief
Challenges & Needs
Solution & Approach
Business Impact
Project Samples

Project Overview

Client Profile

Consumer goods company

Location

United Kingdom

Industry

FMCG

Inputs

Forecast quantity vs. actual sales reports

Deliverables

Smart forecasting solution

Business Needs

The health, hygiene and nutrition products company failed to predict weekly and monthly demand accurately for each SKU.

They needed to:

Get real-time visibility and accurate status of stock level at various stores
Strengthen the forecast model that accounts for sales, trends and seasonality parameters
Ensure timely response to inventory optimization, procurement, and stock related concerns

Challenges

A thorough review of their existing forecast model posed challenges like:

Low product data visibility to assist in accurate forecasts across product lines.
No mapping between customer databases, supply systems and marketing insights.
Legacy system that couldn’t cope up with fluctuating demand patterns in different markets.
Difficulties to meet product standards and stocks due to erroneous inventory data and inefficient analytical tools

Our Solutions & Approach

Delivered a customized demand forecasting solution to integrate supply chain from raw material procurement, product availability, management and operations for all SKUs across locations.

Share Your Thoughts

With analytical insights to enhance their existing data models, it helped in accurate monitoring of stock levels and meet fluctuating future demands.

A team of Hitech data scientists analyzed the client’s existing algorithms and understood the exact pain points and needs of various stakeholders.
Identified the loops in the algorithmic data and implemented right course-correction approach. For example, used data cross validation and feature selection to bring more value to the weekly and monthly forecasts.
Used data pre-processing techniques to pace up existing databases, applied classical time series algorithms to enhance forecast accuracy.
Helped with automated forecasting support and seamlessly manage SKU-level forecast for multiple stores.
Owing to the success of the model in the USA market for several months, the model was deployed for different markets like UK, France, Brazil, Mexico and Australia.