Significant savings identified & strong revenue growth in selected segments through improved targeting
Who are you?
Speedy Hire PLC is a FTSE-listed equipment rental and support services provider based throughout the UK, Ireland and the Middle East. We serve customers across multiple industries, including construction, infrastructure, industrial, manufacturing, facilities management, retail, leisure, and events, as well as local trade.
We operate out of 220 depots across the UK and Ireland and manage a hire fleet of around 3,000 different products. Demand forecasting and inventory optimization can have a significant impact on our business efficiency, as does our ability to generate and convert new business leads into revenue.
What was your challenge?
We aim to provide our customers with excellent service through having the right assets available when and where they are required. This has led us to achieve a strong market position with outstanding brand advocacy. However, our challenge was to optimize our forecasting and inventory management while maximizing revenue. We needed to be able to hold minimal assets to reduce costs, while also maintaining service levels and ensuring that product availability was high.
What did Peak do?
We are working with Peak, an Amazon Web Services (AWS) Advanced Machine Learning and Retail Competency Partner, to improve two significant elements of our business. We’ve defined these areas as inventory/asset optimization and revenue maximization.
Peak worked closely with Speedy’s asset team to understand vital business goals and processes. They have developed a range of bespoke forecasting algorithms based on factors such as; product popularity and seasonality to optimize our forecasting accuracy and inventory holdings across thousands of SKUs.
Our data is ingested into Peak’s Decision Intelligence system, CODI, and stored in Amazon Simple Storage Service (S3). It is processed using Apache Spark, running on EMR. Peak uses several AWS services to power CODI, including Elastic Compute Cloud (Amazon EC2), Lambda, Fargate, and SageMaker.
Peak and its AWS-powered AI solutions are helping us to maximize revenue through better management of lead generation and lead scoring. Previously, we segmented customers based on historical spend and region, but now, thanks to AI, we’re able to reclassify customers based on their potential opportunity.
Speedy Forecasting Algorithms
- Seasonality – determines product stock levels based on historical seasonal demand patterns.
- Anomaly detection – ignores spikes in the training data as well as significant changes in the demand profile, which may skew the true demand.
- Stock Levels – sets different stock levels dependant on the different products and locations. Some need to be available to customers immediately, whereas others can have some lead time.
- Substitute Products – will consider similar products when necessary.
What’s the upshot?
Peak has developed a strategy that supports our current processes and is compatible with our existing business systems. It is also flexible enough to change and adapt as the company grows. Peak spent a lot of time understanding our business and core processes, ensuring the solution considered key operational constraints that could not be changed.
We are implementing their recommendations and to date have experienced no dip in day-to-day performance. We are benefiting from being an AI-driven business, with significant asset savings identified and, through re-segmentation, strong revenue growth in focused segments.