FMCG supply chain

Uplifting planning and replenishment capability for a leading regional fmcg group

Trusted Data were asked to support a regional FMCG group with a major supply chain improvement programme, developing an AI planning & replenishment solution to reduce waste in a direct to store distribution environment

How we recovered 4% net profit for the Client in year one

Trusted Data supported a well established food and dairy group to overcome a long-standing issue that had impacted bottom line performance.

Supplying over 35,000 retail outlets across several countries, in a highly competitive regional market, supply chain and sales needed to find common ground to ensure that retail outlets were not under-supplied and that waste, brought about through store-level overstocking, was minimised.

Several global supply chain solutions had already failed to successfully overcome this problem for the Client.

Trusted Data were therefore tasked to develop a solution, which could forecast demand by outlet, by SKU, on any given delivery day.

 

Machine Learning fuelling more accurate forecasts and a dynamic replenishment approach

Challenge

The Client was supplying (direct-to-store) fresh produce to a huge network of multi-format retail outlets and was liable for overstocking (i.e. unsold, expired goods on-shelf).

Highly competitive shelf space and so cutting supply to reduce waste would impact market share.

SKU’s had a varying shelf life and Customer accounts, based upon format / size, had different expiry terms with our Client.

Replenishment schedules, which had pre-determined delivery frequency to each outlet, were not effectively aligned with store demand, often leading to unprofitable store deliveries.

The Client wanted to reduce on-shelf waste by over 50%, yet maintain a competitive market share.

Solution

Trusted Data developed a specific solution methodology based upon the Clients’ business requirements and internal forecasting and replenishment processes.

A 6-week process to transform missing and anomalous data, which had not been adequately addressed during prior solution implementations.

SKU-level forecasts with the predicitve ability to adjust to actual store-level demand and replenishment patterns, achieving upwards of 94% accuracy.

Replenishment schedules, by store, re-engineered to align with store demand, to eliminate non-value add deliveries that did not meet a specified revenue threshold.

Deployment of forecasting and replenishment models, into the Clients’ sales & planning module, which then connected with the Distribution team and Salesman handheld devices.

Active monitoring for a brief period, post implementation, to ensure shelf replenishment compliance with the forecast models to eliminate stock dumping.

 

Annualised Savings in year one

$26 mn

Reduction in aggregate waste level

+55%

SKU / Store forecast accuracy

94 – 98%

Annual Planning ‘man-hours’  saved

16,200