How artificial intelligence can improve NPD and supply chain efficiency

By Katy Askew contact

- Last updated on GMT

FuturMaster AI can predict what will go in the grocery bag to strengthen NPD and supply chain performance ©GettyImages-a_namenko
FuturMaster AI can predict what will go in the grocery bag to strengthen NPD and supply chain performance ©GettyImages-a_namenko
FuturMaster has developed artificial intelligence software that, it says, can predict demand levels, improve supply chain efficiency and even forecast whether innovation will be successful.

FuturMaster provides supply chain planning software for CPG clients including the likes of PepsiCo, Lactalis and Danone. The company’s expertise spans supply chain solutions and demand forecasting. And, according to the company, it is now launching a tool that leverages AI to help companies successfully launch new products.

Each year more than 30,000 new consumer products hit the market – and 80% of these are destined to fail, research from the Harvard Business School suggests.

This is where FuturMaster believes its new software can help. The tech innovator has honed its sights on using AI to increase success rates in new product introductions.

In tests with a Chinese textiles manufacturer, FuturMaster’s AI programme was more accurate in over three-quarters of forecasting scenarios and twice as fast at interpreting data than teams of trained forecasters using manual methods.

Moreover, the automated forecasting solution was found to be ‘much better’ at predicting how sales of new products will typically evolve from the moment they hit the shelves. Trials showed that rival solutions were often too quick to replenish products and that the resulting costs of unsold stock could be reduced using AI-assisted tools.

This development has uses across the Fast Moving Consumer Goods (FMCG) sector. “It is applicable for any sectors with lots of new products introduced every year,”​ Gilles Lefebvre, product manager at FuturMaster, told FoodNavigator.

Past data points to future performance

The machine learning software works by using algorithms to group products based on the sales history and behaviour of items with similar characteristics, a process known as clustering.

Products can be clustered using a range of attributes, Lefebvre explained: “Products attributes can be the product group, the size, the colour, etc.

“The system will take the characteristics of the products - but also the customers, the product launch and so on – and identify in which clusters these characteristics are the most represented [in] to assign this product to this cluster. Thanks to that, we know that the sales profile of the new product is similar to the sales profile of the cluster.”

The data collected by customers is based on past sales data of similar products, clients or launches, Lefebvre continued. The autonomous system then applies a set of rules to automatically calculate volume requirements over a variable period, adjusting and constantly adapting over time.

Gilles Lefebvre
Gilles Lefebvre, FuturMaster

“Our objective at FuturMaster is to calculate accurate forecasts, which can be measured against actual sales during the launch period. The success of the launch itself is defined by our customers and we can configure these Key Performance Indicator in our application.”

The development has significant implications for the NPD process – enabling companies to forecast whether new launches will have sticking power and consumer appeal.

“Launching new products is a complicated process that involves numerous teams collaborating together and requires close monitoring of sales and demand, not to mention harvesting accurate, clean data. Decisions - which can make or break a company - have to be taken early on and it's an area ripe for the benefits of AI technology." 

AI for stronger supply chains

“It is essential for companies to know how the new product will behave (not only for the supply chain) and thanks to this module companies can calculate the forecast, understand it, change the proposition if necessary, and react if things are not going as expected,”​ Lefebvre told this publication.

And FuturMaster believes that its AI can be used for more than managing innovation. The same principle of using AI to forecast demand can also help build more efficient supply chains.

French vegetable giant Bonduelle is leveraging FuturMaster technology as part of a supply chain transformation project targeting reduce waste.

Through FuturMaster’s forecasting algorithms, Bonduelle can predict with ‘pinpoint accuracy’ what’s most likely to sell in different parts of the world, enabling tighter management of its supply chain across Europe, North America and Asia.

Bonduelle’s challenge: Complexity

Bonduelle generates sales of around €3bn, employs almost 15,000 people and operates 58 production sites across ten countries.

With over 130,000 hectares harvested and over three thousand farmers, Bonduelle produces around a million tonnes of canned goods a year, 450,000 tonnes of frozen goods and 350,000 tonnes of fresh produce.

Following a series of global acquisitions, including most recently the Del Monte business in North America, Bonduelle increasingly has to cater to different tastes in different markets. This has included a shift to more locally produced, healthier products.

“There are significant differences across markets. Countries can have different delivery and logistics setups, not to mention different product requirements,”​ said Nathalie Morandière, S&OP and Methods Leader at Bonduelle Europe Long Life.

Bonduelle cans
Bonduelle has a complex supply chain with multiple brands across three categories (canned, frozen and fresh vegetables)

‘We needed to transform our supply chain’

Morandière acknowledged that as Bonduelle grew through a series of acquisitions the group faced a new set of challenges.

“We needed to transform our supply chain with a customer-focused approach. Before, we were used to working in silos, by region or department,”​ she explained.

“We needed processes supported by technology to help us look forward over a minimum 18-month rolling period, so we could manage our businesses by looking at shared assumptions, risks and opportunities and be sure that we could take the necessary actions to close the gaps.”

Since deploying demand management processes supported by FuturMaster software, Bonduelle says it can optimise sales forecasts and synchronise demand across Europe, where it operates 14 factories and warehouses across five key regions.

The company completed the rollout of the software across all regions in 2018 for its retail and foodservice businesses. This year, it plans to focus on how improved integrated business planning to advance its sustainability objectives.

Digital transformation boosts accuracy

Bonduelle said the benefits of transitioning its supply chain to digital models that leverage AI are far-reaching. They include a ‘better understanding’ of consumer needs alongside a reduction in obsolete and slow-moving stock.

Moreover, Bonduelle said it has been able to use this forecasting system to launch ‘successful new products and innovations’ through improved sales monitoring.

Short-term planning accuracy is on track to increase up to 10% and long-term planning accuracy has remained stable and very high (above 85%), despite moving to a more decentralised approach, the company added.

Morandière did say that some challenges remain, including some ‘volatility in the product mix’, with smaller retailers and private label requiring a more ‘reactive approach’.

Related news

Follow us

Featured Events

View more

Products

View more

Webinars