Skip to content

    15.04.2021 — 5 min read

    Bringing demand forecasting to another level

    Over the past 15 years, retailers and especially food retailers have been trying to find and adapt their models to the online economy. I remember having my first online food order delivery at home from one of the French pioneers just before I moved from France in 2005.

    The idea was simple. In Paris area, there were not so many middle size supermarkets, but lots of more expensive convenient stores and the fact that you might want to purchase for instance bottled water, as the water quality was not that great. It had everything on the paper to become a revolution. Saving time, no need to carry heavy goods with you in public transport and potentially cheaper products even with the delivery cost.

    In 2020 we were far from adoption of online shopping concerning the food retailers, even if numbers of customers using click-and-collect or drive-in in other countries than Finland was slowly increasing. As a foreigner, I could spot the differences in shopping habits here in Finland – more small to medium size supermarkets and fewer hyper markets, where people in France would traditionally spend lots of time and shop food for the coming week or two, without the need to come back. Here in Finland people would have smaller baskets per shopping and would come back more often. Only a few customers would be using online ordering for food shopping until world was hit by COVID-19 pandemic.

    Everyone, both retailers and customers, had to adapt and change their shopping habits (less often, closer location). Lots of key players offered even call service for the older people, who were not so familiar with online shopping. New players came to the table to propose not only restaurant's delivery but order delivery from shops and supermarkets. Those new players like Foodora and Wolt are now even opening their own dark stores. This will again ask the sector to adapt. One of the key issues for the supply chain is how to predict the demand and how to forecast. How to make sure you have the right quantity of products at the right time? The retail industry and the software industry have been developing tools to improve their inventory management. Better forecasts improve the wastage, which is not only good for retailer’s economy, but also participates to more sustainable practices. Better forecasts avoid out of stock and missed sales.

    We have seen an accelerated interest of demand forecasting tools and services around it. Especially around one of our partners latest innovation based on their long knowledge on demand forecasting tools combined with the power of AI. The new tool of Symphony Retail AI brings demand forecasting to another level as it removed some of the biggest pain point for our customers - data cleansing and data aggregation. This huge task can now be done by the AI and let forecaster to concentrate on added value tasks. During exceptional conditions as the COVID-19 pandemic, it is important to understand, if peaks in product consumption are cause of an exceptional behaviour moment or is there a trend for a product at the same time. The more data is submitted to the AI, the better the forecast will be accurate as pattern and external factors can be identified. Weather, promotions, holidays – all those factors have impact on the demand.

    The epidemy has shown changes not only in the consumption trends, shop’s customer behaviours and services proposed by retailers, but also how fast retailers need to adapt. More and more projects are turning to agile method that allows our own customers to be at the centre of those changes. It helps them to pick the most valuable functionality for their business with a shorter delivery time scale, and to implement new tools and processes faster and in a more modulable way that support their customer and their business challenges.

    eCommerce, Omnichannel commerce, Supply chain management