Never has the grocery retail market’s focus on forecasting been more relevant. Growing concerns over food wastage combined with potential market disruption around Brexit and global trade wars are casting a spotlight on what has always been a major issue for retail supply.
Forecasting is key to driving quicker and better decisions, minimising cost and maximising availability. Yet despite emphasis and guidance from the Grocery Code Adjudicator (GCA) on the topic, there remains a huge amount of confusion over what is meant and no examples of best practice. From the vague and often contradictory definitions of a forecast to the diverse form in which information – if available – is provided by retailers, suppliers simply do not have a consistent, usable set of trusted insight to support better, more efficient and cost-effective operations.
However, by using existing data combined with best practice visual analytics, forecasting demand accurately and consistently is both possible and achievable. With the data, technology and expertise now readily accessible, it is possible to achieve the collaboration and consensus between retailers, suppliers and regulators that is required to change current thinking around forecasting, insists Guy Cuthbert, CEO, Atheon Analytics.
The grocery retail industry has never seen more talk of improving efficiency throughout the supply chain. The need is well understood. Facing escalating competition alongside rising uncertainty and global attention on waste, there is a strong desire to achieve far better control over every aspect of the supply chain. However, few retailers have yet to truly get to the crux of the matter. Arguably, across the board the single, most critical area of operational performance remains inadequate: the forecast. Despite the depth and breadth of sales and supply chain data now routinely collected by virtually every retailer, variable, inconsistent or inaccurate forecasting remains the norm.
The GCA’s June ’18 Best Practice Statement’s call for ‘greater transparency of [retailers’] communications with suppliers about forecasting, to allow suppliers to meet orders and to anticipate and calculate the full costs of supply’, makes plain the lack of progress made in recent years. Indeed, while advice from the GCA in March 2016 suggested that all ten of the largest UK supermarkets which are subject to the GCA’s code of practice (GSCOP) were at that time compliant, forecasting was one of the top three issues in its most recent (2017-2018) report. Suppliers reported poor forecasts from retailers, significant variations between forecasts and orders, penalties for failing to meet service levels as well as incidents of being left with significant amounts of stock.
So what is going wrong? In an era of unprecedented data availability and a clear commitment to improving operational performance throughout the supply chain, why are these organisations failing to deliver the depth and accuracy of forecasting required not only to meet regulatory expectations but also address the key issues of wastage and availability?
The importance of accurate forecasting is undeniable: better forecasting reduces waste by avoiding unnecessary over production and avoids sales lost, by ensuring the right products are in the right store at the right time. For too long, therefore, has the industry hidden behind how difficult it is. This must – and can – change. With an increasing amount of data now available, combined with improved data analysis tools and, more recently, the use of Artificial Intelligence (AI) and machine learning to drive ever more accurate forecasting, retailers have the chance to drive significant improvements in efficiency and cost throughout the supply chain.
In large part, the difficulty around forecasting comes from a lack of consensus. There is no consistency of forecast model – even forecast definition – within the retail industry. Ask the top ten UK retailers for a definition of a forecast and the replies will range from an annual prediction of the volume of every product to be purchased from a supplier to a three month prediction of weekly product sales by store, a six week aggregated weekly order prediction, a 14 day daily prediction of short term product sales, allowing for weather and promotions or a seven day demand forecast. Each of which will be delivered with varying degrees of accuracy – and none of which come close to providing the key information required by a supplier, namely individual product forecast by order volume, by depot, by day.
This latter point is key: data can be made to provide unprecedented levels of insight, but you have to ask the right questions of it. If retailers want to drive better, more efficient, less wasteful supply chains they need to deliver the information that suppliers need, in a format that suppliers can use. It needs to be accurate, consistent, frequently updated, and provided in an accessible tool that suppliers can quickly embed within their decision making processes, especially if this is the information used to measure supplier performance.
Achieving 100% forecasting accuracy is unlikely, but there is huge potential to make better use of available data, trends, known variables and machine learning techniques to achieve improvements that far outperform current levels.
The GCA has recommended closer collaboration between retailers and suppliers – and that collaborative retailer-supplier relationship should be using all available data to maximise the accuracy of the forecast. But this insight could also provide a platform that enables better communication between suppliers and buying teams to share intelligence and learn from past issues.
The challenge is to encourage that collaborative model and to ensure the insight shared is based upon a consistent and agreed definition of the forecast. Without access to timely, robust and consistent data suppliers will struggle to interpret information, drive effective decision making and hold retailers to account.
The GCA is doing great work to reinforce the need for better forecasting but clearly, the recent investigations reveal retailers need help to implement the spirit of the code effectively by delivering better forecasts to suppliers. Indeed, while the entire supply chain will benefit from the provision of accurate daily product by depot order forecasts, retailers’ continued struggle to accurately and consistently forecast demand underlines the complexity of forecasting.
The good news is that growing numbers of retailers are making the necessary data available – and whilst retailers may not have the time and resource to look at all of their ~40,000 SKUs daily, suppliers can. By using best-practice visual analytics tools and techniques, supply-chain and sales data can be blended to highlight indicators for demand, such as sales spikes or low-depot stocks, and provide an easily understood picture of recent and historic sales patterns. The benefits are both significant and immediate – increasing the accuracy of forecasting will drive performance improvements at every stage of the grocery flow-of-goods.
The GCA has thrown down the gauntlet; the onus is now on the industry, on suppliers and retailers to collaborate and create a consistent forecasting model that can support far more efficient supply chains that eradicate wastage, reduce cost and optimise product availability.