Valimotie 27, 00380 HELSINKI
Authors: Mikko Kärkkäinen
Managing Director, D.Sc. (Tech.)
Director, Scandinavian Operations, D.Sc. (Tech.)
In retail, store ordering has an important impact on profitability. The accurateness and efficiency of store ordering affect sales through shelf availability, but also has an effect on handling, storage, and obsolescence costs in the stores and other parts of the supply chain. As even small outlets often have thousands of items and larger hypermarkets have tens of thousands of items, it is clear that accurate, item-level control is virtually impossible to achieve when store ordering is carried out manually. Many companies have solved the problem by taking into use computer-assisted ordering, with good results!
In retail, shelf availability is a prerequisite for sales. In grocery retail, approximately 3.9% of revenue is lost due to stock-outs1 and it is estimated that the situation is even worse in specialty retail.
Figure 1. In grocery retail, shelf availability is approximately 91.7 percent on average (Source: Corsten & Gruen, 2003).
The most common reasons for poor shelf availability are store ordering and failures in demand forecasting2. The difficulties are accentuated during promotions and seasons. It is estimated that over 15 percent of promotional sales are lost1.
Figure 2. Store ordering causes over half of the stock-outs (Source: Coca Cola Retailing Research Council, 1996).
In addition to stock-outs, problems in store ordering and demand forecasting also cause high obsolescence costs. Spoilage costs of fresh goods and mark-down costs of seasonal products or products with limited life-cycles have a significant negative impact on retail profitability.
Retail profitability is also affected by ordering and handling costs in the retail outlets. The cost impact is notable – studies maintain that more than half of retail supply chain costs originate in the retail outlets3. Furthermore, as store staff typically divides its time between customer service and the ordering, receiving, and shelving of goods, time spent on ordering and product handling is time away from sales and customer service.
Figure 3. Most of the logistics costs in retail supply chains originate in the retail outlets
(Source: Zelst etc., 2005).
It is important to keep in mind that the store staff is not to blame for poor shelf availability or high obsolescence costs. As even small outlets easily have over a thousand items, and larger outlets have tens of thousands of items, there is simply not enough time to visually monitor the availability of each item or to manually calculate the correct amounts to be ordered. Consequently, in order to improve efficiency and shelf availability, many retail companies have begun to automate their product replenishment.
Computer-assisted ordering is a great way of tackling many of the shelf availability and efficiency challenges in retail:
The basic idea of automated store ordering is simple. The replenishment system monitors each outlet’s sales and compares it to the calculated stock position and possibly to a demand forecast. The system uses this information to make an order or order suggestion that keeps stock at the desired level.
Figure 4. Basic idea of automated store ordering.
Typically, orders or order suggestions are generated either using rather simple re-order point logic or by use of more sophisticated replenishment models based on demand forecasts:
The forecast-based replenishment model has many advantages in comparison to the re-order point model. First of all, the forecast enables the system to automatically take into account anticipated changes in demand. Sophisticated forecast models are able to detect both trends and seasonal changes in demand. Second, forecast-based models offer better support for management of special situations. Instead of adjusting the re-order point up and down, changes in demand can be tackled through the forecast, which makes it possible for the store ordering system to take realized demand into account when placing orders and to automatically reduce the amounts ordered at the end of a promotion or season.
Efficient automation of store replenishment requires that some basic processes are in order: sales of items need to be accurately identified and recorded at the check-outs, backroom storage management must be in place, inventory records need to be sufficiently accurate, and a clear and consistent assortment management process needs to be in place.
In order for the replenishment system to be able to monitor item sales at the different outlets and to calculate demand forecasts based on this information, items need to be identified and their sales recorded accurately enough at the check-outs. In retail, this criterion is usually met. However, problems may occur in, for example, fashion or sports retail if items of different colors or sizes do not have individual product codes.
As automated store ordering leans heavily on inventory information, the accuracy of the inventory records has a great effect on replenishment accuracy. Goods receipt must be timely and accurate. It is also essential that sales are registered using the correct product codes; for example, different taste or color variants of an item should not be sold using the same product code. In addition, it is important that the accuracy of the inventory information is constantly monitored. Especially when there seems to be too many or too few items on the shelves, inventory counts should be conducted.
In retail outlets, it is rarely possible to keep separate inventory records for products located in the sales area and products located in the backroom. Therefore, the replenishment system has to assume that all received goods are on the shelves and available to consumers. This makes backroom storage management extremely important. The aim should be to minimize the amount of items and the time they spend in the backroom. The backroom should also be kept tidy and clear enough so that nothing is lost or forgotten in it.
Another requirement for automated store replenishment to work is that it needs to be completely clear to the replenishment system which items it should order automatically and which it should not. Thus, a systematic assortment management process, which determines which products belong to a specific outlet’s stocked assortment, needs to be in place. This may present a challenge if assortment information on the outlet level has not previously been maintained in an IT system. In specialty retail, for example, store personnel have in many cases been able to use their own discretion in deciding which products to order and which not to order from the entire range of products available in the retail chain’s IT system.
The automation of store replenishment comes with some challenges that may slow development or reduce the benefits attained. However, these challenges can be tackled if they are recognized in advance.
One challenge that companies are often faced with is the inadequacy of the chosen replenishment model. Most companies start out using the simple and easy-to-understand re-order point model. This model works well for products that have low or very stable demand. However, when automatic replenishment is expanded to products that have higher and more variable demand or shorter shelf-lives, the re-order point model often proves insufficient, with stock-outs or obsolescence problems as a result. Dealing with more challenging products typically requires implementation of a forecast-based replenishment model. Roll-out of automated store ordering should be planned with this in mind.
Even when sophisticated replenishment models are used, poorly set replenishment parameters can still lead to poor results. As the number of items to control is typically very high – in large retail chains easily hundreds of thousands when the number of items is multiplied with the number of outlets – setting and continuous updating of parameters presents an important challenge. When parameter management is done manually, the only option is to rely on simple rules of thumb and to use the same parameters values for large groups of outlets and products. As demand for different products within a product group, or the same product in different outlets, often varies a great deal, this approach cannot lead to very accurate replenishment results; even when the parameter values are good on average, there are typically many products and outlets where stock levels are too low or inventory turnover is too slow. It is, thus, important that the chosen replenishment system supports the use of product and outlet specific replenishment parameters and that the system continuously monitors replenishment results and reacts to deviations by automatically updating parameter values or suggesting changes to them.
Automated store replenishment often improves shelf availability, decreases stock in the outlets, and makes the store personnel’s work more efficient, simultaneously. However, there is typically a need to prioritize between the different goals: is it more important to achieve great shelf availability or minimal obsolescence costs, efficient deliveries or attractive shelf presentation, a level order flow or fast response to changes in demand? Poorly set, unclear, or conflicting goals make it difficult to efficiently develop automated store ordering. Defining replenishment goals is an important management task and typically a great change to how things were done before automated store ordering was introduced. When orders are placed manually by a large number of store staff, replenishment goals set by management give, at most, some guidance to the replenishment process and do not necessarily reflect on the actual ordering taking place in the individual outlets. When replenishment orders are system-controlled, it becomes necessary to clearly define what concrete results are desired, so that the replenishment parameters can be set accordingly. Attaining a perfect score on all performance indicators is not relevant, and hardly even possible, especially in the early stages of the transformation. From a business perspective, it is more important to identify the most critical success factors and to focus on them. Adequate minimum levels must be set and met for the other goals.
Nevertheless, the most crucial obstacle to automated store ordering is giving up! Often companies use various excuses in order to avoid making changes to their replenishment systems: ”Our demand is so heavily promotion-driven that replenishment systems cannot manage it” or ”We tried this in 1992 and it did not work then”. Special circumstances affecting demand, such as promotions or seasons, can usually be managed – a suitable forecasting and replenishment model just needs to be identified. This applies to almost all difficulties that companies face – a more challenging starting point usually requires more work in the initial stages, but the payback is also higher in the end. It is important to keep in mind that automation of store ordering is a long-term development project that involves and has a big impact on numerous employees and several business processes. It is, therefore, important to plan and reserve adequate resources for the change management needed.
Very concrete results have been achieved through automated store replenishment in a very short time. For example, the customers of RELEX have attained:
We at RELEX have a great deal of experience in improving the replenishment processes of retail companies. With our solutions, our customers have been able to improve their service levels and inventory turnovers as well as to make their replenishment processes more efficient. If you seek to improve your company’s profitability, do not hesitate to e-mail or call us: email@example.com or +44 7546 124031.