How do I get POS sample data

Use POS data in your supply chain - react faster to fluctuations in demand

I took my doctoral thesis from 2005 out of the drawer and realized with horror that it is still up to date. My research fields included various approaches to be able to react more quickly to fluctuations in demand by using sales data from shops, so-called POS data (point of sale), in other sections of the supply chain. The basic idea is to continuously follow the sales in the shops, so to speak to have your finger on the sales pulse in order to be able to recognize fluctuations in demand before they are reflected in the orders at the central warehouse and the suppliers.

This basic idea was correct, and the theory proved resilient. However, ten years ago few companies had the resources to take advantage of the vast amounts of information that POS data provides. Most companies have now recognized the benefits of POS data and see it as an interesting field for future developments - with an emphasis on "future". Most of them are not aware that this future has long been the present.

Thanks to developments in data management and computing capacity, the systematic, comprehensive use of POS data in the supply chain has long since become a reality. Especially for companies in the retail sector, who mostly run the stores and the corresponding warehouses themselves, there is no longer any excuse for leaving the POS data in the branch's POS system unused. It is time to move from visions of the future to using POS data in everyday life.

POS data shortens the response time

The value of POS data is measurable. In my dissertation, I examined the start of sales of several products at two well-known providers in the consumer goods sector. By analyzing POS data, it was possible to determine the demand for the new products on average 30 days earlier (with peak values ​​of up to 100 days) than was possible if the data was restricted to the orders from the stores at the central warehouse revealed. The difference in response times was even more pronounced: it was an average of 61 days (with peak values ​​of up to 128 days) if you compare the informational value of the POS data with the information that can be derived from the orders from the central warehouse to suppliers.

There are many reasons for this delayed flow of information. When new products are launched, the most important factor is the buffer inventory that is initially created in each section of the supply chain.

The buffer stock is created so that all sections can safely meet the demand. It is also common to have a few sales packages in every retail store and a few pallets in the central warehouse. Since weeks or even months are predominantly expected for the rotation of goods (even with so-called "high-speed movers"), it takes some time until the inventory has decreased to such an extent that a reorder is necessary.

If only the order flow at the central warehouse can be viewed, without taking into account the sales in the shops, this leads in the worst case to the following scenario: First there is a large increase in sales (the first deliveries to the shops), then the product seems to be a failure (none for weeks Reorders from the stores) and finally there is a bottleneck when reorders do come in. Figure 1 uses a specific example to show how much POS data and order flow can differ from one another.

POS data is key, but it takes a little more effort

The requirements for the transmission and processing of large amounts of data have changed fundamentally over the past ten years. The analysis of sales figures at the level of items, stores and days is no longer a particular difficulty. For many retailers it is part of everyday life. (It is now also possible to process the sales figures in more detail per receipt line at moderate costs. However, we will not consider this case here.)

It is clear that access to POS data accelerates the flow of information within the supply chain and enables inventory levels and production to be adjusted more quickly to actual demand. In this context, however, one important restriction must be made: Regardless of all other factors, it should be noted that the central warehouse and suppliers must react to the actual incoming order.

If the POS raw data is used directly for a forecast of the stocks in the central warehouse or at suppliers, there is a risk that the stocks are not in line with the actual delivery requirements. For example, in the case of new items that are delivered in relatively large units, a forecast based on the POS data can lead to central warehouses or suppliers building up large stocks very quickly, which are only needed many weeks or months later. For items with a short life cycle or limited shelf life, this can lead to significant problems due to obsolescence.

A forecast of future demand must be based on several factors: In addition to the analysis of POS data to estimate customer demand, the buffer stock built up in the supply chain should also be taken into account and an estimate of when excess stock will be reduced. In addition, delivery times, delivery schedules and multiple orders should be taken into account, which lead to bundling in the order flow. In addition, the short-term effects of marketing measures should be filtered from the demand data.

Sounds complicated? Don't worry: this is where the enormous computing power of modern IT systems comes into play.

More control through refined POS data for order forecasts

The refinement of the POS data for order forecasts reduces uncertainties in the supply chain and the need for interpretation of the data - while the reaction speed remains the same.

Order forecasts provide the best possible estimate of when and to what extent an order will be placed in a particular store or warehouse. In contrast to POS raw data, the order forecasts can therefore be used directly by the next link in the supply chain - for example from the central warehouse - without the risk of excess inventory that is not in line with the actual delivery requirement (see Figure 2).

A good order forecast takes into account three sources of information:

  1. the expected customer demand, calculated based on the current POS data
  2. Existing stocks as well as reorders that have already been made
  3. Control parameters (reserve or display stock, package size or minimum order quantity plus delivery schedule or delivery time) that affect the point in time and the order quantities

An order forecast may seem simple, but it is actually a complex process that evaluates all available information.

In practice, the creation of a reliable order forecast requires an intelligent inventory control system. An intelligent inventory control system not only calculates the current order requirement, but also simulates future requirements. On the basis of the current forecast based on POS data and the control parameters, the system calculates future fluctuations in the inventory and indicates when new orders will be placed and goods will be received. (Figure 3 shows an example of simulated orders.) In principle, the simulation can extend as far into the future as you want, but the value of data on current stocks and control parameters is of course greater in the short and medium term.

The simulation of the order flow places high demands on software and computing power. With current IT solutions, this challenge can now be mastered easily. Such calculations are carried out daily for our customers for the next few weeks or months. In certain cases, the calculations are carried out with several intermediate steps so that the order forecast of the shops or on-site intermediate storage can flow into the order forecast of the central warehouse.

When are the benefits of order forecasting greatest?

When the order flow is steady and there are no changes, the difference between a complex order forecast and a simple forecast based on issues at the central warehouse is generally relatively small. But how large is the proportion of the range where there is a steady flow of orders? Shorter product life cycles, more flexible opening times and the increasing importance of seasonal ranges mean that the proportion of products with a steady flow of goods is continuously decreasing.

The order forecast is an important tool for controlling the supply chain when introducing and discontinuing products, for seasonal items, for marketing measures and for changes to the delivery schedule (see Table 1). When introducing articles, for example, a good order forecast enables the central warehouse's forecast to be automatically adjusted upwards or downwards if demand is higher or lower than expected - before the stores have received their first reorder.

Order forecasts have more than one area of ​​application

In addition to a faster response to fluctuations in demand and a balanced total inventory in the supply chain, order forecasts have other advantages. Our customers have found that their suppliers can optimally adjust their inventory levels if order forecasts are passed on. You will then benefit from better customer service and receive fresher goods.

Simulations of the future order flow can also be used for internal capacity planning. If it is estimated more precisely how many units are to be accepted or picked on certain days or how the inventory will fluctuate in the future, personnel requirements and transport capacities can be determined more precisely and situations can be dealt with in which the central capacity limits may be exceeded. Our customers in the consumer goods sector have observed, among other things, considerable improvements in the coordination of the order flow with public holidays. In these situations, demand usually rises while the number of delivery days is reduced, so that goods are quickly no longer available due to bottlenecks in order picking and transport capacities.

Summary

The ability to use POS data to control the supply chain has become a reality:

  1. Due to developments in IT, POS data can be collected, transmitted and analyzed by any retail company. Thanks to POS data, it is possible to react much more quickly to fluctuations in demand.
  2. Leading retail companies have already taken steps to create order forecasts that take into account all point-of-sale data, current stock levels and all relevant control parameters that affect the order flow. Order forecasts can greatly improve the accuracy of forecasts for central warehouses and suppliers. This applies in particular to the introduction and discontinuation of articles, for seasonal articles, before public holidays and for marketing measures. Order forecasts place high demands on system support and computing power. However, companies that use modern merchandise management systems benefit from automatic order forecasts and ongoing order simulations.
  3. The development opportunities are not limited to that part of the supply chain that the company itself controls. Some retailers have already gone a step further and have started relaying order forecasts to outside suppliers. In this way, they benefit from better service and fresher goods.

Are you still waiting for a future that has long been the present?