Retail in particular, both online and stationary retail, has great potential and offers diverse areas of applicationto generate competitive advantages through data-based decisions. E-commerce in particular is predestined to implement innovations in this area due to the naturally high level of digitization and the large amount of available data. This is not only denied to large retail companies and e-commerce platforms, but smaller retailers in particular can also use their data treasures to create added value with the help of AI.

One prominent use case is forecasting: Whether it's sales planning, incoming goods or returns forecasting, a good forecast provides the basis for every planning. Unfortunately, previous planning tools, such as the frequently used Excel spreadsheet, the well-known gut feeling or even conventional planning software, are quickly reaching their limits due to the increasing complexity and volume of available data. Here, the use of machine learning algorithms makes it possible to consider many different data sources and significant quantities, so that more consistent, accurate and transparent forecasts can be made and thus ultimately increased planning security. Using five examples from the world of online and stationary retail, we will show you what this means in practice for typical forecasting applications in retail.

‍

WHAT IS FORECASTING?

Forecasting is about predicting future events as precisely as possible on the basis of historical observations and data as well as knowledge of future events that could influence the forecast.

Machine learning algorithms are perfect for a wide range of forecasting applications: They automatically learn patterns and relationships in historical data, which can then be applied to new data to make predictions. Assuming the appropriate data, a wide variety of things can be predicted, as the following examples show.

EXAMPLE #1: THE FORECASTING OF SALES IN THE FOOD TRADE

Which goods are needed in the individual retail outlets on a specific day? A large supermarket can have up to 30,000 items in its range, which means that the supermarket needs a forecast that is as meaningful as possible for all 30,000 items. A particular difficulty is the different sales frequency of individual articles. This is compounded by the fact that many assortment items have varying and, in some cases, extremely short sales histories, such as newly introduced items.

Especially in the food trade, machine learning-based Sales forecasts (demand forecasting) great potential, not only for retailers' purchasing and inventory planning, but also in terms of sustainability. By predicting the buying behavior of end consumers more precisely per day and per branch, the purchasing manager knows how many items must be placed on the shelf to avoid empty shelves, full warehouses or throwaway campaigns of perishable items. So could a current study According to Bitkom and BVE, intelligent sales forecasts will reduce food waste to zero by 2030 (“zero waste”).

On our blog, you can read the joint Read the success story with a supermarket chainHow we managed to reliably predict the sales of racers and bums.

EXAMPLE #2: THE FORECAST OF INVENTORIES IN TEXTILE RETAIL

The main goal of inventory optimization is to keep inventory levels as low as possible while still being able to offer customers the right product at the right time in the right place. With constantly changing customer demand, short product life cycles and the increasing spread of omni-channel models, this is no easy task. This is how a cause IHL Group study According to trade shortages alone, global sales losses of $634 billion each year, while excess inventory due to copies leads to sales losses of $472 billion. As a core part of a fashion store's purchasing department, a forecast can ensure that precisely these scenarios are avoided.

In this way, historical sales figures together with the price development of individual products can be analyzed using machine learning algorithms in order to forecast future product demand. Taking current inventories into account, the best possible order quantity can then be directly recommended to ensure that neither too many nor too few goods are in stock to meet the predicted demand. The automated creation of the individual order quantity forecast significantly makes work easier in purchasing planning. As a result of the new time, buyers at the fashion house can once again invest more effort in evaluating the goods in order to guarantee the best possible quality.

EXAMPLE #3: THE FORECASTING OF INCOMING GOODS QUANTITIES IN RETAIL LOGISTICS

A logistics service provider for a large German fashion house wants to know how incoming goods behave in warehouses in order to enable more precise personnel planning. In the fashion industry in particular, delivery dates and incoming goods are often difficult to plan. Instead of exact delivery dates, approximate time periods prevail. If delivery dates are not met and suppliers deliver their goods seemingly indiscriminately, it is almost impossible to optimally plan incoming goods and therefore personnel and warehouse use. There are regularly too many or too few warehouse employees, which leads to high costs and loss of productivity.

Machine learning can also help here and precisely predict incoming goods from various suppliers. Through a combination of supplier evaluation and machine learning, a concrete forecast can be made for the expected delivery time of the respective goods from the suppliers. This not only allows personnel to be planned more efficiently but also to make maximum use of the warehouse.

EXAMPLE #4: THE FORECAST OF RETAIL VISITORS

Fluctuating visitor numbers in retail make reliable personnel deployment planning difficult. Especially for businesses that are characterized by products with a high level of advice and support, it is important to know how many customers can be expected in a day and over the course of the day. In order to ensure optimal customer service, an optician therefore wants to know how many customers can be expected over the course of the hour, day and week, in order to plan the required employees in the individual branches accordingly. Thanks to frequency counters at the entrance to branches, information about the flow of customers is already available. The daily volume of customers consists of walk-in customers and customers with agreed appointments.

Based on historical visitor numbers and with the help of machine learning algorithms, a reliable forecast of the customer volume of each individual optician branch can be made and a recommendation of personnel requirements can be made automatically based on the calculated forecast. When training the machine learning models, a variety of other influencing factors, such as the weather and information on special opening hours or shortened working weeks and vacations, can be included and checked for their impact on the forecast. With the help of forecasts, personnel planning and customer care can be optimized, costs reduced and at the same time generate direct added value for customers.

EXAMPLE #5: THE FORECAST OF RETURN QUANTITIES IN FASHION E-COMMERCE

Despite all efforts, returns are simply part of online retail. This applies in particular to the fashion industry. However, processing returns is a complex and cost-intensive process. In order to keep returns processing costs in check and to speed up returns handling, a logistics service provider responsible for a fashion retailer's online shop therefore wants to know how many returned packages can be expected today, tomorrow and in the next few weeks.

Machine learning also helps here by predicting the number of returns. This makes it possible to predict how many returned parcels can be expected and when. In this way, the daily volume of returns can be predicted and therefore personnel deployment can be better planned and logistics resources can be managed in a targeted manner. The return quantity forecast thus simplifies capacity planning for logistics, warehouses and personnel in the face of fluctuating package quantities.

Would you like to know more about how the returns forecast works? Then read our joint Success Story with Mode Logistik GmbH & Co. KG, the logistics operator of Fashion ID, the online shop of Peek & Cloppenburg KG Düsseldorf.

CONCLUSION

Regardless of whether it is B2B or B2C, whether stationary or online and regardless of which product category it is: The possible uses of forecasting are more than diverse and the examples mentioned above naturally do not cover all fields of application. The examples should only be a source of inspiration and help you to get an idea of what you too can predict in your company with machine learning.

Is your interest piqued? Then our forecasting experts will be happy to help you find out what your individual forecast might look like.

‍