So, according to the University of Bamberg's return ticket Around 280 million parcels and 487 million items returned in Germany in 2018. This corresponds to approximately every sixth package delivered and every eighth item ordered. Despite all efforts, returns are simply part of online retail. This applies in particular to the fashion industry. The increasing number of items and packages being returned poses major challenges for online retailers and, in particular, the returns process. From goods receipt, inspection and classification according to recyclability, to reprocessing and storage of resalable items in new inventory or the disposal of non-resalable products, to customer communication and reimbursement of the purchase price: Returns processing is a complex and cost-intensive process. It must therefore be constantly improved and optimized. But how? In order to speed up returns handling, capacity planning is required that is based on forecasting the expected returns. The better the forecasts, the more efficiently returns can be processed.
Does your warehouse often overflow due to the large number of returns? Do you sometimes have too many and sometimes too few staff to process the number of packages? With this article, we'll show you how you can avoid exactly these scenarios by using machine learning and predicting return quantities.
THE EXTENT OF THE RETURN PROBLEM
Online retail is currently jumping from high in turnover to high in turnover and further growth is also expected in the future. However, this also means an increasing number of returns. Returns present online retailers with major financial, logistical and organizational challenges. The ultimate goal of online retailers is therefore to avoid returns through preventive measures (preventive returns management). The options here range from classic measures such as detailed product information or the display of genuine customer reviews to more innovative approaches such as online consulting services in the form of virtual dressing tools or the use of machine learning algorithms for personalized product recommendations or the determination of customer-specific return probabilities.
Regardless of the efforts, there are still returns. This is the average return rate, one EHI study According to this, at around 20%. Depending on the product category, however, there are huge differences here: While less than 10% are returned in the food and beverage sector, for example, the fashion and accessories category is the clear leader with almost 40%. Returns are therefore simply an unavoidable part of online trade.
For online retailers, high return rates mean high costs. For example, the average process costs of a return shipment are €7.93. The drivers here are transport costs for return shipping as well as personnel and material costs for the return process. There are also other return-related costs, for example due to a loss in the value of the goods or even for call center personnel. The costs vary greatly depending on, among other things, the number of returns, the industry and the item, as shown in the table below.
As a result of the increasing number of items and packages being returned, returns represent an enormous cost factor that needs to be controlled. The aim is to process returns quickly and efficiently in order, among other things, to reduce the costs per return and to keep the turnaround time from receipt of goods to ready-to-sell again as short as possible. Effective curative returns management starts with planning the return quantity. This requires a forecast of the expected number of returns and packages. The availability of more and more machine learning data and technologies opens up completely new opportunities to create return forecasts that are as accurate as possible.
THE RETURN FORECAST AND ITS BENEFITS
Forecasting returns is about predicting the quantity and timing of returns. Planning logistical processes and resource distribution requires, in particular, predicting the expected returns at package level. Because personnel are needed to enable efficient returns handling. However, how many employees are required per day or per shift depends heavily on the number of packages to be processed. The returns forecast thus provides answers to questions such as: What quantity of returns and therefore parcels can I expect tomorrow, in the coming week and in the next few weeks? Under what circumstances is an order returned, what are the most important return drivers? And how many items are in a package on average?
The better the future return volume can be predicted, the easier it is to plan personnel deployment and adapt it to fluctuating capacities. This is important for companies to be able to process returns efficiently and quickly. As a result, items can go back online more quickly and are available for new customers to buy. Ultimately, not only can sales be increased in this way, but the customer also benefits from a quick refund. In the end, this also contributes to increased customer satisfaction and loyalty. On the other hand, forecasting the number of returns naturally also enables targeted management of either internal logistics resources or corresponding logistics service providers.
The returns forecast therefore has great potential: It enables you to get a clear overview of the associated costs and, above all, the required capacities for logistics, warehouses and personnel. In this way, the returns forecast not only ensures that your warehouse does not sink into chaos, but also supports your financial planning. But how does the whole thing work? And how does machine learning come into play here and what data does it require?
DATA, MACHINE LEARNING AND RETURN FORECASTING
The aim of returns forecasting is to train machine learning algorithms to recognize patterns based on historical returns that allow conclusions to be drawn about the quantity and time of future returns.
Machine learning (German machine learning, ML) is a sub-discipline of artificial intelligence (AI) and a generic term for artificially generating knowledge from experience. It includes training ML algorithms that automatically recognize patterns and relationships in historical data. These identified patterns can then be applied to new data to make predictions.
In order for the returns forecast to be successful, there must first be a sufficiently large data history of returns. However, due to the naturally high level of digitization, online retail in particular has much more data that can be used to create a return forecast that is as precise as possible. These include the following:
- number of orders
- Information about the product being sold (e.g. size, color, manufacturer, price)
- Customer information (e.g. gender, age, place of residence, return behavior)
- Shopping cart information (e.g. number of items in different sizes, order total)
Other data sources that can also be used as harbingers of returns include information on marketing campaigns that have already been carried out and planned (e.g. discount campaigns and campaigns) or a change in product range or collection (e.g. winter to summer collection).
In addition to these internal data sources, it can also help to include external data, such as the weather, holidays and upcoming major social events (e.g. soccer World Cup), as indicators in preparing the returns forecast.
Based on this data, it is then possible to predict how many returned parcels and items are to be expected and when. The insights gained about return drivers not only help to create precise return forecasts, but are also very valuable in terms of preventive returns management.
With the help of returns forecasting, online retailers have a valuable tool to quickly and efficiently manage the necessary evil of returns.
CONCLUSION
Due to the increasing share of online retail and even though online retailers are currently investing a lot in preventive returns management, it can be assumed that the number of returns will continue to increase. Efficient returns processing is therefore essential. On the one hand, to be able to offer returned items back to the customer for sale as quickly as possible (assuming resalability) and, on the other hand, to meet customer demands for a quick refund of the return or a product exchange. Intelligent forecasts are needed to meet this challenge. This is the only way to make optimal use of warehouse capacities and is the only way to better plan order quantities and adapt personnel requirements to a fluctuating number of returns. A returns forecast is therefore a valuable tool for all retail companies that process their returns themselves, as well as for return service providers or logistics companies.
Would you like to know what the return forecast actually looks like for a logistics service provider in the online fashion sector? 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.