According to estimates, 30-40% of clothing purchased online is returned currently, resulting in an increase in substantial costs for retailers.
There are several reasons why online channels are experiencing high return rates today. This includes wrong orders, incorrect size and fit, colour differences, quality issues, or simply because of change in preference.
To add to that, Covid-19 accelerated the trend of buying the same article in multiple sizes to test which one fits best and returning the rest. The above trend, which stayed post the pandemic, contributes to high return rates.
How does the complexity of handling returns impact retailers' bottom line and operational efficiency?
Given the returns add significant costs to retailers’ balance sheets, it becomes a key business concern for them to address and solve. The process of refunds introduces a notable disparity between gross sales and net sales, leaving businesses unsure about their actual sales figures.
The time required to calculate net sales increases as return policies become more flexible. For instance, HUGO BOSS offers a return window from 30 to 60 days, Tommy Hilfiger allows up to 60 days, and Zalando extends it to 100 days. Consequently, this extended processing time hinders a company’s ability to respond to market trends promptly.
Apart from the refunds, companies also incur returns-related operational costs involved in packing, shipping, and refurbishing. Moreover, by the time the returned items are ready for resale, the articles may already be out of season.
The above factors make returns a burden to corporate profits. A study by Narvar, a return management company, found that cutting returns in half can increase profits by 25%.
Optimising return management strategies through advanced analytics
Fashion companies today rely on advanced analytics to address their return management concerns. This provides accurate sales forecasts by estimating the probability of return for every article sold, which will then determine the probability that an article is retained.
Besides that, advanced analytics methodologies can now help fashion brands generate valuable insights through reliable estimates of future product returns that enhance their operations, strategies, and bottom line. But that’s not all:
Real-time monitoring of net sales targets
Reliable return estimates provide businesses with real-time information on net sales performance. This enables them to track their progress on sales targets and make data-driven adjustments as needed. By assessing net sales on a daily basis, companies can take immediate action to address any shortfalls or capitalise on newer opportunities, leading to improved financial outcomes in their balance sheets.
Agile response to market trends
With the ability to estimate future product returns, businesses can make informed decisions and adapt their strategies more quickly. By understanding the potential impact of returns, they can respond promptly to changing market trends, adjust inventory levels, and optimise supply chain processes. This agility allows companies to stay ahead of the competition and remain relevant in a rapidly evolving industry.
Reliable return estimates help to create accurate budgets for research and development, retail operations, and e-commerce infrastructure. By considering the probability of returns in their calculations, companies can better allocate resources. This knowledge empowers marketing and design teams to make data-driven decisions, resulting in more effective campaigns and collections tailored to the preferences and behaviours of their target customers.
Estimating conversions from online advertising campaigns
Taking the probability of return into account will help assess the value of conversions generated from online advertising campaigns. By incorporating return estimates into their analysis, businesses can gain a comprehensive understanding of campaign performance and accurately evaluate the ROI of their advertising efforts. This information enables them to refine their strategies and optimise advertising spending for maximum impact.
Operational and logistics planning
Estimating the number of returned products expected at the warehouse is invaluable and allows businesses to allocate resources efficiently, streamline processes, and reduce costs associated with returns. By optimising warehouse operations, businesses can enhance customer satisfaction, improve turnaround times, and minimise the impact of returns on their overall operations.
Net Merchandising Value: an advanced analytics methodology for accurate return estimates
Net Merchandising Value is a fully automated solution that is designed to estimate return probabilities for every item sold in the future, both for businesses’ own websites as well as for marketplaces.
With highly accurate predictions, this estimation allows businesses to stay ahead of the competition, respond to market trends more swiftly, monitor net sales targets in real-time, allocate budgets with more accuracy, and optimise the value of online advertising campaigns.
By predicting the number of returned products expected to be received at warehouses daily, Net Merchandising Value also enables improved operational and logistics planning, ensuring that stock flows are optimised at significantly higher levels.
Empowering Net Merchandising Value with advanced analytics
Many companies still rely on traditional methods like time series models or historical averages to assess their return forecasting. However, these methods don’t just yield poor results, but they also sometimes fail to provide insights into products that are likely to be returned.
The advantage of adopting sophisticated machine-learning techniques tailored to the fashion industry is that businesses can overcome these limitations and achieve more accurate and actionable return forecasts.
Net Merchandising Value is powered by advanced analytics to generate reliable estimates of future product returns by closely considering various attributes such as colour, size, product category, and more.
This advanced analytics methodology takes into account order features such as the number of articles purchased together, the location to which the order is shipped, and the date of purchase. These parameters increase the accuracy of Net Merchandising Value’s return estimates.
Anonymised customer data, such as the number of historical purchases and returns made by a customer, is also leveraged, ensuring privacy while enriching the predictive capabilities of the model.
Moreover, the advanced analytics model accounts for the days since an order was placed, recognising that the probability of an article being returned decreases over time. This comprehensive approach considers the return status of every historically ordered article at every point in time.
The large data sets generated as a result of these comprehensive records are then handled effectively with Metyis’ extensive expertise and experience in automation, advanced analytics and artificial intelligence.
Net Merchandising Value’s accurate estimates come with rigorous methodologies
Net Merchandising Value adopts a rigorous methodology for model development and evaluation. A variety of models are tested and compared extensively to identify the best-performing model for return estimation.
Rigorous tests are conducted to assess the model's performance and robustness during times of external shocks or unexpected occurrences. Additionally, the model is retrained regularly to exploit the most recent trends in the data.
This thorough approach leads to Net Merchandising Value estimates that are accurate, reliable, and actionable, enabling businesses to make informed decisions based on intelligent insights.
Making an impact with Net Merchandising Value
At Metyis, we understand that accurate and actionable insights help optimise processes, minimise overheads, increase profits, improve the customer experience and grow businesses to the next level.
That’s why our experts, solutions, services, and methodologies come together to push the boundaries of innovation and provide transformative business impact for our customers across from multiple industries.
We are happy to give you a walkthrough on Net Merchandising Value or tell you about how our other technology, marketing, advisory, and digital commerce solutions and services can help make a business impact for you.
About the authors behind the article
Sanne Krom is a Partner in Metyis’s Amsterdam office; Arshiya Nagi is a Director in the Amsterdam office; Arpit Gupta is a Principal in the Madrid office, and JeanLuc Oudshoorn is an Analyst in the Amsterdam office.