Online Fashion Returns: Why They Persist and What Can Be Done

Commissioned by: IDNTFY me AS
Authored by Eduardo Escobedo

Introduction

Between 2014 and 2024, the global apparel market expanded at an average annual rate of approximately 3%, growing from around USD 1.3 trillion to an estimated USD 1.8 trillion (FashionUnited, n.d.; LeRolland, M., 2024; UniformMarket, 2025). This growth was mainly driven by e-commerce. Over the same period, global online apparel sales nearly doubled from roughly USD 350 billion to close to USD 700 billion, corresponding to an average annual growth rate of about 11%, significantly outpacing the overall market. As a result, the online share of total apparel sales increased from around 15% to around 30% between 2014 and 2024 (Statista, 2025).

This shift reflects not only technological progress but also a change in consumer behaviour. Online purchasing has reduced friction, expanded choice, and increased convenience (McKinsey & Company, 2023). At the same time, it has normalised purchasing patterns that contribute directly to high return volumes. Average online apparel return rates are estimated between 20% and 40%, compared to approximately 8 to 10% in physical retail (National Retail Federation, 2023; Statista, 2025). In some segments, return rates exceed 50%, driven in part by practices such as ordering multiple sizes or styles with the intention of returning a portion (Coresight Research, 2023).

Importantly, this challenge is not new, nor is it purely technological. Over the past decade, a range of solutions—including virtual fitting tools, size algorithms, augmented reality, and enhanced product visualisation—have been introduced. However, their impact on aggregate return rates has remained limited. Despite growing adoption, return levels have remained broadly stable over time (McKinsey & Company, 2023).

Throughout the development phase of IDNTFY – a tailored online ecosystem for fashion that seeks to significantly reduce waste – many information sources and insights from fashion experts pointed to an already large and growing number of solutions addressing the online returns issue. However, after further research and confirming that many years of global technological investment have not significantly reduced return rates, it became evident that there is more to this than is currently being discussed and that the problem is not purely technological (McKinsey & Company, 2023; Coresight Research, 2023).

Based on expert interviews, literature review, and surveys of participants in its online community, IDNTFY developed a hypothesis that high return rates persist because they are supported by both consumer behaviour and the commercial logic of e-commerce (Janakiraman et al., 2016; Hjort & Lantz, 2016). On the one hand, return behaviour can be understood as a rational response to uncertainty around fit, quality, and expectation. On the other hand, elements of the current e-commerce model, including frictionless purchasing, flexible return policies, and growth strategies focused on conversion and volume – may implicitly sustain or even reinforce these behaviours. As a result, incentives across the value chain are not always aligned with the objective of reducing returns.

Against this backdrop, IDNTFY detected a potential gap in research and thought leadership in relation to addressing the online returns issue, and decided to commission this White Paper to examine the central question: what integrated set of next-generation technologies, behavioural mechanisms, operational tools, and incentive structures is required to meaningfully reduce return rates by both improving purchase decisions and reshaping the behaviours that drive them?

State of the Art: Consumer Return Data, Interpretations, and Blind Spots

The current body of literature on online apparel returns seems to converge around a relatively stable set of empirical observations. Return rates are consistently estimated between 20% and 40% for online fashion, significantly above physical retail benchmarks (NRF, 2023; Statista, 2024; McKinsey & Company, 2021). Within this, a widely cited claim suggests that approximately 70% of returns are driven by size, fit, or related product expectations (McKinsey & Company, 2021; Patel et al., 2025).

However, a closer look at these figures reveals a lack of consistent empirical validation and relies heavily on repeated industry benchmarks. Many widely used statistics originate from industry reports, consultancy analyses, or proprietary datasets rather than harmonised, peer-reviewed studies. While these sources provide useful insights, they often rely on specific samples or methodologies that are not fully transparent or directly comparable.

In addition, a number of recent studies reproduce similar figures – particularly the “~70%” estimate – by referring to earlier reports rather than independently validating the data (Patel et al., 2025; Marriott, 2025). Over time, this repetition has helped establish certain figures as accepted benchmarks, even when their empirical foundations remain uneven.

Another limitation concerns how return reasons are captured. In most cases, consumers select from simplified categories at the point of return, which introduces potential response bias. Faced with predefined options, consumers may choose the most straightforward or socially acceptable explanation, such as “size” or “fit”, rather than providing a fuller account of their behaviour. More complex motivations, including intentional over-ordering or exploratory purchasing, are likely underrepresented, either because they are not explicitly captured or because they are less readily disclosed (Tourangeau, Rips, & Rasinski, 2000; Janakiraman et al., 2016).

More broadly, the literature tends to converge on a relatively narrow explanatory framework. By repeatedly highlighting similar drivers, existing research reinforces a product-centric view of returns, where the issue is primarily framed in terms of sizing accuracy or product representation. While these factors are clearly relevant, this framing systematically underestimates behavioural and systemic drivers of returns (El Kihal et al., 2025; Stöcker et al., 2021).

This limitation becomes more visible when contrasted with observations from applied market contexts. For example, in various market insights activities conducted with a community of approximately 1,000 women developed by IDNTFY over recent years, size-related factors tended to emerge only after broader issues linked to style, wardrobe management, and better buying practices had been addressed. These exchanges frequently surfaced a recurring tension in consumer experience: the paradox of owning a full wardrobe while still perceiving they have nothing to wear. Over a one- to two-year period, as this community was supported in refining these broader behavioural and stylistic dimensions, an average online return rate of 5% was observed, despite most purchases benefiting from free or highly flexible return policies.

Considering this, these observations suggest that high return rates cannot be explained by size and fit alone. Instead, they appear to result from the interaction between consumer behaviour, technological mediation, and business practices. The limited attention given to these interactions, particularly the possibility that current commercial models may accommodate or even benefit from high return levels, points to an important gap in the literature.

Addressing this gap requires moving beyond the repetition of established statistics toward a broader analytical perspective. This involves reassessing the reliability of commonly cited data, incorporating behavioural and systemic dimensions, and questioning the assumptions that underpin current interpretations.

From the perspective of return data interpretation, these findings suggest that observed return rates cannot be understood independently of the policy and incentive environment in which they occur. High return rates may reflect not only product-related issues (such as size and fit), but also the behavioural responses induced by retailer-designed incentives. Any attempt to reduce returns must account for the underlying trade-offs embedded in current business models.

The Business Model: The Systemic Trade-Offs Impacting Corporate Return Strategies

Returns are often presented as a cost problem. They are. But they are also part of how the current model works. In this context, reducing returns is not only an operational challenge, but it also directly challenges revenue-driving mechanisms.

From an operational perspective, the impact is clear. Returns generate direct costs through shipping, handling, quality control, repackaging, and potential loss of value. They also create indirect effects—pressure on customer service, inventory planning, and overall operational complexity (Asdecker, 2015). At scale, the implications are significant. Returns are estimated to generate over 25 billion kilograms of waste annually, underlining their economic and environmental footprint (BBC, 2023; The Interline, 2023).

More recently, companies have also faced a rise in opportunistic or fraudulent behaviours, such as wardrobing (purchasing items, using them, and returning them) or counterfeit product substitution. While difficult to quantify, these practices add risks and cost to an already complex system (NRF, 2023; Appriss Retail, 2023).

These elements suggest a strong incentive to reduce returns. In practice, the picture is more nuanced.

Lenient return policies such as free returns, extended windows, and minimal checks, play a central role in driving sales. They reduce perceived risk at the point of purchase and make online shopping more attractive. Evidence shows that this effect is material: higher flexibility leads to higher conversion and larger basket sizes (Janakiraman et al., 2016). In many cases, the additional revenue generated outweighs the cost of processing returns.

This creates a structural trade-off. Returns are costly, but they also support growth. As a result, they are not simply tolerated; they are structurally embedded in the current e-commerce model (Hjort & Lantz, 2016; Janakiraman et al., 2016).

Marketing dynamics reinforce this logic. Social media and influencer platforms increase both the frequency and the volume of purchases, while also promoting aspirational and imitative consumption behaviours (Djafarova & Rushworth, 2017; Lou & Yuan, 2019; Statista, 2023). More recent developments suggest that these platforms are also beginning to influence post-purchase behaviour. Content formats such as haul videos, “GRWM”, and unboxing increasingly incorporate “keep or return” interactions, effectively embedding returns within the consumption experience itself (Vogue Business, 2023; Coresight Research, 2023). Practices such as over-ordering, bracketing, and purchasing for content creation are becoming more visible and normalised, with 69% of Gen Z consumers reporting intentional over-ordering with the expectation of returning items (Coresight Research, 2023).

This broader industry pattern also aligns with insights gathered by IDNTFY, suggesting that underlying financial incentives may reinforce these dynamics. Insights from industry experts indicate that in financing and investment contexts, equity markets tend to prioritise revenue growth over profitability metrics (Financial Times, 2024). Where sustained revenue expansion is perceived as the primary signal of performance, some players in the online fashion industry may consequently treat the cost implications of returns as a secondary concern, rather than as a core driver of operational efficiency.

In this context, returns cannot be viewed as isolated inefficiencies. They are the outcome of a system designed for low commitment and high flexibility. Efforts to reduce them therefore interact directly with the commercial drivers of the business.

These trade-offs are critical when assessing potential solutions. Any approach that aims to reduce returns must account for the fact that current incentives do not always support that objective. Without addressing this imbalance, even well-designed tools are likely to have limited impact.

Consumer Behaviour Dynamics: From Decision-Making to Systemic Patterns

Returns persist because consumers are behaving rationally within the system they are given (Janakiraman et al., 2016; Oghazi et al., 2018).

Return behaviour in online apparel is often presented as a series of individual decisions driven by product mismatch. In practice, it reflects something broader: a set of learned behaviours shaped by the retail environment itself. The persistence of high return rates across markets suggests that returns are a normal and expected outcome of online apparel consumption (NRF, 2023; Coresight Research, 2023).

As previously stated, return rates are on average high, ranging between 20% and 40% globally, with significant variation across geographies and demographics. In Europe, return rates reach particularly elevated levels in countries such as Switzerland (45%), Germany (44%), and Austria (36%) (Statista, 2025). Generational differences are also pronounced: Gen Z consumers return approximately half of purchased items, compared to around 40% for millennials and lower rates for older cohorts (Statista, 2025). These patterns are not incidental. They point to behavioural models that are reinforced over time, rather than corrected through experience.

A central factor of this dynamic is how consumers perceive and integrate choice and flexibility into their buying decisions. Free delivery and free returns are no longer peripheral services; they are core expectations (Janakiraman et al., 2016; Luo et al., 2025). More than 90% of European consumers consider them important, and a majority review return policies before completing a purchase (Statista, 2025; Bennet, N., 2025). In this context, purchasing and returning are not distinct phases, but part of a single, integrated decision-making process, where the option to return shapes the initial act of buying.

This shift has important implications. It blurs the boundary between purchase and post-purchase behaviour. Ordering multiple sizes, testing products at home, and returning unwanted items are not deviations from intended use. They are embedded practices, enabled by system design and reinforced by convenience (Janakiraman et al., 2016; Rokonuzzaman et al., 2021). Over time, these practices become routine.

At the same time, the perceived cost of returning remains low. Available data suggests that only a portion of consumers associate returns with environmental consequences, and fewer recognise that returned goods may not be resold (European Environment Agency, 2024; BBC, 2023; Statista, 2025). Observations from educational activities conducted within IDNTFY’s community of approximately 1,000 women point in the same direction: a large proportion of participants did not connect apparel returns to waste generation and were unaware that a significant share of new returned items is discarded rather than reintegrated into inventory.

This weakens any behavioural constraint linked to sustainability. Even in markets and demographics typically associated with higher environmental awareness, this awareness does not consistently translate into purchasing behaviour, as can be seen from the geographical and generational data presented above.

Social and cultural factors further reinforce these dynamics. Digital platforms have reshaped not only how consumers discover products, but how they engage with them. Content formats such as haul videos, “get ready with me,” and unboxing have made trial-based consumption visible and socially acceptable. Purchasing multiple items with the intention of returning some is no longer hidden behaviour; it is often presented as part of the experience. In this environment, returns are not simply tolerated – they are, in some cases, implicitly encouraged.

Beyond observable behaviours, there is also a less visible psychological dimension. While “size and fit” dominate reported return reasons, these categories often mask more complex evaluations. Perception of fit is influenced by body image, expectations, and context. An item that is technically correct may still be rejected if it does not align with how the consumer perceives themselves. These factors are difficult to capture through standard return interfaces, yet they play a material role in decision-making (Stöcker et al., 2021).

Importantly, these behaviours do not appear to diminish with experience. Evidence suggests that return patterns stabilise over time rather than improve (Stöcker et al., 2021). Consumers learn how to navigate the system efficiently, incorporating returns into their routine. In this sense, returns are not a temporary inefficiency. They are a learned response to a system that accommodates and in some cases rewards experimentation.

All these elements shift the interpretation of return behaviour. Returns are not driven solely by imperfect information or product mismatch. They are the result of an interaction between incentives, habits, social norms, and psychological factors, all operating within a system designed for flexibility and convenience.

This has direct implications for the effectiveness of current solutions. Tools that aim to reduce uncertainty at the point of purchase address only part of the problem. When behaviour is shaped by intentional trial, low perceived cost, and system familiarity, improvements in accuracy alone are unlikely to produce structural change. Addressing return rates therefore requires engaging not only with what consumers know, but with how they behave, and why the system allows them to behave that way.

State of the Art: Existing Technological Tools to Address Returns

The current landscape of tools designed to reduce online apparel returns is both active and fragmented. Most solutions focus on a specific moment in the customer journey: the point at which a consumer decides whether to purchase an item. The underlying assumption is consistent across the market: returns are primarily caused by uncertainty, particularly around size, fit, and product representation (McKinsey & Company, 2023; Coresight Research, 2023).

This assumption has driven significant investment in visualisation and fit-related technologies. Virtual try-on tools, augmented reality applications, and avatar-based systems aim to help consumers better anticipate how a product will look. In parallel, sizing engines and recommendation tools use historical data, brand-specific measurements, and customer inputs to suggest the “right” size. More advanced approaches extend to 3D body scanning, attempting to match precise body data with garment specifications.

A first major group of tools focuses on virtual visualisation and simulation. These include virtual try-on solutions deployed or tested by players such as Google Shopping, Amazon, Old Navy, Zyler, and Zeekit-related applications, as well as avatar-based or augmented reality solutions developed by Yoox, Zara, Zalando, and ASOS. Their objective is to help consumers better imagine how a garment might look on their body, thereby narrowing the gap created by the absence of physical fitting.

A second category focuses more directly on sizing accuracy, through recommendation engines and virtual sizing support, such as those offered by True Fit, Fit Analytics, and similar providers. These systems typically draw on garment measurements, brand-specific size variations, and customer profile data to suggest the most appropriate size.

A third set of tools pushes this logic further through 3D body scanning and digital body modelling, with providers such as 3DLook, Fit3D, Styku and others seeking to create more precise body data that can be matched against garments.

These tools have improved aspects of the purchasing experience. They reduce certain types of friction and can increase confidence in product selection (Coresight Research, 2023). However, their impact on return rates at scale remains limited. Adoption has grown, but aggregate return levels have not shifted significantly. This raises a practical question: are these tools addressing the main drivers of returns, or only a subset of them?

A similar pattern appears in the use of data-driven analytics. Machine learning models are increasingly used to predict return risk, analyse customer feedback, and identify product-level issues. Providers such as Returnalyze, Newmine, and Fit Analytics operate in this space. At the operational level, companies have invested in systems developed by firms such as AI2Easy, PrimeAI, Optoro, ReturnLogic and others, that optimise what happens after a return, routing items to resale, refurbishment, or recycling channels. These tools improve efficiency and recover value (European Environment Agency, 2024; Hvass & Pedersen, 2019), while third-party logistics providers such as GoBolt, Loop, AfterShip, and Frate support the execution of reverse logistics processes, but they do not fundamentally change the conditions that generate returns in the first place.

In response to these limitations, some technology providers are beginning to combine previously separate functionalities, integrating elements such as 3D body scanning, size recommendation systems, and product visualisation within unified frameworks. In parallel, a smaller number of initiatives are taking this even further by exploring extensions beyond product-level optimisation. For example, IDNTFY is incorporating consumer-driven components, including structured feedback loops and community-based inputs, aimed at better understanding purchasing behaviour.

Building on these emerging approaches that incorporate behavioural dimensions, there are also early signs of movement on the market and policy side. Some governments and retailers are testing stricter return conditions, including shorter windows or return fees. Others are experimenting with incentives for lower return rates or alternative models such as “keep and donate” (McKinsey & Company, 2023; NRF, 2023). These approaches remain uneven and are often implemented cautiously, given the potential impact on conversion and customer satisfaction.

Across all these efforts, one feature stands out: they are largely developed in isolation. Most initiatives are led by individual companies or technology providers. There is limited evidence of shared standards, common data frameworks, or coordinated industry approaches. This contrasts with other systemic challenges—such as sustainability reporting or supply chain traceability—where pre-competitive collaboration has been more common.

The current landscape shows that the industry is not lacking solutions; it is applying them in the wrong frame. Most tools are designed to reduce uncertainty, yet a growing share of returns is driven by behaviour, incentives, and system familiarity. As long as these drivers remain unchanged, improvements in accuracy will deliver only marginal gains. The limitation is not technological. It is structural.

Sustainability Implications of Returns: From Hidden Externality to Systemic Impact

The environmental impact of returns is often underestimated. Not because it is small, but because it is not fully measured.

Most sustainability assessments focus on forward supply chains—production, transport, and distribution. Returns are typically treated as a secondary effect. In practice, they introduce additional flows, repeated transport, and, in many cases, products that are never used. When these elements are included, the overall footprint changes significantly (European Environment Agency, 2024; Niinimäki et al., 2020).

The most important impact does not come from logistics. It comes from products that are produced but not used. Studies show that emissions associated with unused returned apparel can be several times higher than those generated by reverse logistics alone (Hvass & Pedersen, 2024). A meaningful share of returned items is never resold. These products carry the full environmental cost of production without delivering any functional value (IDNTFY, 2025; European Environment Agency, 2024; Vogue Business, 2023; BBC, 2023).

At scale, this represents a structural inefficiency. It is not a marginal issue linked to transport or packaging. It is a question of overproduction combined with non-use (Niinimäki et al., 2020).

The logistical dimension remains significant. In Europe alone, return transport is estimated to generate over 10 million tonnes of CO₂ annually, reflecting the scale of reverse flows within e-commerce systems (Statista, 2024). Multiple delivery cycles can significantly increase the footprint of a single item (Statista, 2024; Niinimäki et al., 2020).

What is often missing is a full lifecycle view. When return rates decrease, the impact is not incremental, it can be substantial. Modelling shows that reducing returns from typical levels to lower ranges can lead to large reductions in emissions, precisely because fewer products are produced and wasted (IDNTFY, 2025).

Regulation is starting to reflect this reality. The European ban on the destruction of unsold apparel under the Ecodesign for Sustainable Products Regulation (ESPR), marks a shift in how waste is addressed. By restricting disposal and requiring greater transparency, it pushes companies to reconsider how excess inventory is managed (European Commission, 2024). It also signals a broader move toward internalising environmental costs that were previously external. As regulatory pressure increases, the environmental cost of returns is likely to translate into direct financial and compliance risk for companies (European Commission, 2024; CMS, 2026; Transition Pathways, 2025).

However, the scope remains partial. The regulation targets what happens after products are produced and returned. It does not directly address why return volumes are high in the first place. As a result, there is a risk that the system adapts without fundamentally changing, redirecting goods rather than reducing the need for them (European Environment Agency, 2024; Vogue Business, 2023).

This highlights a broader issue. Returns sit at the intersection of production, consumption, and behaviour. Treating them purely as a waste problem captures only part of the picture. At the same time, the cost of returns is shifting from an externality to a managed liability.

Returns do not only reflect inefficiency in distribution, they are directly linked to production decisions, particularly in models that prioritise volume and availability over precision. To fully account for their environmental impact, they need to be considered as part of a system, one where production decisions, consumer behaviour, and business incentives are closely linked.

Systemic Challenges to Reducing Returns: From Issue Isolation to System Misalignment

The persistence of high return rates, despite widespread investment in tools and processes, points to a deeper issue: the system is misaligned with the actual drivers of returns. It is optimised for flexibility and growth, not for return reduction. Several constraints contribute to this misalignment, and they reinforce each other.

A first constraint lies in the absence of clear ownership of returns within organisations. Responsibility is typically distributed across multiple functions, including logistics, customer service, merchandising, sustainability, and marketing, without a single point of accountability. At the same time, the impacts of returns are fragmented across different areas of the business. Functions such as logistics, planning, inventory management, and finance are affected in distinct ways, making it difficult to form a consolidated view of their overall impact and to prioritise coordinated action.

A second challenge relates to the lack of standardisation across the industry. Sizing systems remain inconsistent between brands, product descriptions vary in structure and quality, and detailed technical data is not always available at the SKU level. In many cases, companies do not maintain comprehensive product technical sheets that can be reliably used by digital tools. This limits the effectiveness of sizing algorithms, virtual fitting systems, and data-driven recommendations. Technology can process data, but it cannot compensate for incomplete or inconsistent inputs.

The practical limits of current technologies pose another challenge. Tools such as 3D body scanning or avatar-based fitting promise greater accuracy, but their real-world use is more complex. Data must be captured correctly, interpreted meaningfully, and integrated into decision-making. In practice, this chain is fragile (Gill, Simoen, & Hoque Tasmin, cited in The Interline, 2023). Consumers may not follow scanning protocols, may not understand the output, or may not be willing to engage at all. At the same time, many systems rely on approximations or self-reported inputs, which reduces precision. The result is a gap between technical potential and actual usage.

Data itself introduces another layer of complexity. The effectiveness of advanced tools depends on access to detailed personal and behavioural data. This raises questions around privacy, ownership, and trust, particularly in a context of evolving regulation and heightened consumer sensitivity. Even when data is available, companies face constraints in how it can be used and shared. This limits the scalability of solutions that depend on rich, individual-level information.

More fundamentally, many existing approaches are built on the assumption that returns are driven by uncertainty. As Section 4 highlights, this is only part of the picture. A significant share of returns is linked to intentional behaviour—trial purchasing, low commitment, and system familiarity. In these cases, improving accuracy does not address the underlying driver. The issue is not what consumers know, but how they choose to act within a system that makes returns easy and acceptable.

Regulation, while evolving, reflects a similar pattern. The European ban on the destruction of unsold goods under the ESPR addresses a visible consequence of the system – waste – but does not directly influence return behaviour or purchasing incentives. There is a risk that the problem shifts rather than disappears, with goods moving into secondary markets or alternative channels (European Environment Agency, 2024; Vogue Business, 2023). The underlying dynamics remain largely unchanged.

As a whole, these challenges highlight a consistent theme: returns are sustained by the interaction of multiple factors that are rarely addressed together. Fragmented standards limit technology, data constraints limit scalability, behavioural patterns limit effectiveness, and incentives often point in the opposite direction.

This misalignment demonstrates that reducing returns cannot be achieved through isolated initiatives. It requires coordinated action across the industry, combining technological, behavioural, and economic levers (Hjort & Lantz, 2016; Janakiraman et al., 2016).

Conclusion and Recommendations

This White Paper set out to examine whether current approaches effectively address the root causes of high return rates in online fashion, or whether they primarily optimise around them. The analysis indicates that existing responses remain partial. While technological tools, data analytics, and reverse logistics solutions have improved elements of the purchasing and post-purchase process, they have not materially reduced return rates at scale.

The analysis suggests that this limitation comes from a mismatch between how returns are usually explained and what actually drives them. Much of the literature and industry practice continues to emphasise product-level factors such as size and fit. However, the findings presented in this paper point to a broader system in which returns emerge from the interaction between consumer behaviour, technological tools, and commercial incentives.

Purchasing practices shaped by social and marketing dynamics, low perceived cost of returning, and business models prioritising revenue growth may collectively sustain high return levels, rather than reduce them. These dynamics also carry direct sustainability implications, as high return volumes contribute to increased transport emissions, increased use of resources linked to handling and repackaging, and the disposal of a substantial proportion of returned goods, making return practices a direct driver of waste generation and environmental impact.

Within this context, current interventions tend to be undertaken individually by companies and focus on isolated components of the system. Fit technologies and visualisation tools address specific issues at the point of purchase. Data analytics and reverse logistics improve efficiency after the return has occurred. More recent integrated approaches begin to connect these elements, and a limited number of initiatives extend further to incorporate behavioural dimensions. However, these remain exceptions rather than the norm.

As a result, returns cannot be understood as a purely operational inefficiency. They are a structural feature of a system designed for flexibility, volume, and low commitment (Hjort & Lantz, 2016; Janakiraman et al., 2016). Any meaningful reduction in return rates therefore requires interventions that operate across this system, rather than within a single layer of it. This includes aligning incentives, increasing transparency on the true cost of returns, and addressing consumer decision-making processes alongside technological improvements, while recognising that reducing returns is both a commercial and a sustainability imperative.

In light of these findings, addressing returns requires moving beyond isolated improvements toward a more coordinated and practical approach. The following recommendations outline concrete actions that can be initiated in the short to mid term, combining pilot initiatives, internal coordination, and simple changes to purchasing processes to support measurable progress over time.

Initiate cross-industry pilot programmes to address high-return areas

In the short to mid term (12–24 months), a limited number of pilot projects should be set up with 2 to 4 partners across brands, retailers, wholesalers, research institutes, and policy institutions, and with technology developers. These pilots should focus on product categories and customer segments where return-related costs are highest, with the aim of reducing return rates and improving profitability after returns.

Carried out in controlled settings, the pilots would test practical measures such as enhanced purchasing guidance, simple wardrobe-based decision support, and post-purchase reflection prompts. The main outcome should be a measurable reduction in avoidable returns, together with a clearer understanding of which approaches most effectively influence purchasing behaviour.

In the mid to long term (24–48 months), the results from these pilots could be brought together into a small set of shared indicators and measurement approaches. This would allow partners to compare findings and build a common evidence base through voluntary participation, without requiring full alignment across the industry. Over time, this shared base could support gradual improvements in areas such as sizing, product presentation, and return-related metrics, where differences between players continue to create inefficiencies.

Establish cross-functional task forces to address returns at the company level

Defining clear roles and ownership within an organisation is an essential step in managing returns more effectively.

Companies should set up small, cross-functional task forces bringing together merchandising, logistics, customer service, marketing, and finance. Rather than redesigning existing structures, these groups should concentrate on a limited number of categories or customer groups where returns have the greatest impact.

To support this work, practical training sessions should be introduced at the outset. These should focus on building a shared understanding of return drivers, key indicators, and the links between commercial decisions, operations, and customer behaviour. This helps ensure that all participants approach the issue with a common frame of reference.

The role of these task forces is to act quickly on concrete issues. This may include identifying products that are frequently returned, reviewing how they are presented to customers, adjusting communication, or revisiting certain commercial practices. Responsibilities can be clarified through this work, helping to reduce fragmentation and improve coordination across teams.

At the same time, initial efforts should be made to record and organise relevant information across the return journey. This can begin with straightforward and consistent tracking of return reasons, customer feedback, product performance, and selected indicators of purchasing behaviour, such as intent of purchase or patterns of repeat returns. As this information becomes more consistent, it can gradually be shared across functions to support better decision-making.

Over time, the experience gained through these focused actions can help strengthen internal alignment and lead to a more consistent approach to managing returns.

Deploy integrated, lightweight solutions to improve purchasing decisions

In the early stages, companies should work with technology providers and research partners to introduce simple, combined solutions that support customers in making more informed choices.

These solutions may bring together elements such as size and fit guidance, clearer product information, and prompts that encourage customers to reflect briefly on how and why they are buying an item. The focus should remain on tools that are easy to introduce and do not require major changes to existing systems.

Examples include improved product descriptions, visual cues that help set expectations, or short prompts that encourage customers to consider use, fit, or frequency of wear. These additions can help reduce uncertainty and improve the relevance of purchases.

The aim is to improve the quality of buying decisions, leading to fewer avoidable returns and more stable revenue. These tools can be introduced step by step and adjusted based on observed results, in coordination with internal teams and, where relevant, external partners.

Reallocate part of return-related budgets to prevention efforts

In parallel, companies should gradually shift a portion of their return-related spending toward actions that help prevent returns from occurring in the first place.

Rather than focusing only on handling returns once they happen, part of the existing budget can be used to test simple measures that influence purchasing behaviour. These may include basic customer segmentation based on return patterns, adjustments to return conditions for frequent returners, clearer communication at checkout, or small incentives for customers who consistently keep what they buy.

The aim is to identify practical measures that improve purchasing decisions without undermining overall sales performance. These actions can be introduced progressively and observed over one or two product cycles to assess their impact.

Early results may include lower return volumes in selected areas, reduced handling costs, and improved customer retention quality. Over time, this can support a more balanced allocation of resources between managing returns and reducing them at source, based on demonstrated business impact.

References

Asdecker, B. (2015). Returns management in e-commerce. Springer.

Appriss Retail. (2023). Consumer returns in the retail industry report.

Bennet, N. (2025). Retail Trends 2025: Planning for the New Era of Retail. Vena.

BBC. (2023). The environmental cost of online returns.

CMS. (2026). New EU rules to prevent the destruction of unsold textiles.

Coresight Research. (2023). US online apparel returns survey.

Djafarova, E., & Rushworth, C. (2017). Exploring the credibility of online celebrities’ Instagram profiles in influencing purchase decisions. Computers in Human Behavior, 68, 1–7.

El Kihal, S., Erdem, T., Schulze, C., & Zhang, W. (2025). Customer return rate evolution. International Journal of Research in Marketing.

European Commission. (2024). Ecodesign for sustainable products regulation (ESPR).

European Environment Agency. (2024). The destruction of returned and unsold textiles in Europe’s circular economy.

FashionUnited (n.d.), Global Fashion Industry Statistics, retrieved on 15/09/2025. https://fashionunited.com/statistics/global-fashion-industry-statistics

Financial Times. (2024). US retail’s multibillion dollar returns problem.

Hjort, K., & Lantz, B. (2016). The impact of returns policies on profitability: A fashion e-commerce case. Journal of Business Research.

Hvass, K. K., & Pedersen, E. R. G. (2019). Toward circular economy of fashion: Experiences from returns management. Journal of Business Strategy.

Hvass, K. K., & Pedersen, E. R. G. (2024). Environmental impacts of product returns in fashion e-commerce. Journal of Cleaner Production.

IDNTFY ME AS. (2025). Life cycle assessment report – AI-driven fashion shopping assistant.

Janakiraman, N., Syrdal, H. A., & Freling, R. (2016). The effect of return policy leniency on consumer purchase and return decisions: A meta-analytic review. Journal of Retailing, 92(2), 226–235.

LeRolland, M. (2024). Top Trends in Global Apparel and Footwear in 2024 and Beyond. Euromonitor

Lou, C., & Yuan, S. (2019). Influencer marketing: How message value and credibility affect consumer trust. Journal of Interactive Advertising, 19(1), 58–73.

Luo, X., Wang, L., Wu, Q., & Moriguchi, T. (2025). From stopping to shopping: A field experiment on return policies. Manufacturing & Service Operations Management, 27(4), 1224–1241.

McKinsey & Company. (2021). Returning to order: Improving returns management for apparel companies.

McKinsey & Company. (2023). The state of fashion 2023.

National Retail Federation. (2023). Retail security survey.

Niinimäki, K., Peters, G., Dahlbo, H., Perry, P., Rissanen, T., & Gwilt, A. (2020). The environmental price of fast fashion. Nature Reviews Earth & Environment, 1, 189–200.

NRF (National Retail Federation). (2023). Consumer returns in the retail industry report.

Oghazi, P., et al. (2018). Online purchase return policy leniency and purchase decision. Journal of Retailing and Consumer Services.

Patel, P. C., Karlsson, S., & Oghazi, P. (2025). Fits like a glove? Knowledge and use of size finders and high-end fashion retail returns. Journal of Innovation & Knowledge, 10(5), 100779.

Rokonuzzaman, M., et al. (2021). Return policy, no joke: An investigation into the impact of a retailer’s return policy on consumers’ decision making. Journal of Retailing and Consumer Services.

Statista. (2023). Influence of social media on fashion purchasing decisions.

Statista. (2024). CO₂ emissions from e-commerce returns in Europe.

Statista. (2025). Online apparel return rates and consumer behaviour by geography and generation.

Stöcker, B., et al. (2021). New insights in online fashion retail returns from a customers’ perspective and their dynamics. Journal of Business Economics.

The Interline. (2023). The billion-dollar return dilemma: Does body data hold the answer?

Tourangeau, R., Rips, L. J., & Rasinski, K. (2000). The psychology of survey response. Cambridge University Press.

Transition Pathways. (2025). EU draft rules on the destruction of unsold apparel and footwear.

UniformMarket (2025), Global Apparel Industry Statistics: Market Size and Trends 2025.

Vogue Business. (2023). Haul culture is fuelling returns. What can brands do?