Leveraging consumer review data to reduce product returns

---

Product returns can be costly for retailers. See how consumer data can be used to improve the shopping experience and decrease the likelihood of returns.

---

Product returns have been an issue for as long as retail has existed. According to new research from Appriss Retail and the NRF, total returns in the US retail industry are expected to reach $743B in 2023 (14.5% of total retail sales). 

That figure is down from $816B in 2022, but it’s still a staggering amount. In addition to significant operational challenges and financial costs to retailers, product returns can negatively impact customer satisfaction, not to mention the environmental impact.

It’s no surprise that fraud accounts for a significant portion of total returns. The Appriss Retail/NRF study estimates that over $101B worth of returns are due to fraudulent and abusive returns (approximately 14% of total returns). Fortunately, technology innovations are leveraging machine learning and advanced video capabilities to help prevent such losses.

While combating fraud is critical, it’s important to note that almost $650B of merchandise is returned for reasons other than fraud. Understandably, the percentage of returns for online sales (17.6%) is considerably larger than for in-store (10%) purchases. The inability to touch and demo products sold online can lead to purchases that don’t meet the shopper’s needs and may result in returns. A large portion of these online returns are likely preventable through a better data strategy.

To illustrate, I compared the online shopping experience for popular running shoes. A shoe’s aesthetics are relatively easy to communicate online with high-quality images and 360-degree videos. However, for avid runners, the fit of a running shoe is more important than aesthetics. If the shoes don’t fit, there’s a high likelihood of returns. Yet, communicating fit is an area with vastly different user experiences when it comes to online shopping of major footwear brands.

Nike: Excellent style cues, but fit is more difficult to discern. 

Nike’s Pegasus 40 includes several photos and videos to provide a sense of style. Nike encourages buyers to upload their own photos and tag @Nike on Instagram. They also feature cross-sell suggestions in a “complete the look” section. From a style perspective, Nike nails it.

Nike’s filtering at the product listing page (PLP) provides detailed filtering by product features such road vs trail, suitability for cold/wet conditions, everyday runs vs road racing, and more. These are important attributes, to be sure.

Shoe fit, however, was less obvious. A size guide, is linked from the main product display page (PDP), but it is essentially a generic comparison of different sizing scales (US Men’s, US Women’s, UK, EU). Clicking Reviews reveals a few ratings/ review summaries. It’s only upon further clicking More Reviews that you see more detailed information regarding fit and comfort.

Nike clearly has valuable data regarding product fit that it certainly uses as part of the product development process. However, accessing this data is not straightforward and is likely overlooked by many shoppers. More seamlessly integrating this data into the online shopping experience can improve customer satisfaction while simultaneously reducing a portion of product returns.

Adidas: Data drives enhanced filtering and comparative fit finder

Like Nike, adidas does a great job showcasing product aesthetics. The PDP for the Pureboost 23 running shoes contains high quality product images and videos. However, unlike Nike, additional fit information is more easily accessible. Upon expanding the Reviews section, comfort, quality, size and width information is clear.

Reviews can be filtered by other important criteria such as running activity (road, trail, treadmill, etc), distance, frequency, pronation (I over-pronate, so this is helpful information), and more.

adidas further leverages fit-related data with a Fit Finder tool from Fit Analytics.  This allows the user to input a comparison brand/item to recommend the equivalent adidas size based on thousands of users. According to the Fit Finder, my 8.5 in Nike running shoes are equivalent to an 8.0 in adidas. With a purported “82% chance”, knowing this half-size difference in fit likely prevents me from ordering the wrong size. 

Brooks: A fully tailored experience

The online experience for Brooks is similar to adidas. Expanding the reviews immediately summarizes fit (roomy vs snug), sizing and width. Similar to the adidas’ Fit Finder, Brooks uses True Fit which uses AI to recommend sizes based on data from 82M shoppers.

But Brooks goes one step further in leveraging its data to drive the right product for shoppers with its Shoe Finder. The Shoe Finder asks a few questions such as weekly running mileage, training plans, and recent injuries. Then, it asks the shopper to remove their shoes and observe a few things such as if their feet point inward or outward while walking, if you experience instability while standing on your non-dominant leg, etc. Along the way, it provides short “behind the science” explanations and concludes with product recommendations tailored to your needs. The process takes less than five minutes and feels a bit like speaking with an expert at a local running specialty store.

Better experience, fewer returns?

While data regarding the return rates for the above brands is not available, their return policies can serve as a proxy for the expected level of product returns.

Both Nike and adidas offer free returns. From experience, I can say that their process is easy and straightforward.

It’s notable, however, that Brooks has the longest return window of the three brands and their “run happy promise” is visible throughout the shopping experience. The extra attention paid to fit —a critical element for their customers— seems to result in the company’s greater confidence that customers will select the right products.

	Brooks	Nike	adidas
Return Window	90 Days	60 Days	30 Days*
Return shipping	Free	Free for Members	Free for Members
Worn shoes returnable?	Yes	Yes	Yes
Exclusions	None	Nike Clearance items	Personalized products

One size doesn’t fit all

While the concept of “fit” may not apply to all brands, integrating customer feedback data into the shopping experience can help reduce returns in other industries as well. Understanding the primary reasons for returns and integrating that data into the product selection process can improve the shopping experience and help ensure the right product is purchased. 

For example:

• In toys and games, “age appropriateness” from users is more informative than the age coding on packaging that is largely based on safety.
• Electronics such as televisions could include user assessments of sound, picture quality, suitability for video games, and more (in addition to technical specs).
• Appliance manufacturers could leverage users’ assessments of durability, power, specialty applications, ease of use, etc.

To be clear, some brands and retailers are already incorporating summarized consumer data in their product reviews. For example, Amazon reviews include quantifiable ratings for top features across many product categories. For others, leveraging consumer data as part of the shopping experience can help reduce returns, improve satisfaction, and ultimately improve the bottom line.