Tech tools for combating poor fit and returns: data analytics, AI and VR

What is causing returns on this scale? A massive 46 percent of US
consumers in a recent Harris Poll gave ‘poor fit’ as the prime reason
for returning garments, ahead of damage (38 percent) and poor quality
(36 percent). This issue not only impacts the bottom line but also
contributes to textile waste and environmental concerns. Little wonder
a top priority for a third of global retailers is ‘making returns
processes more efficient and less costly’, as revealed in Asendia’s
recent ‘How to sell direct in the age of the conflicted shopper
survey’ and report, where 800 retailers were questioned.

However, many sales and operations directors want to see returns
stopped in their tracks, not just managed more efficiently. By
offering inventive ways to achieve a better fit and a more committed
purchase, online merchants could alleviate the operational
complexities – and cost – of returns.

Big names in e-commerce are swinging into action. Amazon launched
‘Prime Try Before You Buy’ earlier this year, as a renaming of an
older Amazon programme – Prime Wardrobe. The service delivers a batch
of clothes and shoes to signed-up customers, to try on before they
commit to buying anything permanently.

They have seven days to test out up to six items at a time.
Unwanted pieces can then be returned in a resealable return box with a
prepaid shipping label, and only the items the customer wants to keep
get charged for. This service is open to subscribing Prime members – a
smart way for Amazon to reward its loyal customers, and grow sales
while managing returned items in batch form, which should mean fewer
individual returns. With items coming back so rapidly, reselling them
should be easier too.

If retailers are serious about cutting transport miles though, the
ideal scenario is higher ‘fit-first-time’ rates, with no reverse
logistics required at all.

AI-powered sizing and style recommendations

AI algorithms can generate personalised product recommendations
based on customer data. Machine learning is also improving the
accuracy of sizing recommendations. A new generation of personalised
sizing plug-ins, and virtual try-on tools are coming to market to
boost consumer confidence in fit and style at the point of ordering. A
massive 85 percent of retail respondents in a recent Coresight
Research study either currently use or plan to use this
technology.

Over time, it is hoped these tech advances will both drive up
conversions, and reduce the likelihood of items not fitting and
needing to be sent back. Personalised online shopping services like
Stitch Fix, and Tog + Porter use machine learning and in-house
stylists to create personalised outfits after consumers take a quiz to
specify their sizes and style. These ventures are designed to help
customers find clothes that will suit and fit them, in a more
supported way, before clothes go out for delivery.

Similarly, personalised fit recommendation apps like True Fit and
uSizy can be easily integrated into fashion websites. These use basic
data provided by individual shoppers and algorithms to help people
choose the right size across lots of different brands. Understanding
customers’ preferences based on what they wear and their actual
measurements, allows web merchants to turn hesitation into confidence,
and bad size choices into good ones.

This July, Zalando launched a new AI-powered tool that enables
customers in Germany, Austria, and Switzerland to receive size
recommendations for certain garments, using predicted measurements
from photographs of them wearing tight clothing. Robert Gentz, co-CEO
of Zalando, called this “a step change solution in the industry that
will help customers find the perfect fit before delivery.”

Heralding the virtual try-on experience

Many in the fashion industry have high hopes for VR try-on
technology, to reduce customer uncertainty, and ill-judged purchases.
Google’s new virtual try-on feature shows what is possible. Users in
the US can select models across a wide size spectrum from XXS to 4XL,
encompassing diverse skin tones, body shapes, and hair textures.
Google is harnessing the power of generative artificial intelligence,
available through its search engine, and, to begin with, only usable
on women’s tops.

Shoppers can click on products with a “Try On” badge to select a
model to virtually try on tops from brands like H&M, Loft, Everlane,
and Anthropologie. The technology aims to provide consumers with a
means to envisage how the clothing might appear on them.

Meanwhile, the John Lewis Fashion Rental website lets customers see
how the dress looks on them online before they rent with a new “Try it
On” feature. Customers upload a headshot and sizing information to try
the outfit on virtually, at home. The aim is that the chosen item will
suit and fit first time, cutting out the need for several deliveries
and returns.

Data analytics can alert when returns are high

For smaller online sellers without generous IT budgets to invest in
new tech, there are alternative tried and tested ways to help shoppers
make better purchase decisions, which will result in fewer costly
returns. For example, customer reviews can inform purchasing
decisions, as can video clips of how clothes look, and
realistic-shaped models. Updated, accurate product descriptions with
high-quality visuals such as 3D images also help enormously.

Data analytics can identify patterns and trends related to poor fit
and returns. If brands and their logistics partners are carrying out
returns analytics to pinpoint problems, they can be proactive in
addressing sizing and design issues, as well as problems like damage
in transit, or hold-ups along certain delivery routes.

Being alerted to problems causing higher returns means brands can
see big improvements over time. Parcel shipping service providers are
investing in dedicated portals for their retail customers to access
data and receive real-time alerts to problems. For instance, the new
Asendia e-PAQ Returns platform, provides retailers with powerful data
analytics tools to address and reduce returns effectively. The
platform allows retailers to access a returns data dashboard, which
provides insights into which specific products (SKUs) are being
returned, the regions or countries they’re coming from, and the
reasons for return. This data enables retailers to identify trends and
take timely actions to minimize return rates.

With a dedicated portal, retail customers benefit from real-time
alerts and data that can help pinpoint problems related to returns,
such as sizing issues, design problems, damage in transit, or delivery
delays. Data-driven insights and tools can help smaller online sellers
in addressing returns effectively, even if they don’t have the
resources for the latest AI and augmented reality technology.

When changes are made the improvements can be tracked – for
instance, if sizing is wrong in a particular country, returns should
reduce once the issue is resolved. If parcels are arriving too late in
a certain city causing higher-than-average returns, the retailer knows
they probably have a problem with the outbound carrier
under-performing.

Excessive reverse logistics cost the earth

The imperative to act is clear. By leveraging data analytics,
AI-powered sizing recommendations, and immersive VR try-on
experiences, the retail industry can significantly reduce return rates
and bring an end to the era of wasteful multiple parcel
deliveries.

This isn’t just a matter of convenience; it’s a global call to
action for a more sustainable and responsible approach to retail. As
the fashion industry embraces these technologies and as consumers
adapt their shopping habits, we must take a crucial step towards a
greener, more sustainable planet—one where textile waste no longer
burdens our landfills, and the carbon footprint of e-commerce is
significantly diminished.