Why
it matters:
- Total merchandise returns are expected to have reached $890 billion in 2024, according to the National Retail Federation (NRF).
- Losses due to return fraud totaled $101 billion last year, according to the NRF.
- Merchants are increasingly tapping artificial intelligence to mitigate the high cost of product returns in a variety of ways: They’re using AI to make more detailed product descriptions and sizing information readily available to shoppers before they make their purchase, and they’re optimizing return logistics via the tech to determine the most economical ways to get items back to a store or warehouse, for example.
Merchants
are wrestling with the ever-increasing challenge of product returns,
seeking to staunch the flow of such items and minimize the costs of
processing them, while at the same time ensuring a seamless returns
process for their most loyal customers.
Total
returns are projected to reach $890 billion in 2024, according to a report from the National Retail Federation.
Now
artificial intelligence and machine learning are helping merchants
throughout the pre- and post-purchase processes. These solutions not
only seek to ensure that customers buy the right size and fit when it
comes to shoes and apparel, for example, but also help implement
customized return policies based on customer profiles.
Minimizing
the volume of returns is just one aspect of the challenge, however.
Merchants are also leveraging AI and machine learning to optimize the
return process itself and identify whether returned products are best
resold, donated, or discarded, or whether it might be more economical
to let the customer keep the item.
These
solutions are being implemented amid an increasing volume of fraud
related to product returns, particularly among online purchases, and
merchants are also deploying AI and machine-learning solutions that
seek to reduce these instances of fraud.
“By
embracing AI, retailers can create a more seamless, efficient, and
tailored shopping experience in order to drive customer loyalty and
boost sales,” returns solutions provider ReturnPro
said in the 2024 Holiday Edition of its Returns Report.
“As some retailers lean into discovering new ways to improve their
returns, there is still significant whitespace for the continued
adoption of these technologies.”
The
report found that many merchants are deploying AI and
machine-learning solutions to minimize product returns, including 36%
that analyze
customers’ purchase and return history to predict if they are
likely to return items, 39%
that identify
products with high return rates and the reasons why they are often
returned, and 25%
that use
programs to help determine accurate clothing sizes.
Preempting
returns by leveraging AI to ensure a good apparel fit
One
of the biggest areas of opportunity to minimize product returns is in
helping customers make the right choices before they buy an item
online. This is especially true for apparel and footwear, categories
in which sizing and fit can vary by brand.
Retailers
are increasingly analyzing data from past purchases to generate
AI-automated suggestions for product sizes that are most likely to
fit individual customers. Virtual try-ons, which lets consumers
digitally try on everything from lipstick to loungewear before
buying, are another tool that apparel merchants use to optimize
product selection.
“I
would say preventing returns up front is probably the easiest place
to deploy AI, and where we're seeing merchants use it the most,”
said Kristen
Kelly, VP
of Product at Loop Returns.
She
estimated that more than half of the merchants that use Loop Returns’
solutions are using some type of fit or sizing tool to help their
customers make the right selection before they complete their
purchase.
AI
not only helps retailers use data from customers’ past purchases,
but it can also be used to flag specific items that are most likely
to be returned, said Robert Johnson, Executive VP at ReturnPro.
[Read more: Inside Google’s Bold Push to Help Small Businesses Sell More Online Via the Magic of Generative AI]
I would say preventing returns up front is probably the easiest place to deploy AI, and where we're seeing merchants use it the most.
Kristen Kelly, VP of Product at Loop Returns
Identifying
trusted customers
These
return-prevention tools all start with the ability to identify the
customer and thus determine their propensity to return certain
purchases based on past behavior.
It’s
important for retailers to balance the need to drive customer
satisfaction against the cost of handling a return. Optimizing the
returns process for customers considered “trusted,” based on
factors such as past purchase history, return frequency, and reason
for returning merchandise, is one way retailers are doing this, said
ReturnPro’s Johnson.
These
valued customers can be given incentives to complete their purchase
by including assurances that product returns will be easy to execute,
he said. AI can help merchants streamline this process by leveraging
purchase history data and prompting effective responses.
“It’s
about understanding past purchases, and what has been kept and what
has been returned, and then being able to recommend something,”
Johnson said. “We’re absolutely seeing AI being used from
pre-purchase all the way to post-purchase as well.”
Walmart
is among the companies on the leading edge of this technology, he
said. The retailer has leveraged its vast trove of data to create
trusted customer profiles and reduce returns.
In
fact, nearly half (47.4%) of retailers surveyed said they have used
AI to make more detailed product
descriptions and sizing information readily available to shoppers
before they make their purchase, according to the ReturnPro report.
Fraud
prevention begins during the buying process: ‘AI
can flag those specific high-risk customers’
Losses
due to return fraud totaled $101 billion last year, according to the
NRF, which found that 13.7% of all returns were impacted by fraud.
More than half (52%) of retailers said they are implementing one or
more preventative measures.
In
fact, fraud prevention has emerged a key area where retailers are
deploying AI solutions to mitigate the cost of returns, said Zack
Hamilton, Head of Growth Strategy and Enablement at
parcelLab.
AI
can be used to detect customers engaging in specific behaviors
associated with fraud, he explained. AI can help sift through the
data to generate an appropriate return policy for each individual
customer.
“You
can start to flag those specific high-risk customers and put them
under a specific review for suspected fraud returners versus someone
who has been very loyal to your brand and rarely returns anything,”
Hamilton said.
For
those loyal customers, the return policy might be a little more
lenient, he said.
“It’s leveraging AI to identify patterns in order to really drive a more personalized experience,” said Hamilton.
Kelly
of Loop agrees that AI can be used at the time a customer is
initiating a return and flag returns that may be fraudulent, such as
if a customer may be returning an empty box or swapping the label
from an authentic product to another item.
Detecting
specific behaviors can help merchants make changes in the return
process in real time, such as switching over to manual processing of
the return so that they don't issue a refund until they actually have
the item in hand and they've inspected it, she explained. Merchants
can also limit fraudulent returns by only offering store credit, for
example, rather than a refund.
Optimizing
return logistics and resale opportunities via AI
Another
key area where retailers are using AI to minimize the cost of returns
is through the optimization of return logistics, helping determine
the most economical ways to get products back to the store or
warehouse.
Merchants
are looking for ways to minimize the costs of processing their
returns, and at the same time seeking to optimize the potential
revenues they can obtain from reselling them. They can use AI to help
determine if items can be returned to their original shelves, if they
can be donated to charity, or if it makes more sense to dispose of
them or let the customer keep them.
For
example, merchants should be able to analyze their return data to
determine how often products that are returned within a week are able
to be resold or how often the products are returned damaged.
“Making
a prediction up front about whether it’s worth getting that item
back or not is critical,” said Kelly.
If
merchants can predict that an item is likely to come back damaged or
too late to resell, or if it’s just going to be too costly to
process it, then they can just allow the customer to keep the item
and eliminate the processing costs.
A
key opportunity for minimizing the costs of product returns lies in
routing the products to the right location, based on machine
learning, said Kelly. Products are often reshipped after they have
been returned, which adds to merchants’ costs, she explained. If
more products are returned directly to the most efficient
destination—a specific store, warehouse, or other destination—it
can significantly reduce processing costs.
Johnson
of ReturnPro agreed that resale of returned products should be a key
objective for merchants. ReturnPro leverages AI to perform several
functions that help its merchant partners optimize the potential
revenues from product resales and minimize the costs of returns.
“We don’t even need humans to do it anymore, and that provides speed to get [returned products] back to the market,” he said.
CO— aims to bring you inspiration from leading respected experts. However, before making any business decision, you should consult a professional who can advise you based on your individual situation.
CO—is committed to helping you start, run and grow your small business. Learn more about the benefits of small business membership in the U.S. Chamber of Commerce, here.
