Online retailers since the beginning of online retail have been obsessed with conversion — and rightfully so. Conversion is the win, the buy, the moment at which all that work and expense of getting a customer to your site, showing them the right products at the right time, and assembling the proper price point and fulfillment proposition pays off — literally.

But in 2023, with ecommerce representing an ever-bigger piece of the retail revenue pie, retailers who are winning the market are focused on “un-conversions.” Yes, un-conversions, the biggest of which are realized through product returns.

Here’s one of those Murphy’s Law bits of wisdom: It is always cheaper to wind something up than to unwind something. As expensive as it is to get a customer — at least a new customer — to your site, prompt them to buy and ship them their purchases, it is even more expensive to have the customer send that something back for a refund.

The National Retail Federation says 16.5% of all U.S. online sales were returned in 2022. Furthermore, nearly 11% of those returns were fraudulent returns, according to the NRF. That makes returns a $212 billion problem and fraudulent returns a $22.8 billion problem. That’s $212 billion in un-conversions a year.

Preserving those un-conversions as sales happens well before a customer arrives at your site. Online shoppers pay attention to return policies, so it is wise to have a well-thought-out, easily accessible and easily understood return policy. More than half of consumers (51%) surveyed by Signifyd last summer said they study a merchant’s return policy before deciding whether to buy from the brand.

And the work continues after the purchase is made: 65% of consumers surveyed by Signifyd said they would stop buying from a brand based on a poor return experience. Conversely, 62% said they would buy more from a brand based on having a good returns experience.

Narvar, the leader in online returns solutions, says that almost one-third of shoppers they surveyed attribute a slow refund to a poor return experience. And 48% said that an “immediate refund” is what makes for a positive return experience, according to Signifyd’s survey.

So, the solution is clear: Have a return policy that is easy for consumers to understand and execute and get the consumer their refund as fast as possible. Not as easy as it sounds. Consider that Signifyd’s survey also found that 24% of consumers polled admitted to returning an empty box or a box containing something less valuable than their original purchase as a means to keep an item without having to pay for it.

So what do you do? How do you provide seamless, liberal returns and fast refunds while protecting your business from return fraud and abuse?

Getting returns right requires the ability to instantly understand which return requests are high-risk, medium-risk and low-risk. That allows merchants to provide the best return experience for loyal customers while protecting their businesses from bad actors.

The good news? Such insights are available now. A merchant could use its own return and refund intelligence to establish patterns of return abuse among returning customers. Machine-learning models can provide the horsepower a merchant needs to process the information it has instantly and provide a recommendation or take automated action within a fraction of a second.

One merchant’s return and refund intelligence is a start, but it provides a relatively limited view. However, if that merchant could join with others in a consortium or turn to technology, like Signifyd’s Decision Center, that relies on intelligence from a vast network of merchants it could greatly expand its insights and receive richer risk assessments and recommended action based on risk. With the right risk assessment, you can confidently issue an instant refund for a low-risk shopper to ensure customer satisfaction and really step up the friction in the returns experience for high-risk or repeat abusers.

It’s not only an idea whose time has come, it is a powerful way to begin to make un-conversions a thing of the past.

Gayathri Somanath brings deep experience in ecommerce, payments and mobile apps to her work making commerce more profitable for merchants and more enjoyable for their customers. She has led Signifyd’s growth beyond guaranteed fraud protection and into areas focused on preventing consumer abuse resulting in illegitimate chargebacks and unauthorized reselling. She has a bachelor’s in computer engineering from Amrita University, Coimbatore, and her MBA from the University of California, Berkeley, Haas School of Business.