Almost Half of Subscription Churn Caused by Preventable Friction in Failed Payments

The bane of subscription brands and merchants is churn, but what triggers that churn is often a failed payment that starts a chain reaction resulting in a canceled subscription. This is preventable with the right combination of technology and strategy.

In a conversation with PYMNTS, FlexPay CEO Darryl Hicks told Karen Webster, “If you look at where the churn is coming from in your book … typically 48% of the churn is coming from failed payments. We see that a lot of merchants think churn is happening in customer service or [because] people are dissatisfied with the product or service. But if they’re not focused on the failed payments, they’re missing 48% of the problem.”

Hicks told Webster that about half of that failed payments problem is fixable, but many companies in subscription and recurring billing businesses aren’t addressing critical areas, from the use of transaction data and tech to some basics around the timing of batched payments.

According to Hicks, if companies aren’t taking certain steps to prevent failed payments behind the scenes, artificial intelligence (AI) and machine learning (ML) alone won’t prevent subscriber defections.

“Companies that are world-class top decile performers … put a tremendous amount of focus on back-office optimization, efficiency [and] best practices for managing the book, but it’s nowhere near as prevalent as we would like it to be,” he said.

Hicks told Webster that top decile performers “have a very data-driven approach. Specifically, they’re looking for points of friction where … maybe things are getting in the way, to the detriment of the customer’s relationship with the merchant.”

Payments friction tops his list.

Pointing to the pandemic growth of card-not-present (CNP) transactions, he said that with the “problems inherent in existing payment rails and the ever-increasing problems of fraud, and the constraints that’s putting on issuing banks with being able to approve the transactions they want to, we are seeing way more friction than we have at any point in history right now.”

See also: Better Data Solves Subscription Commerce’s Growing False Declines Problem

Covering the Bases on Failed Payments

Going back to what top performers do — and what others don’t — Hicks told Webster that a handful of best practices, combined with the astute use of technology, can put a dent in failed payments.

Mentioning “low-hanging fruit” approaches that should be engrained in all recurring billing operations, Hicks said that “there’s the timing of when you process your payments, and generally speaking, you want that to be in the early afternoon Eastern time on business days. That’s when you’ll have, by and large, the best success rate in processing subscription transactions.”

“Banks are more sensitive to transactions that are running at off hours,” he explained. “They think it’s a higher likelihood that this is a compromised card with a bad actor running a transaction in the middle of the night, trying to see if it’s a good card or not. All fraud systems at the issuing side are sensitive to the timing of when a transaction is processed.”

However, Hicks added that many organizations run payments batches at night, thinking they’ll get faster processing times due to less server traffic. He called this a “bad, bad idea.”

Another tip for avoiding failed payments is making sure that merchant category codes (MCCs) are accurate.

“Every merchant account … on the Visa, Mastercard or AmEx rails … has a four-digit merchant category code associated with it. Making sure you’re on the right [MCC] can have a massive impact on the approval rate on first transactions.”

Along with these steps is ongoing data hygiene, which runs the gamut from removing special characters from name and address fields to staying within character limitations to deciding whether to submit ZIP codes in transaction data.

“All these little things start to have an impact on your approval rates — things like leveraging the recurring gateway flag, which most modern gateways support,” he said. “When should you use that, when should you not and on which issue” can have an outsized impact.

See also: Subscription Payment Declines Are Usually Random, Often Costly, but Largely Avoidable

Payment Authorization Management (PAM) Steps Up

Hicks is a major proponent of adopting a payment authorization management (PAM) strategy, which he said many top performers use to reduce failed payments as part of an “invisible recovery” strategy.

“[Payment authorization management and avoiding failed payments are] something that top-performing merchants really spend a lot of time thinking about,” he told Webster. “It’s typically someone on the finance team, or it might be someone on the product team who’s responsible for optimizing the P&L, but at a certain scale, there’s someone who’s focused on payment authorization management and looking at what we can do to make … our interactions with our customers as frictionless and seamless as possible.”

Not only would that be better for the merchant in terms of more revenue, more profits and a higher lifetime value, but it would also be better for the customer in terms of less annoyance, less frustration, and getting the goods and services that they signed up for, Hicks explained.

Probing into what happens in the absence of PAM, Webster asked Hicks to quantify the problem as he sees it.

“On average in America right now, debit rails are slightly better than credit rails for most issuers, but it’s hovering in the 20% range,” he said. “That’s the typical decline rate for most high-quality merchants, especially on annual rebills. It tends to be 26%. On monthly rebills, it can be in the high teens [or] slightly better, but … it’s a meaningful chunk of a merchant’s business.”

Hicks said that the use of PAM, together with attention to data detail, is helping more subscription merchants recover up to half of the 50% of churn attributed to failed payments.

This is “where you have to have world-class digital engagement, whether that’s texting or email,” he said. “We even see some merchants using direct mail. But again, you want to do the least amount of that as possible because it’s expensive [and because] we have really good data that shows it’s not just the … failed payment that’s bringing down lifetime value. Believe it or not, [involving the] customers [to solve a failed payment] turns into voluntary churn.”

In a curious reversal, Hicks said that the more a company tries to communicate with failed payment customers, “churn numbers [are actually driven] up. You’re giving them an opportunity to express their dissatisfaction, or you’re just irritating them.”

The better way is to manage failed payments invisibly without bombarding customers with messages. This approach also scales well, delivering value faster to more customers.

“It’s across the entire business … making all the unit economics of the business a little bit better so that you get … this compound effect,” he said. “That’s why it’s also important to make sure that when you’re tackling these problems, you’re using the best possible solution to the greatest extent that you possibly can … because that’s how you’re going to yield the benefits, which is … one of the reasons we started looking at machine learning and AI.”

But Hicks warned against throwing money at AI in the belief that it can sort things out on its own with the AI having been programmed with deep insights and trained properly to be able to act independently to optimize payment recovery.

“Even a rules-based system that knows what all the dials and levers are in order to move them efficiently can perform [fairly] well,” he said. “[But only the] combination of lots of data [and] a trained team who knows exactly what it is they’re trying to influence, combined with the AI and machine learning, actually yields these outsized results.”