Walk into a busy gym, and you can usually tell who’s drifting. Fewer check-ins, less eye contact, the front desk stops recognizing them by name. That’s the churn story most operators expect and, frankly, the one they’ve learned to manage.

The sneakier version does the most damage.

The member is still showing up, still training, still taking classes — and a payment fails quietly in the background. Nobody catches it. Access eventually gets cut. And it looks like “they left,” even though they never made that choice.

Charles Rosenblatt has been sitting in that gap for a long time. As CEO of Butter, his read on the fitness space is direct: a lot of what gets labeled churn is actually failed payments that were never properly recovered.

Who is Charles Rosenblatt?

Charles helped scale Payoneer into a public company and worked across issuing, acquiring, loyalty, and cards. He also built an early Buy Now, Pay Later product in the U.S. through payroll deduction at Capital One in 2001, which is a pretty specific kind of “solve the payment problem first” mindset.

He joined Butter about eight months ago and became CEO five months in. The company itself has been running for six years, founded by a former Microsoft statistician who noticed that subscriptions were disappearing at scale, and the culprit wasn’t dissatisfaction. It was due to failed payments. He built an internal system at Microsoft to recover those algorithmically, saved hundreds of millions in recurring revenue, ran the same model at Scribd, and then started Butter to do the same across industries. Charles came in to lead that next chapter.

That matters in fitness because subscription billing is basically a trust business. Members shouldn’t be punished for a bank error, and operators shouldn’t need a back-office team playing whack-a-mole just to keep revenue steady.

How Butter Works

Butter’s primary mechanism is to separate access from collection, then recover failed payments in the background rather than punishing members at the front door. It sidesteps the blunt retry approaches that frustrate members and, over time, teach banks to decline charges more readily.

One unnamed large fitness provider had 23 people in a room somewhere, pressing buttons to manually retry failed payments. Other gyms handle it by having front desk staff confront members when they walk in. Butter ran an internal test to see just how long that gap could stretch: a team member intentionally let a payment fail at a class-based gym and attended six classes before anyone said a word. At roughly $20 per head per class, the gym absorbed $120 without realizing it. Multiply that across a member base, and the math gets uncomfortable fast.

Algorithmic Recoveries That Avoid Blind Retries

Many systems default to the same retry schedule regardless of context — Day 1, Day 3, Day 5, Day 7 — cycling through the same pattern. Charles calls it the “woodpecker” approach, and the problem is real: hammering failed charges without context can degrade authorization performance over time, even on cards that would otherwise clear.

Butter builds a recovery path for each failed transaction, using machine learning on 128 transaction attributes. If a debit card fails on the 12th, the system may hold until the 15th, as payroll cycles indicate a deposit is likely, then retry early in the morning when funds tend to post. A prepaid card failure might skip retries entirely and move straight to direct outreach. Authorization rate functions a bit like a credit score for your billing operation, and sloppy retry behavior can cost you later, including on charges that had every reason to. 

Predictive Shutoff That Protects High-Overhead Services

Butter also uses predictive shutoff. The idea is to keep access open for likely payers while recovery runs, then cut access for unlikely payers before expensive services get consumed. A coached class has real marginal cost, whereas an open-gym visit usually doesn’t. Treating them the same is how clubs lose money in the messiest way.

What Results Look Like When Recovery Is Done Well

Butter says it drives a five percent average lift in annual revenue, with a reported range of three to 10 percent. Recovered members, on average, stay around four additional months. As for how pricing works, Butter takes the initial recovery month, and the client keeps the rest of the retained revenue. Across industries, Butter reported $1.5B+ in ARR recovered. 

Final Thoughts

Fitness is a particularly clean target for this kind of solution because its underlying infrastructure is centralized. A small set of platforms sits beneath an enormous share of clubs and studios. Integrate at the platform level and recovery reaches thousands of facilities at once, rather than selling the idea gym by gym.

The competitive map tilts in Butter’s favor because most recovery vendors were built within Stripe-adjacent ecosystems, while fitness largely operates outside them. With more than 70 percent of staff on the development and machine learning side, Butter is leaning into the data rather than manual operations to close the gap — and in an industry that’s been handing this problem to front desk staff for years, that’s a meaningful shift in how the work gets done.

About Robert James Rivera
Robert is a full-time freelance writer and editor specializing in the health niche and its ever-expanding sub-niches. As a food and nutrition scientist, he knows where to find the resources necessary to verify health claims.