Customer story · Payment performance management

$3.6M / month · continuity offer · 10+ merchant accounts

They were paying full price for customers the checkout was quietly turning away.

Forty three of every hundred new customers this merchant paid to acquire were failing at the very first charge. Not because of the offer. Not because of the traffic. Because of where the payment was being sent. Six weeks after Beast Insights went live they were keeping seventy of every hundred, from the same ads, the same funnel and the same budget.

Anonymised at the merchant’s request

Six weeks, same traffic, same spend

Revenue secured since launch

$914,758

12 August to 23 September. 18.0% of the $5.08M this account approved in that time, across first charges, subscriptions, rebills and upsells. The figure on the merchant’s own dashboard.

More customers, same ad spend

+22.5%

Not more traffic. The same traffic, converting further.

Effective cost per customer

18% lower

The same media budget spread across 22.5% more customers, without touching a single ad.

Customers recovered

4,780

In six weeks. Each one enters a subscription that keeps paying at every step.

The problem

The most expensive leak in a subscription business is the one that looks like a marketing problem

When four in ten first charges fail, nothing in the business says “payments.” The dashboard says conversion is soft. So the team goes to work on what it can see: creative, landing pages, price points, affiliate mix, traffic sources. Everybody is busy and nobody is looking at the one place the money is actually going.

A decline looks like a customer who changed their mind. Most of these had not. The card was good, the funds were there, the shopper wanted to buy. The charge was simply sent to a merchant account that particular card issuer does not like, and it came back declined for reasons that have nothing to do with the person holding the card.

That is what makes this the cheapest revenue in the business. You already paid to get that person to the checkout. Winning them back costs no additional media, no new creative and no discount. On this account that leak was worth roughly 18% of everything it now approves: $914,758 in the six weeks since launch, counted across every step of the funnel.

Every quarter you wait, you are leaving five to ten percent of your revenue on the table inside your payments.

Which is why “doing nothing” is the competitor we lose to most often.

Of every 100 new customers they paid to acquire

First charge outcome on the traffic Beast managed, against the rate the account was running before. Same traffic, same offer, same spend.

Before

57 kept · 43 lost

With Beast Insights

57 kept + 13 recovered · 30 lost

Cleared anyway Recovered by Beast Still lost

Thirteen customers in every hundred, bought and paid for, that the business was writing off before anyone knew they existed.

The work

What the merchant had to change: nothing

This is the part that usually gets disbelieved, so here it is plainly. The merchant did not rebuild the checkout, did not open a new processor relationship, did not touch a line of code and did not change a single campaign. They signed on 10 August. Beast was live on 12 August. The first measured week closed on the 16th.

Payment performance management means Beast takes responsibility for how the payment stack performs, rather than handing over a report about it. It sits in the flow of funds across the processors already in place. On this account it starts with the first charge: when one arrives, Beast reads the card’s issuing profile and sends the transaction to whichever of the merchant’s existing accounts has the best track record with that profile. Same cards, same customers, same money, better destination.

The programme on this account covers both halves of the problem: getting the first charge approved, and recovering the charges that still fail. The figures on this page are the first charge result over the first six weeks, measured net of anything that later cancelled or refunded.

Rollout went from 15.5% of new-customer traffic in week one to 97.4% by week four, largest opportunity first, and has held above 97% since. The figures below are the performance of every first charge Beast engaged, including the ones it could not improve, which is what a merchant at full coverage gets.

Share of new customers whose first charge cleared

Every first charge Beast engaged, week by week, against the rate the account was running before.

69.9%Week 112-16 Aug 73.6%Week 217-23 Aug 67.7%Week 324-30 Aug 70.9%Week 431 Aug-6 Sep 70.0%Week 57-13 Sep 68.4%Week 614-20 Sep

57.1%, the rate before Beast went live. Measured over the sixty days prior, fixed, and never recalculated since.

The line does not climb because the gain arrived at once: week one already sits 12.8 points above the old rate, and six weeks later it is still there.

The part most people miss

One saved customer does not pay once. They pay at every step for as long as they stay.

A declined rebill costs you one payment. A declined first charge costs you the entire customer, because nothing downstream ever happens: no subscription, no renewals, no upsells. So the 4,780 recovered charges are not 4,780 transactions. They are 4,780 subscribers who would not otherwise exist, each one now moving through the same funnel as everybody else.

And the compounding is measurable, not theoretical. Traffic Beast engaged outperformed traffic it had not yet reached at every step of the funnel in the same window, most of all at the step that matters most after the first charge, the subscription start:

Every charge Beast engaged against traffic not yet reached, same days, same funnel, 12 August to 20 September 2026.
Funnel stepNot yet managedManagedDifference
First charge58.5%70.0%+11.5
Subscription start76.9%84.4%+7.5
Upsell, step 361.3%63.2%+1.9
Upsell, step 461.7%63.3%+1.6

Rebills are excluded from this table: almost all rebill volume in the window ran on traffic not yet managed, so there is not enough managed rebill data to compare honestly. It is left out rather than reported thin.

That is the whole argument for fixing the first charge in a subscription business. The gain does not stop at the front door. It follows the customer through the subscription start, through the upsells, and through every renewal after that, which is why the revenue below is counted at every step and not just the first charge.

Revenue secured since launch, 12 August to 23 September 2026, as reported on the merchant’s dashboard.
Funnel stepRevenue securedOf revenue approved
First charges$561,604$3,123,755
Subscriptions$168,889$933,839
Rebills$163,234$906,192
Upsells$21,031$115,367
Since launch, 43 days$914,75818.0% of $5,079,161

How it is counted. 18.0% of the customers this account approved since launch were rescues, customers who would have been declined at the old approval rate. That share is applied to every dollar approved in the same period, at every step, because a rescued first charge carries its subscription, rebills and upsells with it. It is the same figure the merchant sees on their dashboard and the same basis the engagement is measured on. At the six-week rate it runs to roughly $640,000 a month.

Because you should not take our word for it

How the number is kept honest

Approval lift is the easiest claim in payments to fake, so here is exactly how this one is measured. The merchant sees these same figures on a live page that refreshes daily, and can reproduce every one of them in their own dashboard.

01

The benchmark was fixed before the work started

57.1%, measured over the sixty days before go live and never recalculated since. A benchmark that moves absorbs the vendor’s own gains until the comparison quietly reads zero.

02

Refunded and cancelled orders do not count

A charge that clears and refunds a week later did not make anybody money. It is stripped out of every figure here. Most approval numbers in this industry are not.

03

The headline includes our misses

Beast engaged 37,202 first charges and could not find a better home for 8,681 of them. Those went through as they would have anyway and cleared at 39.6%. Every one of them is counted in the 70.0%. On the charges Beast did place, the rate was 79.2%. We do not headline that number, because which charges get placed is not random, and quoting only the placed set would flatter the result.

04

Coverage is stated, always

A result quoted at partial coverage without saying so is the oldest trick in performance marketing. Rollout ran 15.5% in week one to above 97% from week four on, and it says so under the chart. Across all first charges, including everything Beast had not yet reached, the account moved from 57.1% to 68.1%.

Whether this applies to you

The conditions that made this work

This result is not universal, and the shape of the account matters more than the vertical. It went the way it did because of a handful of conditions, and if you recognise them in your own account, the same leak is very likely sitting in your stack right now.

Recurring revenue

A subscription, continuity or membership offer, where the first charge opens a relationship rather than closing a sale.

Ten or more merchant accounts

Managing performance means having somewhere better to send a charge. A single account stack has no alternative destination.

Real volume

Roughly $300K a month and upward. Below that the patterns are too thin to act on with confidence.

A first charge rate you cannot explain

If new customer approvals sit well below your rebill rate and nobody can say precisely why, that gap is the money.

Willingness to let us act

Beast works in the flow of funds. Reporting alone would have found this leak and fixed none of it.

About the anonymisation. This merchant asked not to be named, and we would have withheld the identifying detail anyway: no vertical, no processing partners, no issuing banks, no merchant account identifiers and no decline profiles. Approval rates, customer counts, revenue scale and dates are reported as measured over 12 August to 20 September 2026. No figure here has been rounded in Beast Insights’ favour.

Beast Insights provides payment performance management for subscription businesses: we take over the payment stack’s performance as an outcome, acting in the flow of funds across every processor to get more first charges approved, recover the ones that fail and keep every downstream step healthy. We are paid a share of what we produce.

Measured 2026-08-12 to 2026-09-20 · six weeks · 37,202 first charges engaged · Prepared 2026-09-23

Find out what your first charge rate is costing you

Most first conversations surface a specific dollar figure inside thirty minutes, using your own data. If the number is small, we will say so.

Talk to us