Payment inefficiencies (technical issues) impacting retention of subscribers, can eat up to 10% of total revenue! 

In this on-demand webinar, Vijay Menon of Butter Payments emphasizes that subscription businesses must actively manage and rectify payment failures to mitigate involuntary churn. He highlights that about 70% of payment failures are potentially fixable with the right strategies.

Understanding and addressing the nuances of payment processing can serve as a pivotal factor in boosting a company’s retention rates and overall financial performance.

This is highly valuable information for subscription businesses looking to improve their bottom line by tackling one of the more technical and often overlooked aspects of customer retention—payment processing efficiency.

This on-demand workshop will arm you with:

• Actionable Insights: Learn why payments fail and how to turn these failures into opportunities for retention and growth.
• Strategic Frameworks: Get equipped with the strategies that have helped top companies recover millions in revenue.

On-Demand Playback

Presentation  Slides (PDF)

Payment Teardown Presentation (PDF)

About Our Experts

Vijay Menon, founder/CEO of Butter Payments – where churn ends.

A decade ago, Microsoft, Dropbox and Scribd were all losing millions a year to subscriptions that were being “accidentally” canceled. Vijay’s work shoring up the financial backend helped his previous employers retain millions they otherwise would have lost. In the process of solving this problem, he developed the world’s leading core technology around payment authorization and recovery. On this premise, Menon started Butter Payments.

Butter is a payments platform laser-focused on ensuring that legitimate payments don’t accidentally fail, rescuing the 5% of total subscribers that silently churn every year. This delivers millions of dollars in found revenue every year for our customers and ensures their end-users continue to get the products and services they want.

Vijay is also a TED Speaker and author of A Brown Man in Russia.

Kathy Greenler Sexton, CEO, Subscription Insider

Kathy Greenler Sexton is the CEO of Subscription Insider and a recognized expert in subscription business models, market strategy, brand development, and information products. Subscription Insider is an information company focused on delivering news and insight for growing profitable subscription businesses.

Transcript

Kathy Sexton:
So welcome everybody. My name is Kathy Sexton. I am the CEO of Subscription Insider and welcome today to today’s session Payment. Tear Down, Unlock the Revenue You Didn’t Know You lost. I’m really excited to welcome Vijay Menon. He is the founder and CEO of Butter Payments, and VJ has really spent his career for companies like Dropbox and Microsoft and others really analyzing and understanding where inefficiencies are to recover payments, to prevent payments from failing in the first place, to really help the businesses he’s worked with in the past and to inspire the company that he founded Butter Payments. So I’m going to let Vijay kind of talk a little bit more about his background, but that’s why I am so excited to pick his brain today and really walk through where there are inefficiencies that we all should be looking at in our recurring revenue businesses.

So Subscription Insider, we are a media business that are focused on helping you subscription executives run and operate your operations more efficiently and more profitably. At Subscription Insider, we host news. We host webinars like we have today, and this fall we will be hosting subscription show. I’m excited, Vijay, you’re going to be with me on stage as we host to everybody coming to New York and I want to make sure it’s on everybody’s calendar. It will be October 22nd and 23rd in New New York, and we’re diving in to issues that hire recurring revenue companies face payment strategy. So it’s a great opportunity for you or your team who focus on retention to really help understand how to maximize all the payments and all the subscribers that you work so hard to bring in. So with that, I’m going to turn the microphone over to UVJ so you can get running and get this program started.

Vijay Manon:
Awesome. Thank you so much Kathy for the introduction and I’m looking forward to seeing you in person in New York and October should be a fun, timely. Thanks everyone for joining today. I’m hoping that you get a lot of concrete insights into solving what we call accidental churn or involuntary churn here today. Hopefully get a chance to go back to your teams, maybe learn a little bit of something new. I like to hopefully make sure that you take away some practical tips from this and that ultimately gets you excited about an area of the business that you’re likely under investing in. So we’re going to spend the next half hour or so talking about what we like to say, finding money in the sofa or unlocking revenue that you didn’t know existed or that you didn’t know that you’re losing and that is churn on account of payment failure.

And one place for us to start, Kathy would just be around defining the surface area of what we’re talking about. So when people ask why we’re called butter, it’s pretty simple. Butter is where churn ends, but the type of churn that we’re talking about is not your classic conception of active churn that your retention teams at large subscription companies are usually focused on. Active churn is what happens when you are paying Netflix $15 a month, you decide you don’t want to pay anymore, you log into Netflix, you click through a cancellation flow, they show a bunch of pictures of sad dogs to you and they ask you, are you sure you want to leave? And then ultimately you say, yes, I’m sure and I cancel and you lose your subscription. That’s not what we’re talking about. We’re talking about when Netflix charges you on May or on May 8th and they charge you $15 and the card payment doesn’t go through and you get a response from Stripe or from Aian or from Payment Tech or whatever you use to process payments and that payment fails.
Now you’ve got a choice. Do I churn the user immediately? And if you’re Comcast, you don’t get payment from a user, you cut cable, and if you are pg and e and you don’t get payment from a customer, you cut the lights. And for the subscription panelists in the room today in your business, when you don’t get payment from a user, you’ll cancel them or churn them from your service. But what I’m going to pause it to you is that there are a lot of users who are trying to pay you and actually just can’t and we want to fix that. So that’s what this session’s going to be about.

So we’ll spend some time on some introductions, getting to know me, spend some time on the state of involuntary churn today, connect it back to this broader topic of payments health, which I think is under talked about because a lot of organizations don’t actually have a pure payments team. Then we’ll run the tear down and talk about the impact to your business, which hopefully will be something that you guys can take away and take back to your team as an initiative for the year. So here is where I will start, which is that I did not come in to build butter because it had always been my dream to be a company builder or entrepreneur, be very honest about that. It had been my dream actually to be a basketball player and probably to play a lot of video games. And so getting really into sports got me into statistics and being really into video games got me my first job at Xbox Live back in 2013.
And the confluence of those two things is what brought me into the world of payment failure because when I was put in charge of retention for Xbox Live, the first thing I started to do was go look at the hours that people were playing games. So people are more engaged with the service, they’re more likely to stick with the service. I would look at how many friends they would add people add more friends on the service, they’re more likely to stick with the service. But beyond the basic engagement activities, I started to recognize that a quarter of all the churn at Xbox Live had absolutely nothing to do with how many hours Kathy was playing Halo. And actually a quarter of that churn, it was just happening because we couldn’t get payment from Kathy. So that’s what was happening and I had this aha moment of that seems like a lot a quarter of our churn is just because we’re getting error responses from payment tech and Adian and Stripe and Braintree.

That feels like a lot. And the feedback I got from higher ups at the time was, wow, you are right. That is lot. And secondarily, you can do nothing about it because you can’t reach into Kathy’s wallet and run her card. And I just didn’t believe that was true. So I started to look into the data and I started to recognize, well, there are certain times when payments go through at a higher rate and one thing I recognize is we were running a batch process for all of the users who signed up on May 8th. We would go bill them on June 8th, but when will we bill them? We’d bill them at 12 o’clock in Redmond, Washington. Well, if I signed up from India, when you’re running my card at 12 o’clock in Redmond, Washington, it’s one 30 in the morning in India. So now you’ve got a card that’s issued by the State Bank of India and it’s a Visa card and it’s being run in a different country and it’s being run at one 30 in the morning.

Now we’re not thinking about that when we run the billing, but this is what’s happening. And when that happens, a fraud response comes back and when the fraud response comes back, what do we do with Xbox? We churn the user. And when the user wakes up in the morning and they try and play the game, they’re confused. They can’t log onto the service. So this is what ended up happening. This was my aha moment. I’m like, okay, there’s something here. Now I’ve learned over the years that that’s a very reductive example. That’s probably about 5% of the reason why payments fail, right? And there’s a whole variety of reasons why payments fail. That was my introduction into realizing this is a huge problem and it’s one of these things that makes so much sense when you think about it backwards, but in the moment you’re not thinking about this at all. And as I continued my career, I started to see this was a problem everywhere. So I moved to Dropbox in the year prior to their IPO and the number wasn’t 25% of overall churn. It was 55% of overall churn. In fact, the majority of Dropbox users that were churning six months before an IPO were being churned by Dropbox. They weren’t asking to leave Dropbox. They’re being churned by Dropbox. 

Kathy Sexton:
Crazy.

Vijay Manon:
So it was a wild thing to discover and that’s where I’m like there move to scr. I saw the same thing happen. So there’s a trend here, a hundred million dollars subscription, recurring revenue businesses, huge black hole to payment failure, not a lot of awareness. There’s a minor awareness payment failure is a cost of doing business, but there’s not the awareness that it’s a driver of growth and that it is a massive problem. So that was my framework for, okay, if these businesses haven’t solved it, we got to build a business to solve this at scale. And you’ll notice that these are $300 million B2C businesses. But now that I’ve built out butter, really seeing this in the direct to consumer space, you’ve got a box, got to ship. If you can’t get the payment, the BarkBox isn’t getting there and my dog Bernard is not happy. So we have to fix that problem. And similarly in the V two B self-serve space, your Zoom, your Figma, your notion, you’ve got licenses and card volume and signups, you have this problem too. So I want to talk about fixing it and that’s where we’re going to go here.

So clearly you can see the problem involuntary churn is stealing your profits and your subscribers and we want to fix it, but we want to be smart about fixing it. We want to find the cases that are remediable. And the reason we’re talking about it is because most folks again don’t understand that the number one cause of churn for a hundred million dollars plus recurring revenue or subscription business is going to be payment failure. It’s going to be involuntary churn. What we’re talking about involuntary churn. You want to take the universe of failed payments and you want to be smart because you want to differentiate between the cases that are fixable and the cases that are not fixable. And so what we tend to find in the market is 70% of payment failure is fixable. It’s not a hundred. Anyone who tells you it’s a hundred is lying.

Anyone who tells you it’s zero is lying, it’s somewhere in between because some payment failure should be real. There is actually fraud going on. My card was actually stolen. That stuff, we don’t want to fix the payment failed, let it fail, let the user go. But about 70% of payment failure, when you look down to the deep error codes that are coming back from the issuing banks, and we’ll talk more about that, are actually just purely a configuration error. We didn’t run the payment at the right time or we didn’t pass the right fields and flags with the payment. And I’ll talk a lot more about what that means. But we want to identify this chunk of failed payments that we can fix without having to put the onus on the user to go call their bank. And it is the number one cause of churn.
And I say that because when we have any engagement with any subscription company, we just look at the hard data. So before we go into any sort of relationship with a subscription business, I will ask that subscription business to give me read only API keys. And the reason I’ll ask for that is because I want to look at their last 12 months and I want to look at the transaction history from Stripe, from Braintree, et cetera. And typically what I’ll find is for a hundred million dollars recurring revenue business, there’ve been at least $10 million in failed payments and sometimes $35 million or more. So I’m not looking one to 2 million, I’m talking 10 to 35. And that should get your attention of something you want to solve. And then you think about the compounding impacts, absolutely you think through the compounding impacts of failed payments, and it’s a lot deeper, right? It’s not just the initial subscription payment, but the organizations that are watching this, you lose a $10 subscription in January, well the user is not back in February, March, et cetera. You lose the LTV. And there’s an even broader concept that if a payment fails, my merchant ID starts to get more unhealthy, which means the banks actually make it less likely the next person who signs up can actually clear their payment. So it’s a really complicated negative feedback loop and we need to solve the problem efficiently.

So it’s a silent killer. I like to call it a black hole, a silent killer. We want to fix it. Solving it is the number one way to drive revenue growth. It tends to be about half of overall subscription churn. If you class between active churn and passive churn, about 50% of that passive or involuntary churn, it tends to be about 10% of your top line subscription revenue across the industry. It’s 440 billion. And I say solving it is the number one way to drive revenue growth. And the reason I say that is because again, if you’re a hundred million dollars business and you have $10 million in failed payments and you can get back five to 7 million bucks, that’s five to 7% of top line a RR. And if you run a product growth team and your product AB growth team runs an AB test and it delivers 0.5% to top line a RR, you guys are probably running the best experiment you’ve run all year. And now I’m saying if you focus on this, you won’t get 0.5% to top line ar. You’ll get 5% to top line ar, you’ll get 10 x what the number one AB test you run this year is. And that’s the opportunity that we’re talking about and it’s a verifiable opportunity because we have a mantra of showing, not telling. Well, you can just look in your own data and find this.
I’m seeing a question around is it B2C solely or does it include B2B? And there’s really three elements. There’s three businesses where this applies. The answer to your question is it does include B2B. So we really are talking B2C digital subscriptions. Think your Netflix is your skill shares, your coursera’s. Then we’re talking B2B self-serve businesses. So we’re talking your notions, your Figma, your Zooms, your Dropboxes, right? There are card payments that are being made self-serve. There’s not a salesperson calling your signing up for that service that applies to B2B businesses, your dial pads, et cetera. And then it’s your direct to consumer boxes, right? So your bark boxes, your farmer’s, dogs, et cetera, any of these businesses, they’re going to have this problem and payments get declined due to communication issues across a bunch of layers. Now, you might’ve thought payments is a very uninteresting topic.

I’ll be honest, I thought payments was going to be the driest topic in the world when I got into this 10 years ago. But it is increasingly complex. And when you look at the complexity, you see why it’s hard to solve because there’s the billing platform. So you guys are selling a subscription product. You may sell the New York Times, you might be selling BarkBox, all of you use a different billing platform to manage the state of the subscription. You might manage it in-House, you might use recurly, you might use Shopify, you might use Order Groove. There’s a million out there. Now all of you are also using different payment processors. You might use Stripe, you might use Braintree, you might use add, and you might use payment tech, you might use any variety. Next one is the card networks, visa, MasterCard, there’s Discover. If you do work in Japan, there’s JCB, then there’s the issuing banks, there’s a thousand issuing banks in the us, there’s PNC or there’s Chase or there’s Wells.
But for your users, there’s also Ban Columbia, Banamex, state Bank of India. And I’m just going to tell you, they all have their own rules. They all have their own authentication requirements. And you need to understand for each set of users who’s trying to pay you for each of these platforms, something could go wrong that you have to be aware of. I see a question that says do the metrics assume no credit card retries are updater when we’re talking through a hundred million dollars business with $10 million in declines, we’re talking through post-process. So every business has some process in place to do some card updater or to do some retry. And I’m talking through at the end of that funnel, there’s still a 10% opportunity. That’s a good question and maybe a little bit more advanced question for some of the folks in the room.

But to answer your question, Maria, very directly actually three ways in which you can solve a failed payment. One is an active user can call up the phone or go into a web portal and update their information. The second way is this notion of a card updater, which many psps implement. And when I say PSP, I mean a payment service provider that when a card fails, they will ping the issuing bank immediately and say, is there an updated card on file? The third way to recover failed payments is through algorithmic retries of those failed payments, figuring out the right time to run it, figuring out what attributes to pass along with it. And it’s the third way that I’m really lasered in on. So any metrics that you’re going to see in this presentation that talk about here’s the opportunity and here’s how much we can give back are focused almost entirely on that algorithmic retries piece because that’s what I want businesses to focus on is you can get uplift from figuring out and solving this problem for your users, not making them call someone up, not making them take an action.
And you can see how this can be challenging as we get through and move towards the tear down, you’re kind of hearing a couple different things. Why is it so hard to solve? Because you heard the reductive example upfront and you’re like, well, can I just run everything at 12 o’clock noon, right? That’s what the VCs asked me. They’re like, what’s the 80 20 rule? Isn’t this easy, right? It’s not easy. And the reason why it’s not easy is because few different things. Number one, when you run a card transaction, there’s 128 different data points on that transaction. Almost every business, in fact, I’d be shocked if there’s a business in this room that’s optimizing those 28 data points, you’re all sending it the same way and I’m going to get into a tear down that shows you that in a second. But you actually need to be sending it different ways.

And that’s the important part to take away. There’s also 2000 unique error codes, lots of different reasons why payments fail and you got to solve that. There’s the timing and then there’s changing regulations. So if you’re on different markets, for example, if you’re in the EU and the EU passes legislation that says PSD two and it all needs to be done by 2023, and now you have to institute 3D secure as a protocol for only EU users, but you don’t have a payments team and you’re focused on delivering a great product experience, really hard to keep up with that to be honest. So as we start to get into a little bit more of the tear down, what’s the right data to send to the card network meaning visa or to the downstream issuing bank required? And I mentioned earlier there’s 128 different data points when you’re sending a transaction and I’m just going to zero in on one of them.
One of them is the zip code. There’s 127 others, you don’t have time to go into all of them. We’re just going on one of them, right? And even within the zip code itself, you wouldn’t think this is complex, but there’s crazy complexity here. So now I’m a user, I’m typing in a card number onto a checkout form on the client side. I typed 4 1 4 7 2 0 on the client side. I know from that six digits, which is a bin, that this is a Visa card and that it’s a Chase Visa card. This actually tells me something. It tells me that I need the zip code in a certain format, whereas for other cards, for instance, annex cards, I might not need the zip code, but here’s what your product team is doing. Your product team is not thinking through payments complexity. So your product team is like, we have a checkout form, we want more people to click submit.

I want to reduce friction, so I will get rid of the zip code. And now they’ll measure more people are coming onto the checkout form, less people are abandoning it, they’re clicking submit faster, then they’ll celebrate. Everyone will clap and be like, we produce 5% more submissions and a 10% reduction in checkout abandonment. Nobody’s looking at what happens after you click submit. Now after you click submit for certain types of cards and issuers, these payments will now fail because you didn’t run the zip code, but your team’s not thinking about that. And so that’s the type of decision that needs to be made at scale all the time on the fly. Now by the way, this zip code can be formatted in five digits or it can be formatted in five plus four digits. And that is important for some issuers but not others. And lastly, when I type in that card number, I gave you a Chase Visa example, but if I gave you a Barclays card, that five digit zip code that you’re set up to accept doesn’t work anymore because Barclays in the uk, we now have letters to introduce. I don’t know why they do zip codes that way over there. It’s a weird country. That’s what they do. And so if you’re only set up to accept five digit zip codes, every Barclays card and sign up will fail. And now that’s one of the hundred and 27, 28 different fields and there’s that much complexity just in that field, and that’s telling you that this is a problem that needs machine learning.
And then you guys are on this call because you’re committed to trying to solve this problem, which is great. And the first thing you’re going to do is you’re going to log into your payment service provider and you’re going to see those error codes on the left. You’re going to say, I believe payment failure is a problem. I believe I can fix it. The first step I’m going to take is to go look at the error codes, and this is a customer that works with us and this is what they see in their Braintree dashboard. Generic decline is 40% of my payment failure, insufficient funds is 28%. PSP potential fraud is 12%. And then I’ve got this long tail of other things. First of all, even those error codes, business is not going to know what to do with them. What the hell does the generic decline mean?
What does that mean? What can I do with that? What’s the difference between PSP potential fraud and issuer potential fraud? What’s the difference between a processing error and authentication error? But now I’m telling you what’s on the left is actually not what you really need to see. You need to see what’s on the right because what you’re getting from the PSP is an aggregation, it’s a rollup. And I mentioned there’s a thousand different issuing banks in the world. Let’s keep it simple. There’s a thousand banks in the world, and when your PSP is giving you that error code distribution within that generic decline bucket, they’re taking error codes from BANAMEX and BBVA bank Commer and the State Bank of India and Chase and Carp Bank care, and they’re rolling it up and they’re saying, okay, this class of error codes from across these thousand banks looks kind of like a generic decline.

Let’s throw it in that bucket. What that means is even within that bucket of generic decline, there’s some payments that are fixable and there’s some payments that are not. And you need to know the difference, which means you need to be looking at what’s on the right. But unless you have the bandwidth and the scale to do that, you’re not going to figure that out. And that’s kind of why we say there’s 2000 plus unique error codes. And again, there are some error codes that you really don’t want to rerun, and there are many error codes that you do and that’s what you need to sort out. I’m seeing a question that says, are they failing on day one of new purchase or renewal? And that’s a great question. Both sides are failing, right? You’re failing in two spectrums. Within our business, we classify that as authorization and recovery Subscription businesses want to do two things.

First thing they want to do, make sure the payment goes through in the first place. That’s an authorization problem and that’s a solvable problem as well. Second thing they want to do, once a subscription is created, they want to make sure that they can recover a subscription that has failed downstream. That’s payment recovery, and that’s a large problem as well. And that’s the problem that we’re focused on fixing. So both those sides of the coin. But for the purposes of this conversation, we’re mostly talking about the downstream recovery. I won’t belabor a lot of time on timing. We’ve spent some time there. But again, if you’re running an engineering optimal CR job and you’re running it at 12 local time, but you have international traffic, you’re going to be running payments at one in the morning, easy fix. So if you want to take the easy fix, that’s the easy fix and you’ll get a 5% improvement.
You’ll get a C minus on the test better than getting an F. But if you want the easy fix, start there. We can also talk about the day of the week. And people often use that example of, well, if you know their debit cards and paydays, then you should run those debit cards on Friday or Saturday. Kind of true. It’s kind of the C minus fix again, but we’re in a new world, right? We’re in a new world where the cash app through Sudden Bank is now offering users and customers to be able to get their money early. So now early wage access grows. So now paydays aren’t just Fridays and Saturdays, right? Paydays are now Tuesdays, paydays are now Wednesdays, paydays are now Mondays. Are you looking in your data and figuring that out? Do you know for which types of cards paydays occur on certain days, by the way, in different countries, paydays aren’t paid out the same ways in the us.

So understanding that complexity as well will help you solve this problem. And there’s the representation of how complex this is. This is one payment that’s failed and you’re looking at the time that it failed. There are questions around, are there differences around different attributes? And the answer is all of these attributes have an impact on the strategy that you should run to recover a failed payment. The time at which it fails the amount, if it’s a higher amount, it tends to be something that’s more difficult to clear. It’s a lower amount, tends to be something where there are standards that are lower to clear decline code. We just spent a bunch of time on that. The card scheme itself, whether it’s Visa, MasterCard, discover, et cetera, defunding source, is it debit credit or prepaid? The region, is it North America, Latin America, me or apac, the actual country?
And then all the way down to the issuer and all these things matter. And I see a question that basically says you cannot change the debit the date that the consumer agreed to. So our response on that is that’s absolutely true, right? So a consumer signs up on May 8th. The case that we’re talking about is now we’ve signed up to pay $9 on June 8th. What we’re talking about is the case when that payment fails on June 8th. Now when the payment fails on June 8th, there’s a recovery schedule or a dunning process that we put into place. So we’re not changing the date at which we run the first payment, we’re trying to get as many payments as possible cleared on that first attempt on June 8th, which is when we agreed to pay. But when that fails, the question becomes what do you do next? And that’s where strategy comes into play, and that’s what we’re talking through is the strategy of what do I do? Do I just cancel the user and give up or do I make an attempt for the legitimate users who are trying to pay you to fix that issue for them without asking them to go fix it themselves? And I would believe that everyone in the room has had that experience, right? I get card declines all the time when I’m legitimately trying to pay for a service. That’s what we’re talking through solving.

Alright, so coming up on this, we’re kind of talking through how do we drive that 5% a RR growth? And I want to make sure that we address that. Just understand that churn of a symptom of a larger problem, which is poor payment health. Your payment health is critical to the success of your organization. Your payment health is the entire picture. And you’ll want to understand what do I mean when I say payment health? And I’m talking about at least three things that you can take back Actionably. First thing is recurring revenue failures, which is what we’ve been talking about. You have a subscription payment that is owed and that payment fails. Second thing, refund rate, chargeback rate. We want to minimize those as much as possible. Third thing is the overall transaction authorization rate and mid score. And so that is what we are trying to talk through if we’re into an evaluation.
Here are the things that I would be thinking through. First thing as I walk away from this conversation is let me go to my business and let me just look at how many failed payments there are in the last 12 months. Next thing I want to look at, which comes to the question earlier of the difference between top of funnel and post funnel. Second thing I want to look at how many transactions are failing to even make it into the funnel? Third thing that I want to look at is how much revenue did I lose due to payment failure? Then I want to look at the efficacy of my current strategy, and then I want to look at the impact of payment failures on my LTV and the impact on payment health and total revenue loss. I see a question that says if I reprocess on a different date, it can result in more chargebacks.
And this is where actually is very relevant to the payment health conversation. If you have a payment that fails. What is happening in the market today is we’re running what’s called a woodpecker. So if the payment fails today on May 8th, what folks in the industry are doing is rerunning the payment on May 12th and then they’re rewriting the payment on May 16th, then they’re rerunning the payment on May 20th, then they’re rerunning the payment on May 24th. That’s what’s causing your chargebacks because you are not applying any nuance to a failed payment. Every failed payment that fails, you’re not looking at the data, you’re just trying to rerun it and get your money. Not a smart approach. What you need to do is bring nuance. You need to look at the reason why the payment failed. You need to ask, is this a legitimate user who’s trying to pay me? And that’s that 70% of cases. And for that 70% of cases, you need to apply a strategy, meaning the timing and the attributes that we pass. And if you do those things, you’ll recover more money and you will lower your refund and chargeback rates, that’s what we see in the data. If you don’t do those things and you try and slam the payment retries, you’ll recover less money. Your customers will be more angry and you’ll have higher refunds and chargebacks. That’s exactly what we’re talking about when we talk through payments help.

So some example, insights and what they could mean when you look into your payments health. Number one, look at your transaction authorization rate or your tar. If you have a low transaction authorization rate, that means that you are over retrying. You might be running too many card retries in particular hard error codes, things that can’t be recovered, you want to fix that. Second thing might be the failure rate of the prepaid cards. If there’s a large amount you might consider not accepting them. Well think about definitionally what a prepaid card is. If I sign up for a prepaid card for a subscription, it definitionally will become at some point a failed payment because it will run out of money. And in particular, we can look at non reloadable cards and say, Hey, we shouldn’t accept those for subscription signups. You can also look at your invoice acceptance rate outside of the US or outside of North America, and that’s probably a timing problem. And then you can look at where you’re running your retries. Are you batching them within Aron job or a window that’s likely to be flagged as fraud? So these are tangible areas that you can look at. Kathy, I want to make sure we’re doing okay on time. How are we doing?

Kathy Sexton:
We are doing fine on time. I think let’s keep going. I love all the questions that are coming in from everybody, so let’s just keep going. I think that’s great.

Vijay Manon:
Absolutely. And keep the questions coming. So ultimately, I’ve tried to give you some of the nuts and bolts because that’s the world that I live. I built the systems to go solve this at Microsoft, at Dropbox, et cetera. And I’m building systems at butter at scale to solve it for the market. And I could talk to you guys for three hours on the vagaries of why certain payments go through and why certain payments don’t go through. But the bottom line is all the nuance aside, it’s just the numbers that matter are the numbers there or not? That’s where some of these questions are too. It’s like the numbers have to be there, meaning the revenue numbers, the refund numbers, the chargeback numbers, the transaction authorization rate numbers. This is an example from a customer that we work with without butter versus with butter. And you can go see the metrics and how they impact.
And so you come in and you see the intentional or active churn, which we should never be trying to fix. Somebody who doesn’t want your product, let ’em leave, make a better product. Don’t try and force people into staying, right? So without butter versus with butter, the intentional churn rates are going to be the same. But what we can do with butter is bring the accidental churn rate down. Why can’t we bring three to zero? Because certain payments or fraud, they should not be retried. We should give up on them. We shouldn’t be bad stewards, but 70% of accidental churn is fixable and you shouldn’t put the onus on your customer to fix it. You should do it for them. So we can bring that accidental churn rate down in half. And if we can bring the accidental churn rate down in half for a business with an average churn of 7%, I see churn for subscription businesses on a monthly rate, somewhere between five and 10 as usual.

So we call it seven. If we can bring it to five and a half, then here’s what we can do for customer. LTV. Gross margins stay the same, but the LTV to CAC at $25 now just went up two x. And so now I work with a lot of, I built a startup basically during a pandemic and a global macroeconomic downturn. Nobody wants to pay for enterprise vendors, but they’re paying for butter. Why are they paying for butter? Because their businesses are contracting because of the global macroeconomic downturn. The CMO of the business is like, wow, now I can’t spend because my business is contracting from a top line perspective, but when they put butter into play and we boost that LTV to CAC by two x, well now I can spend, even though my top of the funnel is contracting because I fixed the problem downstream, I can now spend on new acquisition channels or I can spend on an existing channels.
That’s exciting. And you can see the same thing at the bottom where if we’re solving this problem as well from a churn perspective, we’re going to impact ARPU and very minor reductions in monthly churn lead to very major impacts on the business. $15 million impact on $170 million business is 10% of top line a RR. And we’re finding on average by solving this problem, it’s a minimum of 5% to your top line. So those are the things to think about. If you’re solving this problem well, you’ll reduce your churn rates, you’ll increase your LTV to cac, meaning that flows through to your contribution margin, which means it flows through to your marketing spend, which means you can solve the top of funnel problem that you’re dealing with. And then you increase not just top line, but bottom line. These are subscriptions. They’re like a gym membership. If you’re a public company, this is impacting your ebitda. And so we can have that impact on your ebitda.

And we use a lot of butter puns. So I apologize in advance for the proof is in the bread and butter. We like to say it’s your bread. Make sure it’s butter. There’s a million more. I could also spend an hour on that, but I’ll stop. But again, like I said, my business is, it’s a very existential business. I mean, we can have a debate over what works or doesn’t work and I’d be happy to do it, but you also just have to see the numbers. And at the end of the day, it’s a sink or swim business. Our business succeeds if our customers make $10 for every dollar we’re making. And that’s what our customers are saying and seeing. And so we talked to the VP of curology’s, been proud butter user for two years. Through our partnership with butter, we’ve seen a significant lift in the customers.

We recover with each customer translating to two x more revenue. Now, that second part is not something I’ve talked about. So now we’ve talked about this concept of maybe you’re recovering $10 from every $20 that goes into failed payments and butter will bring that to $18. But the other part of recovering failed payments efficiently is that the payments you recover will have a higher LTV because there will be fewer refunds and there will be fewer chargebacks. I think I got a question around the refunds in the chargebacks. The reason you would have more recoveries with a higher LTV, the reason you would have that is because you are being smart and nuanced about payment recovery. I see some businesses that do nothing about this problem, and I see some businesses that tilt too hard at this problem. They’re tilting at windmills, they’re trying to run everything, and when they run everything, those users aren’t happy to have that subscription recovered and they churn.
So you want to recover efficiently to get the extra LTV. And of course, the head of finance at Hair Story says, we’re keeping more subscriptions active, we’re increasing our revenue every single month relative to what it would’ve been had we stuck with recharges solution. So again, it’s a very existential problem. Our mantra is show and don’t tell ironic. I know I’ve been telling for quite a bit. We want to show you your own data. And to that point, and this has really kind of coming up to the end, we do all of this for you and happy to do all of it for you for free. And that mantra show, don’t tell and we have to deliver the results, put the Otis on us to it. We have something that’s called a payment health analysis. And so whenever I talk to anyone about butter, we have a long conversation around payments to the extent that they want to have the conversation around the payments or the technology.
And I think it’s an exciting conversation because it’s an interesting space, but I’m also just like, I’m not going to guess how much we can give back to you. I want to tell you it based on what I see from the historicals. And that’s how we want to do it is I mentioned upfront, we grabbed Read only API keys into Stripe or Braintree. Your Aon. And we will tell you your business had $25 million in lost revenue due to failed payments in the last year alone. It’s not a qualitative statement. We just pull that directly from Braintree. Here’s exactly how much we think we can give back to you. And for every business it’s different because you guys have some process in place right now. You have some mix of skews, geographies, amounts, network distribution. We want to look at the error code distribution and we want to tell you exactly how much we can give back to you.

And typically, of course, you’d go to a consultant, they charge you 20,000, 30,000 bucks, you’d get some insights and you’d move on with your lives. But what we have the software to go do this for you at scale, and our software basically is a web hook listening to all the failed payments that come in making a decision of whether they’re fixable and in that 70% bucket or whether they’re unad addressable and should not be retried. By the way, if we take that out of that bucket, you stop paying the payment service providers for swipes. If you’re running all the a hundred percent of transactions that fail, you’re paying too much. You don’t need to pay them for those swipes. Let’s get rid of that. And then for the 70% that’s fixable, let’s go make more money off of it. And then we give a commitment and we basically say, Hey, we’re committing to giving you $5 million back and we’ll give you 30 days to go see it.
And if we’re committed to give you $5 million back then in the first 30 days, you don’t want to see 500,000 bucks. That’s what we’ll save. We have to show you that. We have to show you that to have a relationship. So it’s an existential business and it’s easy to run this analysis. We can find this data in five minutes and show you what’s on the table. So we want to build an ROI case for you. And by the way, if there’s no money there, we don’t want to work together. This is, it’s something for your business where you want to prioritize it relative to what you could be doing with your time otherwise. So those are the things that I think I really wanted to cover and get through, and

Kathy Sexton:
That’s brilliant. Can you just, why don’t we just go to the previous slide. Well, actually why don’t we just put it on video or this here For anybody who wants to learn more about this, I have an email in the chat for [email protected], and there’s also a URL: @butterpayments.com payment health churn analysis. There you go. So if you want to go and you want to get started on that, that’s fine. We do actually have a few questions. Angela wants to know what strategy tends to move the needle the most in terms of mitigating accidental churn? Said differently, where do we start that’s going to get the biggest bang for their buck?

Vijay Manon:
Yeah. Yeah. So I would say there’s four levels. Angela and 99% of subscription businesses are on level one. And so level one or level one or two, right on level one payment fails. You don’t do a whole lot about it. You might send out some emails, but you kind of have the customer on their own to go figure it out. Level two is you’ve got some sort of aggressive retry strategy that I would call the woodpecker, which is, again, it’s May 8th at 10:42 AM where I am here in Berkeley, California, right? And the aggressive retry will say on May 9th at 10 42, I run it again on May 10th at 10 42. I run it again on May 11th, 1142. I run it again, that’s level two. Level three is what I would call dynamic decline handling, which means I look at the error codes that we talked about earlier and we saw, for instance, insufficient funds.

And when I see insufficient funds, I change the schedule, I see insufficient funds, and rather than retrying every day, I now just retry every Friday, and then I see processor error. And rather than waiting till Friday to run the payment, I actually run the payment in six hours. If this is a processor error, we’re not going to be down for seven days. We’re going to be down for six hours. So I go rerun it in six hours. Dynamic decline handling is kind of where I am not really using machine learning, but I’m being smart enough to say, okay, there’s certain obvious things that I should be doing differently for obvious classes of error codes. If you look into this graph, I know it’s hard to see, but you see insufficient funds, invalid card issuer, invalid transaction, do not honor dynamic decline handling would say, I look at those and I can at least put some smart business logic in place.
So if I would say, what strategy moves the needle and is achievable, I’ll add the achievable part. That would probably be where I’d go. I try to get myself into the level three bucket of at least there’s, I look at my error codes, my high level error codes, and I at least try and do some smart things. I try and change the strategy depending on the error code. That would probably be what I’d encourage is I think I can probably do this in-House in many organizations, and you’ll get a little bit better, you’ll get 10% better, which is good. Now to really move the needle to get the 5% impact, truly, you have to use a machine learning. And if you have a smart capable team that has a machine learning cloud platform and is capable of ingesting thousands of payment failures a day and making those decisions, then go for it.

But that’s a service that we’ve built out over three to four years. And the other thing that I would say is it’s not just the technical expertise to build that, it’s domain expertise because the rules are always changing. There’s no unified payments organization. So the rules that work today don’t work next year, right? Because some things for some banks are hard error codes now and they might become soft later. And so those are always changed. So unless you’re actively monitoring it with the payments team, with the machine learning cloud platform, you’ll never fully solve the problem. And I think for most organizations, you want to have your product people focus on the things that you’re doing best. Like if you’re Roman, you want your people focusing on prescription drugs. If you’re Netflix, you want your people focusing on movie quality. If you’re Spotify, you want your people focused on algorithms. I ran backend teams for search recommendations and payments in my past lives. Everyone wants to outsource their payments. None of the smart PMs come out of college and say, I want to own payments. That’s my thing. People want to outsource payments and they want to keep search and recommendations in house. That is part of their court competitive advantage. So that’s where I would suggest is if you really want to take a quick stab at it, try to do some of the dynamic decline handling. And if you want to get the a plus on the test, it’s an ML cloud platform, and that’s what we provide.

Kathy Sexton:
Don’t ever set it and forget it when it comes to payments. Absolutely. So I have one last question here, Vijay, and then you can leave us all with one final thought. But Tony has a question, and I think it’s really relevant for all of us, is there a minimum revenue that A, we shouldn’t be worried about this, so we need to start worrying about it. And then specifically for better payments, is there a minimum revenue that work with? So a two part question, and thank you, Tony for the question.

Vijay Manon:
For sure. The vast majority of our customers are over a hundred million dollars in recurring revenue. We will work with customers under that threshold, but the vast majority of our customers are at that threshold, but we’ll work down to 10 million or whatever the case may be. Now, to answer Tony’s question direct, and that’s a fantastic question. This is such an interesting part of the business, it’s one of the few businesses where the percentage impact on the business does not vary by size of business. So I can tell you a 10 million business, I expect to find at least a million dollars in failed payments, a hundred million dollars business. I expect to find at least $10 million in failed payments and a billion dollar business. I expect to find at least a hundred million dollars in failed payments. We find that trend across the board. So no matter the size or scale of your business, the percentage impact on top line is almost always the same. And again, within five minutes, we can verify that statement or not. It doesn’t need to be a qualitative thing, and that is true 95% of the time. So hopefully that answers the question Well, and the reason for that is because, again, there’s an engineering optimal way to batch run retries. Everyone does it whether you’re 10 million or a billion.

Kathy Sexton:
Right? Right. Well, any final thoughts for us, vj?

Vijay Manon:
I appreciate the participation from everyone, and I hope that we uncovered something that you guys may have been aware of as a cost of doing business, but may not have known how deep seated the problem is. This whole thing for me has been a course of discovery over the course of my career, and I learn new things about payments every day. I’ve learned them over the last 10 to 12 years. I would just encourage everyone that wants to go deeper on this topic to come have a conversation with us and we’ll talk about the facts and see if there’s an opportunity to do this for your business. I appreciate the time to come on, and I hope to see folks in New York and October as well.

Kathy Sexton:
Absolutely, Vijay, thank you so, so much. Thank you all for coming and participating today. We know your time is valuable, and I look forward to seeing you VJ in New York, October 22nd and 23rd for a subscription show. And I hope all of you will be joining us as well. Thank you and have a wonderful day.