AI & subscriber churn: Mastering recurring payments

First of all, it's the thank you. I wanna we'll start we're gonna do some quick introductions here. We're gonna we've talked about a couple of items that we thought would be relevant for everybody who's attending and, the gentleman to my left who will introduce themselves in a second will share probably better insight than I will on that. And then we wanna, open it up any and all questions or good questions here. So this is I think the the goal from my standpoint is you hopefully walk away meeting somebody new and interesting from this, taking at least one new idea source of inspiration that you might go apply in your business.

And that's really that's really why we're doing this. So I'm Joe Rohrlich. I'm CEO of Recurly.

I'm new into this role. So I've been here for all of four months. I figured it all out.

I actually am a I am a Recurly customer in the past, so I am, quite familiar with, with business and with the product. It was a fan. When I had the opportunity to come join, it was one that I was willing to kind of stop what else I was doing to be a part of. So excited to be a part of this conversation. Sebastian, I'm gonna let's just do quick introductions, maybe tell us a little bit about, maybe your role of the business, And I'm gonna throw like one so one additional question. We were talking about our own subscriptions earlier.

The way this game works is you can only, you have to go back home and cancel all subscriptions, but one, what are you keeping? You only get to keep once.

Water that's a very good question Apple one?

Is it the is it called the Apple one?

I don't even know how it's called.

They they charge twenty five Euros every month, but I'm using all the services.

I keep this one. And water. I mean, electricity, let's say I think you you you refer to, like, the it was first world problem kind of subscriptions.

Right. So tell us what, tell us about Blinkist. What do you do there? What is Blinkist?

Yes. So, I'm Sebastian from Blinkist.

I joined Blinkist in twenty twelve, a week too late for the notary. So I'm not a co founder, I'm co starter.

After twelve years nobody cares about that anymore.

But, we took the world of books and digested them into, really nice digestible chunks of fifteen minutes in audio and text and were, laughed out loud when we said twenty twelve that we go into a subscription model, because everyone said it's never gonna work.

Turns out it worked. It was pretty nice. And I don't know since when, but we've been Recurly customers for a long time. And, the reason why we went for Recurly is because you had PayPal.

And I read the PayPal documentation and I'm the CTO. And I it was three hundred pages long XML and I said, I don't do that. And I Googled for a solution. I found Recolet.

And since then, we are Recolet customers.

Today, I'm still the CTO of Blinkist, but we got acquired last year in May. And now I'm reporting to a VP of engineering at Go one. Who knows Blinkist? Who knows Go one? Simon knows, because he's account manager. Yeah, so we're now part of a bigger organization, around eight fifty people. Global.

Very interesting. I can talk about it more later. In the social time.

Exactly.

How about you?

I'm actually gonna say, my car because I think that's, that's an interesting I know it's not, I guess it's First World? I guess, Last World cars.

But this the new subscription models for cars now I think kind of interesting where you can then bundle in absolutely every facet of owning a car into into a subscription. So that's and I I did that about twelve months ago, and it's been it's been a weight off. It's sort of, takes a lot of pain out of owning a car. So, yeah, I'm Andy Wigan.

I'm the chief product officer for GoCardless.

We are a global bank payments business, so we we look after and process payments for it's verging on about a hundred thousand businesses now across something like thirty five forty countries.

I've been in the role for a couple of years, I joined as a VP Product, and yeah I spend most of my days sort of building building cool shit really and I've spent I was gonna say nearly twenty years in and around payments and subscriptions. So I, I was at Spotify for nearly ten years. I don't know if Spotify are here tonight.

Will my full length? Hello. We haven't met. We will, I'm sure.

So I spent a lot of time there building, you know, their their sort of payments, infrastructure and figuring out how to sort of grow, grow from, you know, two or three markets at inception to fifty, one hundred plus markets with all sorts of different payment methods and subscription models and pricing models and things so I've kind of got a little bit of you know vendor experience from from from my current role but also customer merchant experience from from from quite a bit of time there.

Okay. Well done.

I gotta give mine to right. So I'm, I lived in London until twenty nineteen, but I'm very American and so I'm a huge baseball fan. It's the boring sport that we're, you know obsessed with.

And so that's my subscription to my local, my local sports club, my local baseball team has been right by my side all over the world where I've lived and, I can't I can't ever imagine. It's the first thing I check-in the morning. Can't ever imagine getting rid of that. You mentioned something about automotive which is interesting.

Like when we step back and look at what's happening in subscriptions, we see that we're still at the fairly early stages of this sort of as an industry. And we're gonna talk about that right now, like what are some of the practices that we know work when it comes to things like retention insured and what do we not know today, which is a lot still. One of the interesting things is that we see industries that are more mature in this. Like a number of you are in, various forms of media here.

That's obviously a consumption model that has been subscription for a longer period of time.

But we also have we see a lot of growth in automotive. We see growth in airlines. We see growth in more physical goods. So there's a lot more of our total economy that is maybe in the experimentation phase but moving towards subscription. So I think it's an interesting kind of an interesting time to be exploring this personally. That's that's again why I wanted to be a part of it. So we're gonna kinda start with a little bit of the maybe the things that we know that we're working on, then we're gonna try to end with some of our what are sort of our forward looking ideas and that will include everybody's favorite AI topic, of course.

So, Sebastian, what, obviously, we're gonna talk about the the nuts and bolts of how do you improve, retention or reduce churn, which I think was the title of this.

But before we get to that, there's sort of the subscriber life cycle and experience like where you've been at this for, thirteen, fourteen years now. Like what are what are some of the biggest insights that maybe have been surprising to you or have been key to success when it comes to kind of optimizing the subscriber experience and life cycle. Obviously, it's got the end of the day for your CFO. It's got to relate back to Revit UI suit in some form.

Yeah. There were a few, interesting moments that, we had in the past. The first was we we started always being fair, transparent, and how we want to be treated. We also want to treat our customers. So, when we introduced subscriptions in twenty thirteen, I guess, we started with a thirty day trial. We thought everybody should be able to test the product for a month before they actually commit to a year. Felt really, really good.

But we didn't make much revenue. And then, at, I don't know, ten PM after four beers, my colleague and I have said, like, you know what? Let's make it seven days.

And suddenly, we made more revenue. It said, like, And literally, we got more revenue. And said, like, we make another test. Let's move it to three days.

We made more revenue. We made a free trial in only one day, and we made even more revenue. And we learned, people wanna try the product, but they're actually interested in committing rather early. So we found a speed spot of, three days, after which people are they feel like they had enough time to test the product, and, then they wanted to commit properly and they wanted to pay for the service.

There was no pay trial back in the days. And, after a while, Apple Pay, Google Pay, so we are native in the mobile app stores, where you have to use those payment vendors.

They offered then a trial period, a paid trial. But we learned rather quickly that people don't know what's going on. So, you opted in for a trial, but you didn't know when it's actually charged, what's happening. Was all pretty new.

We iterated a lot on the subscription screen, like the checkout screen basically.

The best winner, the winner that turned out was like a timeline view where people just transparently understood what is happening when.

So day one, this is happening. Day three, you get an email. Day five, you get this email. And so, we iterated on that and we learned that transparency is just super important that people trust in this entire endeavor.

And, that's why we are focusing a lot on on transparent communication and setting clear expectations for the customers of what they can expect by when.

Apple messed that up because they introduced their own notification system that was, notifying customers too late.

Then we get a lot of feedbacks. So this is really important for us that we tell people what is happening when.

Even we were introducing emails to announce that there's an upcoming renewal. And today, this all seems normal. But back in the days, this was all brand new and nobody knew exactly how to do that. So we're investing a lot on this communication part.

In between we're working a lot on activation because in the end, you wanna use a product that you're actually using. So there's this gym moment. You know? You see the onboarding. You feel really good. You read that book finally that you had in your shelf for like five years and never touched it, and you really feel good and you subscribe for a year or at least you start a trial. But after a while you stop using the product and we learned the retention curve is not looking good, so we started investing in a seven day activation mode so that we can activate people during the first seven days to product and feature education, all those things.

We started doing weekly newsletters, push notifications.

We offered a free plan, like, where you have basic, access with a free daily, all these kind of things, because we want to make sure people use the product.

Not just pay for it, but actually use it. And that got us, I think today, we are eight percent over our planned Renewal Revenue.

Then on top, stuff that does, which is a more technical thing, is like this, the stunning, this figure out, oh, there was something wrong with your credit card. All these mechanisms of involuntary churn, improved a lot, especially that credit card expired problem. I think that was a big bummer in the past. It got a lot better.

And recently, we, we learned that, and I said that when when we spoke, when when when we met each other the first time, that we're currently investing a lot in the moment when people decide to turn off the auto renewal.

Then you have it's it's my guess. It's currently a bet I have with the CRM team, we have twenty four or forty eight hours until people forget about us. So you decide I turn it off and then you will stay in the mental space of a user for like a day or two, I assume, before you're just like gone. So use that moment and change people's opinion, like, offer them a different title, something you haven't done before.

Just tell them, hey. You deactivated it. It's great to but, You know? Do you wanna have an offer, like, an alternative offer?

Was it a price thing? You can hook into this and make a team focused on improving that number. And we already see some early results of ten percent uplift in this auto renewal status thing. Very early, unconfirmed, but, it just makes sense to me.

So this entire cycle is super important to us. We think about fairness, transparency and product centrism because people will not buy your subscription, they will buy a product.

I love it. So I said, yeah, I mean I'm kind of saying what you just did at the end there, but I'm hearing kind of transparency over communicating to your in educating your subscribers, your consumers. And the other thing which I really was sort of throughout that is a lot of experimentation.

Feels like, right? So the ability to iterate fairly quickly, go test those ideas, those hypotheses, sounds like that's been key?

Yeah. Hundred percent. We've been doing AB testing, multivariate testing, the subscription screens, the messaging, everything, for the past, I don't know, eight years.

Cool. So then, maybe on like so we sort of said, okay, there's subscription life cycle as sort of the core product in some ways, especially in a digital product, but I think also in a physical one. And then there's the how we monetize how we monetize that product.

And so Andy, kind of curious if you could maybe lay out for us both as a subscription person and a payments person, what are those fundamental building blocks when it comes to successful payments and subscriptions that we should be thinking about?

A father and a husband.

Of course. Yes.

I don't just define myself through pay payments and subscriptions.

Yeah. I so just going back to what, to what Sebastian was saying, because I I think there's there's just something in that, which is about the the ambiguity of your product as well. You know, if you've if it's if it's water, and, you know, you're you're running a water subscription business, then I think you probably have to do a little bit less on the, experience, dare I say it, and, what we what we term sort of involuntary churn side of things. And it's a large year round, you know, pricing and and sort of much more hard, sort of competitive competitive metrics. Whereas something, which is a new concept, which is largely sort of, you know, unknown some, you know, twelve years ago for for for Blinkist, you know, you do have to do a huge amount of testing.

So on on payments, or yeah. Before I go into payments, I think I don't know if it's the same as Spotify now, but when when I was there, it was about a fifty fifty mix. So you had, you something like two and a half, three percent monthly churn for of subscribers net, and, half of that was was what we call active, which is people canceling, and half of that was was passive, which is the, you know, voluntary, involuntary. Passive was all was all payments. So I'm not sure if if anyone sort of sees those those same numbers, but that's that's the experience I had.

The the passive churn, the the payment churn problem is, I I would I would say, a bit more constrained.

If you sort of rewind, you know, fifteen years or so ago, there weren't that many businesses, that many payment, companies offering sort of ways to solve this problem, and it's become, you know, almost sort of table stakes. So I remember when Adyen, as a payments business, first first launched their sort of intelligent retry product, I think it's revenue accelerate or something like that. And it was sort of revolutionary. You know, they're running, five percent of all transactions in in in dark mode through, through a machine learning model and could prove essentially that by doing that, you know, you are saving, you know, x X thousand pounds of revenue every month. Fast forward, you know, a number of years, and it's kind of it's sort of table stakes now. Most PSPs are offering that. So I think that's the largest old problem in in cards.

For bank payments, which is what we do, it's slightly less mature. So we, and I'm trying not to do the hard sell here, but we we launched an equivalent of that, but for bank payments, you're only using our bank payment, network, our millions of bank accounts that we've sort of seen before, to do predictive modeling on on when to do retries and when to actually sort of, try and try and save a payment. And then I think the second one, and this is kind of the obvious one, is really about payment mix. Right?

So when you're accepting payments, unless you're sort of tied to an app store, and you can't have any options, there are payment methods that obviously naturally lend themselves to lower passive churn or lower, in bond churn than others. Bank payments being one of them because there's no card in between. So you really only see, you know, lack of funds and things like that. So those are the two main areas I would say, when I think about churn and payments, but I think there's there's you do need to think about the whole lifecycle.

So one of the things that I was always very hot on when we were doing this at Spotify was sort of knowing your data across the entire lifecycle, You know, making sure you really understand at a super granular level everything that's happening both in terms of usage and activity of the customer when they're on platform, when they're using their products, and consumption habits, all the way through to, you know, super granular analysis on exactly why something's failing, when it's failing, and and and everything that sort of play plays into that. So I think that was that's that's kind of point one.

Know your data, and then, you know, when it comes to payments, it's really about payment mix and, having a means, and we can go into the details of it, but a means of doing retries in an intelligent way.

Yeah. I think, it's the it may be the single biggest thing that we get asked even, you know, met with looking around here. Maybe we hear not here. I met with a couple of folks, earlier today, and this is like it sounds it like you said, it's table stakes, but it is something that everybody I think is asking, trying to understand. Are we getting as much as we can out of kind of squeezing that involuntary churn down.

I know it's an area you all have, again, experimented in, a lot over time. Like what how do you what have you what have you done that you found works? What's kind of what's, you know, where do you see more opportunity on the involuntary side?

On the involuntary side, the biggest driver we had that made was a big success was actually introducing PayPal.

Because PayPal has taken care of a lot of, of a lot of those things that lead to involuntary churn, like, you know, keep your credit card up to date, have a different payment method. Like this wallet approach is really useful. So we see much higher renewal rates from PayPal just because they organize all this for us.

We recently got on the Tokenization thingy, I call it. Like instead of storing the credit card, actually storing a token and then the credit card vendors take care of whatever expiry date and and and la de da de da.

What I also did is I always talked to Rick Crowley. I literally said, Hannah, that's your problem. It's not, it's not my job. You know, I need to make sure book summaries get delivered across the planet really fast so that when you hit play it actually starts playing.

That's why I actually pay Recurly to fix that problem for me, because I need to work on the voluntary turn. That's something I have under control. I don't have involuntary turn under control, so do whatever you can. AI model, turn it on.

I wanna plot see how it works. We had many fun calls where she was just telling me what's what's coming up on the road map and said, let's let's go.

We have not done much ourselves, to be frank.

Yeah, I mean we we saw in twenty twenty three we saw about sixty three million subscribers across our network, almost all consumer subscribers of various services.

And on average, the duration of time that they continue to pay and subscribe after a, essentially, a recovered transaction, There was another one hundred and forty one days. So let's call it, you know, another half year of subscription value once they were able to process that. I mean, I think I think, again, it gives you the opportunity then on the voluntary side to say, okay. How do we get this right, with them?

But there is a like, there's a lot there's a lot there. At the end of the day, I think all of you know, everybody is trying to go network together this data to get more predictive around, you know, whether this is the token thing, this is what Recurly does. Everybody's trying to network here this data to get more predictive and get to a better consumer experience where we're not having to go ask you, you know, your payment failed. Can you go?

You know, can you go update this? We're anticipating that in advance and able to get ahead of it. So that's the, you know, that's that's the goal. But I do think we're all excited to talk about kind of what what we're doing and what might be next on the voluntary churn side.

There's there's sort of, like, what's in front of us and then what might be possible with AI and, you know, other capabilities. I know I'm excited about this personally. So kind of curious, like, where where you see some of the next frontiers and, like, maybe we get this wrong, but, but we'll brainstorm a bit on where are the next opportunities.

Involuntary.

Yeah. Okay. Good.

Okay. So I I think I think the there's a couple of things already happening. Right? So there's there's, you know, I spoke about Adyen before and, you know, most PSPs, processing cards, and and and now more more alternative payment methods and then vendors, you know, like like Recurly and such have got these models where they can do predictive analysis and understand when when things are gonna gonna churn. So I think that's that's gonna continue to to happen.

One of the interesting things that I would love to have done more of, when when I was on the merchant side was sort of actually kind of the crossover between voluntary churn and involuntary churn because there it really comes down to intent and if you can if you could find a way to measure someone's intent of being a customer being a valuable customer whether that's through usage, whether that's through some certain habits, you know, in terms of how they consume your product, then you can really find that sweet spot and figure out, you know, when actually it's a better idea to offer a grace period, for example.

So if you if you see that someone's using the product loads, like, in binary terms, extreme terms, someone's using the product loads but they have a payment failure, don't be hounding them to, you know, try and recover that payment. Give them a grace period and find subtle and kind of more and gentler and more invisible ways to sort of just recover that payment. And lean on networks. You know, if there were one single network that you could just tap into that just gave you access to all all of the data to be able to predict when a payment is gonna fail, then that would be brilliant.

Unfortunately, they're not. You know, the payment world payments world is very fragmented. So tap into the right networks depending on your payment mix.

But But I think it's that it's that predictive kind of understanding the intent of the customer and knowing what to do, about that. I'm thinking, you know, in terms of the the future of, the future of sort of solving payment churn is really about the networks being bigger. You know, so for us, for example, we're on a sort of mission to build the world's bank payment network. We've got you know, substantial amount of data in our platform today.

But if we double our business and we double our customer base and we double the number of transactions that we're that we we we've been we can do, then the, then the model obviously gets much more powerful and so our product product gets a lot more powerful. So for me, that's really I'm sort of maybe it's a bit short termist, but I'm looking at kind of how we can get those networks to be as good as they possibly can be so that really you reduce payment failure down to to almost zero unless someone's intending to fail because they've, you know, not had funds in their account for five weeks or whatever.

Awesome. So let's kind of drift into maybe the, you know, the voluntary side. The like where does where is that going next? You you I think did start to touch on it in terms of more personalized pathways based on user behavior or other data. But, like, what's your like, what's your if you were building kind of, I don't know, envisioning Blinkist five years from now and the voluntary, you know, like, what might be in place from a voluntary, churn perspective that doesn't exist today?

The voluntary churn comes from like what we hear most people say it's either too expensive or they don't use it enough.

So these are the two things. Don't use it enough. Why? You can drill down there, you know, Oh, the content catalog is not enough for me.

Well, we have nearly seven thousand titles. Usually, it's enough. They just don't find what they need because machine learning algorithms aren't good enough, yi di di di di di di. So discovery is a big thing.

I think that as of today, when I say like what's the current state, what's the desired state, today I need to go to and it's the same for Recurly as it is for Google Play, as it is for iTunes, as it is for Stripe. I need to go in there, configure a Product, give it a Price, give it a Currency.

Then I configure the next Product, SKU, give it a price, give it a currency.

Then I do this for every permutation I can think of, and then I need to implement a decision tree logic like if user is from India, if user is in a big city, then show that SKU.

If user is in Austin and is a CEO, then show the other SKU.

For example, is this and and so price elasticity and saying, like, you know, at least four ninety nine. Max, nineteen ninety nine for a monthly subscription would be nice. And then we can use machines to figure out what's the best price. They can learn. They can test. They can evaluate validate. They can take all this data that we can't process as humans, and none of my engineers can put into an if then else statement because it's just way too much.

I think this is one of the things that I that could help scale the customization based on whatever we think is relevant.

Might it be Geo? Might it be whatever we know about a user? Might be the IP address? I don't know.

Whatever it's gonna be. But right now we do this by hand. And I say by hand, of course, we have software to make a decision. But this decision logic is written by humans with a logical if then else statement.

If we get this to a point where a vendor like Ricohulli can just give the best price because they know, I think that would be a really good thing. So price elasticity.

I also think that what we like, we do it at the moment by hand. We try to find the moments in time where we message customers certain offers or recommendations or whatever. There's so much power in triggering these, these communication moments at the right time.

And, and I think here, again, Recurly sits on so much information that telling us it's a good moment to send an offer to this customer. It's a good moment to remind them of the renewal.

I don't know. You know, all those things that are, with all the data that we're collecting nowadays and all the data we have, we can tailor the experience so much more to the individual user.

And by far, by the way, I have just talked about the actual subscription and I said before nobody's buying a subscription. People are buying a product. So I specifically focus here on the payment experience, so to say, because on the voluntary churn, we need to we fight against all other competitors that are out there that either now offer book summaries as well, because they saw with us, it's a viable market. So we have nowadays competitors that are good in this business as well. We have to compete versus large language model where you can just just say give me a two minute summary of the Lean Startup and you get it and you have to buy it on Blinkist.

So there are many things that we need to do on the product side, but on the revenue or like the subscriptions side, I think there are quite a few, let's say, opportunities where it is much more tailored to to your needs. So that you find it affordable and you commit. Whereas, the customer is valuable as long as they give us money.

Some will give us more, for example, a CEO, or some might give us less because they're just students. But it will give us more over time. So as long as they stay with us, great.

You seem to think they pay me to do this.

I have no idea what you earn.

Right.

Well, I look. I just we were talking about this earlier. I really I believe in this pretty strongly. I think, again, if you sort of step back, a lot of these systems including like, a lot of the technologies that exist today are built to go help kind of take something that's really complex, which is managing a subscription business.

Like it's just subscription businesses create a lot of complexity across the organization, across technology, across finance, across, you know, growth and product there. And I think that that's what a lot of these systems help to do and hopefully create speed to market as a result of that. But I do think that the sort of next generation where these, you know, platforms are going, including ourselves, is about helping to anticipate what hasn't yet happened in the subscription business and going to essentially do what you can't do in a b to b subscription business. I run a b2b subscription business.

I can have my team sit down and talk to us about our customers and you can't do that. A consumer subscription business it's just like you know it doesn't scale. And so you need this, essentially prediction and automation, and testing. And so you know an area where you're gonna see us, you know, experiment over the coming quarters is in essentially being able to push more predictive offers to subscribers.

So things like downgrades or pauses, or upgrades or education and being able to push those across you know multiple surfaces. So you know in an app store, in a web app, in a connected tv there. And I think at the end of the day, a lot of these, you know, these sort of problems don't get solved in isolation. Right?

They are by definition, they're a matter of connecting data from a combination of sources like the usage data that you see, around your product, the transactional data that GoCardless sees or that we see.

But I think it's an area where I anticipate there's gonna be a lot of innovation and, you know, I'm kind of challenging our team to think pretty broadly about where that, you know, where that applies. Because it's again, when I've done I've been I've probably done close to three hundred customer meetings in the last, ninety days. And the number one question, you know, businesses, some of which are the biggest names in subscription globally. The number one question is like, hey, how do we how do we go actually get ahead of the curve on subscriber engagement retention and it's not just gonna be payment success rate. It's gonna be how do we anticipate where those kinda where those gaps are and meet the subscriber where they are more individually, like you said.

So I think I think this is actually the perfect time for us to open up for, for questions. I love it. Thank you for thank you for breaking the ice. That's that's perfect.

Hi. It's Lisa from the Telegraph.

I think it's really interesting what you just said because we've suddenly got to a world where we realize acquisition and retention are so closely coupled now. So what we do at the front end is gonna massively pay into the retention play.

So how do we bridge the gap between the low ball offer and how that plays into retention because I sit on the bridge of that all the time I've got stakeholders who are asking me to like just get the best sweet spot for an acquisition I would touch people then well how am I doing these people because how do we bridge that gap?

So I'd be interested to hear, like, what your thoughts are and how Gen AI and all of this.

I'm not yeah. Not sure about Gen AI. I I think that I'm sort of thinking how I would think about that problem because it is it's a sort of universal problem, and it's one where there's kind of there's hard rules and there's soft rules that you would put that you'd need to put in place. Right? So for for for something on on the front end when you're acquiring customers, you are basically you're screening for the value of that customer, and then you're setting your price point accordingly if you've got flexible price point. Perhaps you don't, but you are screening for the value of the customer.

So there's there's hardware rules that you put in place, you know, you might not accept cards from a certain bin range, for example, because they're just taking advantage of the free trial and, you know, will never become a real customer and things like that. So there's kind of I can't think of many more examples to be honest, but there's there's there's hard rules there.

If you can do that well, then obviously, it will have a knock on effect on, you know, your the lifetime value of each of those kind of cohorts of the different intent levels of each customer type. So, so for the customers that you screen well upfront, you're gonna have, you know, less payment failure issues, you know, further downstream.

That's just my mental model, but, I don't know if we can go any further on the gen.

Well, I think well I don't have any answer on the gen AI thing but I'll share with you I thought we did one of these last week in New York. I thought there was a really good example of this which sort of speaks to like the need to create a cultural alignment in the company around this balance between acquisition and lifetime value. Right? And so this was we had a woman who, runs a business called fuboTV which is a online, sports streaming station.

They're one of the one of the largest in the US and, and large in France as well, a brand called Molotov. And they what they know is that in their business, they have moments that bring in big acquisition opportunities. The World Cup or, you know, big big sporting event. Right?

And so there's a tendency to basically drive a lot of, you know, acquisition dollars behind that.

At the same time, they've been able to kind of study these cohorts of customers and get to pretty good clarity that those customers are coming in for an event. They are actually not buying in yet to the value of this ongoing subscription. Doesn't mean they can't forever, but they don't. And so the sort of the business decision they sort of drove alignment around was, okay, do we do we make a different offer to this cohort of customer?

If we know that you've entered for this reason, do we intentionally downgrade your subscription actually to a lower cost after the second month? Say, we know you came in, you know, for the World Cup. The World Cup is is over. We had these other things we think you'd be interested in and we want to proactively we actually want to proactively reduce your cost moving forward.

Now Milcor has come back around and attempt to increase the value and the cost. But I think that, you know, what she described is ultimately like a lot of, I think, activity just to try to get leadership alignment in the company that, hey, we do have this balancing act to go do. And as we identify, as we have these insights, like I kind of need everybody's buy in across because yes we have different groups who have different KPIs, different goals, different bonuses and all those things but I kind of need everybody's alignment that we're gonna balance these two things. I think the thing they used successfully was just a kind of got buy in on a common customer lifetime value model.

Basically that was the thing sort of like kind of run everything through the same model. And, and so I don't again, it's maybe it's not a perfect technology solve because I think I think some of it is a cultural challenge as well.

I don't think I've ever seen them before either like companies using active downgrades just to keep you as a paying customer, you know but if you're not using it then actively downgrade because your lifetime value is still very high so it's super interesting.

Thank you for breaking the odds. What, what else?

Just off your last point there, active downgrade.

Joe Zipcar, the car using my car membership and and car. Funnily enough, I was paying a Zipcar membership annually, realized I wasn't really using it that much, called them to cancel. They said don't worry, we'll just waive the membership. When you use it you just you spend which was that active downgrade but they kept me on but they just they basically canceled my annual membership fee. So it was a really interesting weird way of doing it because if people found out about that obviously they'd lose a load of revenue. So yeah don't all start kind of singular. If anyone from Zipcar is here, I should.

Thank you.

Thank you. I guess a question more for, Sebastian. Right? Is is on, the upsell.

What tactics or finds that you found successful on on the upsell page that really grabs customers and and ensures kind of a consistent conversion to, to actually paying and subscribing.

And so absolutely, you mean the we call it the subscribe screen like the the moment like where you inform you that we also have a paid offer.

Done.

The first thing is you put it in the face of a customer as soon as you can.

Classically, after an onboarding, if you wait too long, they don't see it, they're gone. So time is a crucial aspect.

And then, as hard as it sounds, but the classic marketing tricks. Like, don't say it's seventy nine ninety nine per year. Say it's, sorry, I'm bad at doing a math in my head right now. Well, like a cappuccino a month.

Make it comparable.

Yes.

Is, you know, you spend more McDonald's a month than on knowledge consumption.

So, then add a timer, say, this offer is just applicable for the next two minutes. It's just for you.

Half the price tiers like the unsexy high price for the CEO. Sorry. I'm coming back to this.

The very unaffordable low price and the mid tier that you want, like the silver platinum, you know, like that that stuff always works. It's you don't need a lot of research. Just try it. It it will work.

And like I said, transparency and honesty. Like, you shouldn't trick people.

That fires back fires, like, on your on the reviews and stuff, so be honest. But these simple tricks work. I I wouldn't even call them tricks. It's more you need to make sure people understand the value of your product. And if you make it comparable and it's harder to compare seventy nine ninety nine with that thing you just discovered to something that is such a low entry point, like a cappuccino or whatever. You you know?

That helped a lot.

But sometimes, literally just changing the copy from an inspiration we had and we just started a test, improved conversion rates by ten percent.

And someone told me, I don't know exactly the name, but if or, not the name, the the exact word. They tried out to replace subscribe with with buy now only in the US in the south, and it was a fifteen percent conversion uplift Just because it changed the word on the freaking button only in the south of the US, it's just like, that's crazy. Like, you just need to experiment yourself through that and, be creative, and then find a sweet spot for where whatever works.

Right. Yeah.

Hey, this might be another one for, Sebastian if you don't mind.

What would be your top tips? I know in the UK especially we've got a really busy summer of kind of sports and entertainment.

What's your advice for businesses to prepare from, like, a CTO point of view for, like, unprecedented amount of subscribers? So, like, things need to be at peak performance. So what would you suggest and what what's your advice there?

So peak requests I would say in general like when you have a very fluctuating business, how do you have enough resources at your disposal that you still can collect all the money that people wanna throw at you?

So on a very technical level, you do it by, being elastic. You know, it's a term you so you quickly need to respond to an increase in load.

There are other systems like content delivery networks that speed up your performance and you keep stuff where at where the where the spikes are, you outsource it to to vendors that do that for you. For example, we do it with Recurly. Uh-huh. That's why we're here.

I don't care. That's my point. With payment specifically, I just pay the recurly fee and I expect the, the system to work. And if we would have the CTO or like, a technical person here, they would say the exact same stuff. Horizontal scaling, you know, get more requests, add more notes.

And then you need to build a scalable system.

And it's I think AWS, when they started or before they started selling the services at AWS, it was Amazon and they had, like, loads of computers ready for Christmas, but they weren't utilized for the rest of the year. So they started thinking, we can rent them out when we don't need them. And born with AWS.

So, like, use the, CPU times or the power you need when you need it and make sure you can scale it up and down with a slider or automatically.

That's how you do it. And just trust that partners like Recurly, the App Store or GoCardless, whoever think alike.

And, on top, we tended in the past at least to prepare a lot for those events. So for us, Black Friday is a big thing and I usually reach out to Hannah and we say: Hey, Black Friday is coming again. Are your servers warmed up?

Are the integrations working? Let's not release something. Because usually only what you touch tends to break. So can we make sure nothing breaks at this time? The support, response time is really, really short.

Yeah.

But I would, I would just, by the way, that second point, I think is regardless of who, you know, what partners you have is a really important one. Like we had, we had a partner, who called me up right after I started and said, hey. We, like, kinda have this big thing happening, and we think we're gonna do we think we're gonna add over a million subscribers in one day.

We're like, well, that that is, like, that's big. Yeah. And it was it was an event, you know, big event and they're like, you know, and it can't go wrong basically. But the thing that they did a really nice job of was they sort of got they rehearsed it with their partners.

I mean they literally like we, we created a war room. We, like, we all, you know, ourselves and their other partners, we all rehearsed it together. When the actual day came or that, you know, that time came, we shipped our kind of our lead employees who are going to be accountable for that performance there to an actual war room. We were you know we're there together but I think that communication coordination just sets up a lot of, you know, a lot of success because we, you know, I think from a technology standpoint like we can, I mean, yeah, we can scale our services as well to go meet those needs?

The area where, like, you just don't wanna be a surprise. Broom?

Hey.

Thanks for the chat.

But one of the examples you mentioned earlier, the sports particularly.

How are you thinking about this behavior of customers coming in to get something or something, then churn?

Do you want to make it easy for them to sort of, like, cancel and then hope they come back in the future? Or Or should we make it difficult for them to cancel? What is the sort of optimal way of thinking about that? Because we saw customers budget start timing. They they have a limit now with subscriptions they can actually have a single time. And, you know, do we build trust by allowing them to actually do it really easily or, you know?

So I'm gonna give you, business perspective and personal perspective on this because I don't know. I think they maybe like they're related. Right? Like, because we're all we all subscribe to things, like, probably too many of them. And, so from a business perspective, what I can tell you is that we are seeing a very significant trend towards providing options to consumers. So things like pause cancellation is, like, is becoming very, very common. Downgrade, cancellation.

We're seeing things like we refer to it as stair step pricing but it might be that you know, you're, we've seen in both directions your price decreases decreases after two months as a thank you for, you know, for being here. Your price increases as you're getting more value in usage. But we're seeing, I would say overall, a big trend towards we want to optimize for keeping the person as an active user or let's call it an active active subscriber here because the acquisition costs ultimately are like are too high for us just to be continuing to kind of do that hamster wheel. So that's the big trend overall.

The human being point of view is I think I think Sebastian talked about trust which I think as a brand is like it's a balancing act of course but I think it's like, you know, it's quite important. Like there's there's a difference between giving an option to, you know, to pause and understand I won't be able to access this for a period of time and having to call eighteen phone numbers to try to cancel and still not being able to figure it out. Like that's a, you know, one of them is like I'm gonna tell I'm gonna, you know, I'm gonna tell my wife and I'm gonna tell everybody I know that this was a terrible experience. And and I think in the long run, like, we're in the we are in the we're in a lifetime value economy so to speak so I think that's kind of important but what do you guys want to add?

I just echo that. I mean, there's there's sort of nefarious practices that you just shouldn't do, like, don't hide buttons, don't have ten redirects to different pages, certainly don't you know, I think there's a I won't name the business, but it's a a media outlet in New York that makes you give, like, a mouth swab and all sorts. It's, like, it's full on.

So there's just things you just don't do. But then I think I think it comes down to that personalization point as well. Like, if you have the resources to do this, you know, there'll be if it's a digital script subscription, I don't I don't know what your, business is. But if it's a digital subscription content, look at usage.

If someone's not used for six months, if they come to cancel, just show them the cancel button. And, yes, okay, it's gonna be more expensive to reacquire that person, but you're not gonna reacquire them anyway so they're they're a dead user. If someone's super active and there's you're not really quite sure why they're, why they're canceling, then obviously you can offer more steps. So it's not about making it super easy.

It's about sort of, if you can, making it personalized and not jumping through hoops, but you know offering alternatives, downgrades, etcetera for those customers that you think you can retain.

Yeah. I believe that in what what you said is accurate. It's the same with the acquisition and the retention thing.

In my experience as well, we looked at all those things way too monolithic. Like it's just one customer. The customer always behaved like that. Statistically like so.

But when you drill down and you said that before, you need to understand your customer segments. Like, you acquire different people via different channels, via different platforms, via different intents, via different intents, via different ads.

They saw an ad and they saw they have a solution for me. I click there, I go there.

It's so many different reasons why people use your product. Might be they cancel because they just can't afford it right now because, I don't know, for whatever financial reason, and they wanna turn it back on, but they wanna be a customer that just can't. Others wanna exploit you, so there's a bandwidth of reasons. Please.

Just to add to that I think, what we've experienced at The Telegraph definitely is that subscriptions before the pandemic were were unknown to the user. The pandemic accelerated their experience and they are now super users and we have had to suddenly give them a new experience because they now know what they want whereas our models were built with what we thought they wanted and I think now we're learning that personalization is what they want.

They know when they want to dial it up and dial it down and we need to lean into you guys to know how we adapt to pause, to holidays, to tokenization content consumption. That's a must have changed again.

I would say it in a different way. It's not what they want. It's just everyone's reality. You know?

Yeah. Every one of us has a different financial situation today. It's different next month, and it's very individual. And if we can adapt to this individual situation of your not only financial reason but also whatever is happening in your life, you have a life changing moment and you suddenly cancel all your subscriptions, it's not because you don't like the product.

But if we make it too flat and assume everyone's doing the same thing for the same reason, we can't optimize for the individual needs.

Hi. Maybe this is a question for Joe.

What do you what do you see from companies doing, in this quest against German, towards ARPU, to protecting ARPU while while giving these, you know, massive offers and benefits.

Like, is your question about, like, as customers, like, I talked about downgrading or pausing as an example. Is that kind of the the question?

Yeah. Or a massive, reduction of, let's say, one pound for three months.

Sure. Sure. I mean, I think at the end of the day, it's about I mean, I'm gonna give you my CEO answer, by the way. Like, so I because we we're, again, we're a subscription business.

We face this question all the time. Right? I mean, I'm interested in what is the, you know, what's the average lifetime value of a of a customer? We have customer who's I mean, our business model is based on usage.

At the end of the day, we have customers whose businesses are growing quickly when we grow with them. We have customers whose businesses grew in the pandemic and then shrink and, you know, and and I would rather go kind of through that part of the journey, the downturn, but, you know, but believe that it'll it'll grow again. So I guess I think to answer your question more specifically, like, what we're seeing the same companies I talked about offering things like reductions are very much focused on upsell paths as well, growth paths as well. It comes back to this sort of personalization and segmentation thing, which I don't think anybody at least I'm not aware of examples happening really at a truly one to one level, but a lot of it based today on more logic at a segment level.

So we have this cohort of users who are, like, are heavy, you know, heavy consumers and would be interested in and, you know, an expansion of, price or, you know, added content or added, you know, added product or what have you. So I think that they're really they're sort of saying, look, we're gonna have certain cohorts of users that may be lower ARPU. We have certain cohorts of users that are gonna be higher ARPU, and that those two things can coexist together.

Yeah.

Just a quick question. We talk we're probably all here in the audience fighting for our own businesses or our own subscriptions.

But do you see the power of bundles and collaborations and opportunity or risk for the people in this room?

For for our business specifically, it's very it's a very hard question to answer. I mean, the the less businesses there are, the less payment volume we will do because everything's kind of, you know, obviously bundled into one, although you might have a higher transaction.

Yeah.

It's pretty sort of a churn aspect of that.

Yeah. So more more like the meta question about is it good for the churn problem.

Yeah. Yeah. So I'm sitting here with a, you know, guy from the Feet. Are we better off combining with an Feet and a which bundle to protect churn in the long term? Is that do you see that as an well from your kind of business as well do you see that as a potential opportunity in the future or do you see it as a risk?

I mean, I yeah. You you have obviously an opportunity to potentially wholesale bundle given the amount of data you've got on your on your subscription business.

Well I would say yeah I don't I don't think we're the seller of that but I think so okay here's I think about this a lot. Here's this is what I see happening is that I see as I said at the beginning, there are some industries that are more mature from a subscription standpoint than others. So the and I think the most obvious one is something like publishing, what you're talking about, or streaming media. Right? And that we all know that as human beings, individuals, sort of at some point you can only maintain so many subscriptions yourself.

And so I think it is, it is logical and we see a lot of experimentation amongst our customers when it comes to licensing and bundling agreements.

For us frankly, it's, we're we're pretty neutral in all of that. We're doesn't really change what we do. But the, you know, examples would be, in my Amazon Prime Video subscription now. I can access content from other streaming media services within there. Now I'm I'm upgrading as I'm doing that. So my ARPU is increasing there. And, obviously, those, you know, those payments are then on the back end being split in a revenue sharing agreement there.

And for me as a consumer, it's a pretty good experience, honestly.

You know, kind of gives me one fewer thing to manage, screen to go manage. So I believe in more mature subscription industries we are going to see more of this kind of I don't think it's, I mean, we will see obviously, we know if you look at read the press, we know there are mergers happening from, from that standpoint. But I think beyond that, just more experimentation with bundling and licensing of content.

So I think there's a lot, like, to me, it feels like it's a conversation to be, you know, to be had. And, again, I know in the US market in particular, we're seeing a huge amount of our customers experimenting with that.

Again, I think there are other product categories where that's much further, you know, probably much further off in the future.

Just a kind of tangential point.

I don't know if anyone in the room has this problem, but I quite often go and look at my Google payments sort of subscription management thing, and I'm I've got, like, sixteen there. So I just cancel, cancel, cancel, cancel, cancel, cancel, because I have to because otherwise so I think the the issue is around there's two there's two things here. One is around sort of managing your subscriptions as a consumer and, like, sixteen is too many to to run through one single wallet because one, you know, you don't really pay attention to all the price points. You don't really see that you don't really understand the value of each of those individual things. So Google are causing you a problem there in some ways because they're sort of bundling, passively bundling things next to each other and making it look like, you know, these are all your things through Google.

So I think that's kind of that's both a it's supposed to be a good thing because it is helping to manage it, but it's it can cause it can cause a mass churn event when you go and check your Google, your Google Wallet, you know, every year and you've got fifty subscriptions.

Can I just add to that? So we've seen a massive difference in payment success with Apple Pay because there is that involuntary churn that's so silent to us because of that very thing. You clearly look at your subscriptions, and the last person to know is your provider because you've gone, what? I didn't know.

And it's so silent. And it's great to have that quick entry point, but it's a quick exit point. See, we have to think about your offer, and it's what we were talking about earlier, and you were saying, like, do we give them a longer free trial period to try and build that habitual nature because we know they're the likelier to churn versus the likelihood to convert. So you have to be careful with opening up and that's why like though Carlos is great if we get people on direct debit but then we also have the challenge of what's the grace period or retention strategy for the abolished charges closing down a payment method through a bank transfer we have challenges there as well, which I'd love to talk about, but it's easy and easy out.

It's it's the phone.

We have tried bundling up with others, and it had ups and downs.

It created reach because we suddenly hooked onto someone else's marketing.

So I think we acquired a few customers.

On the other side, if you're in a bundle, you have suddenly five products you need to think about, if they're worth it or not. So why do you make your decisions upon? Because you're using one, you keep the bundle. You use two, you use the bundle.

And if you then have a revenue share model that's based on usage, you're not earning anything if they're not using your product. They bought the the bundle because of something else, but it just came with the bundle, but you're not using it. So you're not making any money out of it anyways. So you might have a touchpoint with the customer, but, if it's worth it, I don't know.

From a management perspective for the end user, definitely cool. Like I said, I've got Apple, what is it called, this iCloud plus thing because it's a storage, family plan, music. My kids play arcades.

Yeah. And I don't use the other stuff. But, I think it has its pros and cons.

Risks. People don't like your product, they won't use it, they won't pay for it. I think it's one of the things you should never forget.

Cool. Okay. We have time for maybe one more question, and then I think food is starting to happen shortly thereafter. But but they need five minutes, so we should ask one more question.

I think I know the answer to this question, but, a a lot of us probably wouldn't be here if, if it was true but, lots of discussion around subscription fatigue and just discussing the the sort of Google pay option to kind of, Oh sorry, the Google Chrome option to get rid of all your subscription Subscription fatigue, miss or reality?

Yes. Of course.

I don't know. I have no data. But what we saw is when the pandemic hit and people started I think in the beginning, everybody went like, we lost a huge revenue portion with the pandemic because people stopped commuting. And that was the sweet spot for us when you work on a commute at twenty minutes time, went for a blink, blah, blah, blah.

What we saw is what what we heard before, that people learned what subscriptions are about. So they suddenly got like an audible subscription, a Netflix subscription, like this subscription, and all the other learning things subscriptions.

And after they realized the pandemic is continuing, I might have lost my job, they started to filter out subscriptions. And so we were not really high up on the prior list because the water subscription is more important. And so, we already saw the effect of people turning off their subscriptions, especially when Apple introduced, I think it was last year or the year before, like, they they were more eager on pushing people in, like, they started active communication with push notifications about renewals.

And we saw a massive drop. So the moment people got educated on what they're actually paying for all the time, they started to say like, Oh, I'm actually not using this. I'm not using this. Chrome extension.

Whatever. So, I think the moment people have less money again available for the disposal, they need to make a choice. Like, do I keep it or not? And, you start cleaning up with the ones you can easily clean up, your water supply likely, well, turn off.

So we've seen that, but I have no data points.

So, alright. So I this was my number one question when I thought about joining Recurly.

It's like, where are we in this subscription thing? You know, is this like the beginning, the end, the middle?

So this is the best data that I can see around it is I think we of course are we're in the middle. So the early days of like I'll just subscribe to anything and it's kind of exciting, it's cool, that's probably over. I'm probably we're in a maybe a more critical state where we evaluate. We're thinking we're checking in once a year on you know our Apple Wallet or our Google Wallet thinking about the subscriptions. I think that's true, but I don't think we're at the end.

If you step back and look at subscriptions sort of role in the economy globally, best estimate is it grew about eighteen to twenty percent last year, which is much faster than the global economy is growing. So despite the fact that, like, we are maybe making choices about which subscriptions do we care about most, how do we optimize those, the reality is that the big picture is it's growing. We look across the data of our own network, just meaning our customers, our existing merchants business at the beginning of a year last year versus how those same businesses grew over the course of last year, they averaged about the same about eighteen to twenty percent growth over the course of the year.

So it kind of lines up like we're, you know and of course we see businesses of different, you know, different kind of curves along there. But I do think in the big picture that like these things can coexist. That consumers are getting more sophisticated and smarter. Like, we're all getting more sophisticated and smarter about how we think about subscriptions.

And of course, luckily, we're not trapped inside of our flat, you know, anymore just having to, you know, subscribe to whatever we can get our hands on. But at the same time, subscriptions I think are still in the early stages of how they kind of shape our life. And like I love the car example at the beginning because I mean we see really interesting evolutions. I mean we have, you know, airlines who, joined us as customers recently, which doesn't seem seems like an odd odd use case to me at first, but they're really they're rethinking the loyalty model.

You know, maybe rather than earning points, you're gonna pay nine ninety nine per month. That's gonna get you access to the best seats and, you know, through the queue and priorities and and all of that. And so I think they're, at the end of the day that's sort of that's that's my my best honest assessment and I've tried to be pretty honest with myself about that but that's where that's where I think we are.

I concur. I think, I I don't have the data either. I mean, I I would, I would I would concur. I think we're we're subscriptions are still growing. It depends on the type of subscription.

From a personal perspective, I'm I I get fatigued by, you know, twenty five different cheese box options when, like, you can just go buy some cheese and, like, does does a cheese box subscription really make sense?

Digital media, that is, you know, that that I think is kind of getting to saturation point, and then we ask that, you know, we see lots of bundling happening in that space anyway.

And then, yeah, but I but I don't I don't think we're we're anywhere near the end because, like you say, there's so many new models, kind of arriving. I mean, I the interesting one for me is the access one where you know all of the delivery services and Uber and all those folks do it I don't know if anyone from Uber or Deliveroo is in the room where you know you pay a subscription as well as paying for females and things like that so I think they are sort of almost supplementing some of that fatigue from the kind of everyone wants to do subscription, dog food boxes, cheese boxes, etcetera.

Cool. Well, I think okay. I think that's probably a good pausing point for us.

This was fun.

Thank you for thanks for the questions. Thanks for the conversation. Again, I mean, I bring us back to the beginning. Our goal is make a new contact here, somebody you can hopefully reach out to, you know, in the future. You have a question, hopefully walk away with a relevant idea from this.

We certainly love I know we have folks from ourselves will be here. We have folks from our teams here as well. We'd love to be of service for you if we can, and we certainly love to feed you, and and give you another drink. So we just we're free to we're free to do our to mini. Yes?

Free to drink. Okay. Cheers.

Thanks.