Paywall Optimization Best Practices

A deep dive on paywall design, timing, and testing

Wednesday, February 22, 2023 at 1:00 PM

If you’re working on a subscription app and not thinking about and experimenting with your paywall, you’re doing it wrong. While a great paywall isn’t going to magically make up for a sub-par product, optimizing your paywall and freemium strategy is key to building a great subscription business. Jake and David will chat for a bit, then take questions from the audience.

Topics we cover:

  • The best content to paywall
  • How to effectively test pricing
  • Paywall design best practices
  • Writing copy that sells
  • When to prioritize annual subscription plans over monthly plans
  • How to offer discounts to low-intent users
  • The best times to show your paywall

Hello, everyone. And hey, Jake. Thanks for joining me today. Yeah, so a little bit of housekeeping. We are going to record this. I like to dilly dally a little so folks have time to show up, so I’ll just kind of ramble for at least two minutes. So we are going to record this now. You know, no need to furiously screenshot or or, you know, screen recorder or whatever. We will release the recording. You’ll get an email after this with the recording that you can share with colleagues and whatnot. Excuse me, another bit of housekeeping. As I as a host, I do a terrible job trying to view comments and chat and also your questions and everything. So what I like to do is go ahead and enter your questions. And the questions have there should be down below a tab specifically for questions versus chat. So we will leave time at the end to do a Q&A. And I’ll start with the questions that are in that tab. So if questions come up while we’re talking, go ahead. In interim there, if we answer them, I think there’s a way to delete. So if we answer it during the session later or something like that, you can go ahead and delete it. So we don’t cover something that was already covered. Yeah and, and that’s about it. So I wanted to introduce Jake Moore. He’s the CEO of SuperRare wall. And so what we’re going to do, Jake Jake is put together this fantastic slide deck on I’m building great paywalls that sell and so we’re going to go through it. And then, you know, as he goes through it, you probably have a million questions pop into your head and I already have a ton of questions for him. So I’m going to try and be your voice asking him questions as we go through. And then again we’ll answer even more questions at the end. So without further ado, Jake, let’s get rolling if you want to go ahead and share your screen and talk to us about it. Walls for sure. Thanks for the awesome intro, David. All righty. So today I would like to talk to you about building paywalls that when I’m going to show you best practices for iterating towards better, more effective paywalls. A bit more about me. My name is Jake. Before founding super well with two amazing co-founders, Brian and Justin, I founded a company called fitness AI. It was an iOS app that told you exactly what to do at the gym. I grew fitness Ii to around 2 million in revenue. It was a top 50 grossing health and fitness app in the country. I handles all engineering, design and growth on my own, and I did that mainly by building a lot of internal tools, kind of like super wall. So the first thing I learned while building fitness I was that growing apps is all about optimizing funnels, whether it’s from add to install, from install to onboarded, you know, paid to, retained from. Most people miss out in this last one. Once they’re a user, you need to get them to share the app with their friends and you need to get each user to leave a review. So all if you break these out into standalone processes and figure out ways to optimize them, then you’ll win. Again today, we’re going to talk mainly about how to get users to go from install to paid. That’s that’s the mission of Super Bowl. And whether you Super Bowl or decide to take these tactics and build them your own, the same way I did with fit and say that totally works. The principles still apply. So the thing is that users rarely subscribe to your app without a good paywall. And the reason is because they haven’t used your app yet. They don’t know who you are or if they can trust you and they don’t know what you do. Right so how could you ask someone to pay you without getting this in their head? Most apps do it by going through the freemium model. So you give parts of your app away for free or you let them tinker and play around with your app. But the truth is, a fantastic paywall is a better solution. Yeah I want to interrupt you there. Where do you think the balance is? Because, you know, it does seem for some apps, you know, freemium can help demonstrate the value even better than a paywall. You know, if you get a really experience the value of an app that can actually potentially even sell better than a really great paywall, where do you think that balance is? You know, are there specific apps you think should maybe lean more heavily on freemium or do you think this really is more universal? Absolutely the most important thing is getting to a user to an aha moment as quickly as possible. And so however many users you lose to get some users to that aha moment is your opportunity cost for building a fantastic paywall. So if you’re losing out on 50% of your signups because they’re not even reaching the aha moment, then it would make much more sense to add a paywall before the aha moment that at least describes what the app does in a really good way. And tells them how much they have to pay to access that. The best app I’ve seen that gets to that aha moment really quickly is photo room. So in their onboarding experience, they literally make you add a photo. And remove the background. So, you know, something like 80 plus downloads probably experience that within the first minute and then right after that they get hit with the paywall. So it’s really just comes down to that aha moment. The thing is, most new apps or apps in general, there’s so much more work to do to get to that aha moment that it’s so it’s such a hard thing to do that. It’s much easier to optimize your paywall, to at least share what that, what that thing is. Yeah, that, that’s a, that’s a great answer, really making me think so. So this begs the question, what makes a fantastic paywall? There are five key elements that we have that we’ve zoned in on. So the first is, is your placement. So you want to make sure that your paywall is being shown at the right time. The second is that you’re locking off the right features. So that’s deciding if you’re freemium, if you’re pay to play, if you get three free workouts, et cetera. The third is pricing. So you want to make sure you’re charging the amount that maximizes LTV for each of your users. The fourth is visuals that provide good context as to what you’re providing. So a great example here is like the Blinkist free trial timeline that falls under the visuals category, and the last is copywriting. If you look at columns paywall, there’s just five bullet points on it. You know, that’s the famous paywall that took them to 100 million plus in revenue. And it’s the reason they were able to get away with that is because their copyright, their copyrighting, explains the why and not what. So that’s how a common piece of advice that all marketers sort of live by. Each one of these, we’ve been able to optimize on average by around 20% That’s like a baseline if you haven’t optimized any of these things. And it sounds crazy, but the opportunity for optimizing all these things, 20% is a 2 to 2x increase in revenue. So Yeah. So here’s an example inside a sit and say I this was two years in on August 16 of last year I added a video paywall. So I added a video to my paywall and I moved to the paywall, too, before onboarding. Right so this is two years of optimizing in this happened. Right? so this is sort of proof, you know, Blinkist also approved of the free trial timeline. Doesn’t matter how long you’ve been iterating. There’s always that Nugget that gold that you find that can double your install to trial. Yeah when you did this experiment specifically I would assume that with the increase in install to trial that you would see kind of a commensurate drop in trial to two paid. Was that the case in this trial? No, I saw two paid in startup. Paid went up to x as well. Wow and this is crazy. And so there are two pieces here. So the first is we move the paywall to before onboarding. So our onboarding completion rate at the time was it floated between 60% and 70% So moving that beforehand was a 50% relative increase, just number of people seeing the paywall. So you’re going to get some gains there. The second is that we added a video to the paywall. So my hunch is that most people didn’t know what they were buying when they got hit with the paywall right after onboarding. But adding the video sort of showed off exactly what the app does. I could show you a demo. Of what? Yeah yeah, that’d be great. So this is the paywall that we updated it with. You can see that the video. Sure maybe not. OK yeah, go ahead. This is this is the pay wall that we designed that. So it shows exactly what the app does. Users know exactly what they’re getting. I think this is really important, especially if you’re going to be aggressive with your pay wall and show it off before they even get to play around with your app. And, you know, the truth is, a video can do a better job of watching a user through the app than just giving them access to it because they don’t know how to use it. Right so videos are probably the highest leverage thing that we. That we recommend. You know, this doesn’t need to be so clear cut. Like with a device model and you could just do a random video like this. This is a video editing app. That a friend of mine created. And it just shows you that you can make cool videos with it and people sort of get it. This is the best performing paywall we’ve ever seen on our platform. There’s like a 20% install Ii trial rate on this one. Wow, that’s fantastic. All righty. So most tactics don’t apply between apps in different categories, right? So for example, with health and fitness, there might be. There’s a weird thing that happens in health and fitness where the more people, people who put their money where their mouth is actually end up going to the gym more. So every category is sort of different. All we can do is share what has worked in the past and the tests that we ran, and hopefully you can run those tests inside of your apps. So you absolutely must test these things for yourself. Yeah, that’s really interesting because like in your previous explanation of how you had much higher trial conversion rates with that video paywall early on, I would imagine there’s maybe a little bit of like a kind of a sunk cost of people who once they start the free trial, they’re actually more invested in trying to actually use the app and see if it’s a good fit during that seven day free trial. So, I mean, do you think that’s maybe part of why those paywalls and putting the paywall so early actually works in your specific case? Absolutely I think a common thing that app developers do is they underestimate how much intent is behind the install. So there’s a user who downloaded your app because they’re there. They specifically are experiencing some issue and they did at least a minute of research and they chose your app and they’re downloading it. And so you got to capture that demand. Yeah and capturing it early to point is that if you wait until they go all the way through onboarding to show them the paywall. And only 50% of people are converting or are completing your onboarding, then you’re missing an opportunity with 50% of these high intent installs to even show them the paywall. Absolutely and yet intent is fleeting. You know, people they’re really down to try out your solution and they’re like, yeah, you know, I’d rather just go back on Instagram or something else. I see. You got to capture it. Right and the thing with paywalls is you miss. You miss 100% of the shots. You don’t take the number one metric that we focus on that we see almost every app company not focusing on is what percentage of your users are even seeing the paywall? And it is crazy. I mean, we see some apps doing three or 400k in revenue a month, up to a few million where only around 20% to 30% of their user of their installs even see the paywall. Right so, I mean, that’s crazy, right? 80% of people download your app and they use it, but they never even realize that there’s a paid option. Yeah and people feel bad when they’re aggressive with their paywall. But the truth is, you need money to invest into your app, and the user can also always exit, exit it out. You know, it’s not like a you’re not forcing them or even it’s not as aggressive as you think people are used to it. Yeah Yeah. That’s that’s really great advice. So what makes a fantastic paywall? Again, there’s placement features, pricing visuals and copy. Let’s talk about placement. So these are your options. You can show before onboarding. After onboarding. Before using a core feature. After using a core feature and on every single app open. We recommend to show your paywall in all of these places, at least at first, because it’s important to know what your opportunity is. So if you’re the kind of person who thinks showing a paywall before onboarding is too aggressive, showing it after onboarding is too aggressive. The real the truth is, is it depends, you know, is it too aggressive if you’re going to 3x your revenue and quit your job to work on your app full time? You know, not really. Right so you have to know what you’re missing out on. So at least run a test with all of these things, do it for a few weeks just to see what happens then? Keep on removing ones until your conversion rate stabilizes and you’re in a place where you feel comfortable, like you’re not being too aggressive. Does that make sense? Yeah, that’s great. So the second is, what features are you going to like? What are you going to feature? Great so there are two options here. You can pay all your best features with a freemium model, or you can paywall your entire app with a pay to play model. Usually when you’re creating a new app, we recommend locking out your entire app. This usually does best. It’s not for the reasons you might think. When you first launch your app. Your app is the worst it’ll ever be. And when you have a user who’s willing to pay for that app without ever trying it right, or at least starting a free trial, that user has such a huge problem in their life that they’re willing to try this app. Right? the shittiest version of your app to see if it’s a solution. And those are the exact kind of early users you want. Those are the people who you want to be building for. And so. Lock out your entire app, at least when you just start and take your lesson to your users and take their problems as gold and solve for them. One of the things on freemium strategy that I’ve been thinking a lot about and I’m curious to get your take, is that. I think the tendency, maybe the wording thereof to challenge you on is that locking your best features, is it always and lock in your best content? I wonder if sometimes that’s actually a bad approach, depending on whether or not your best features and your best content have some level of self-evident value. Because I think what I’ve struggled with in the past is like when I introduce a new feature in my app and I think it’s fantastic and the value is clear. But I think to the average user. And maybe, you know, again, a video paywall can help with this. Good copywriting can help with this. Like there’s a lot that can help with this. But if it’s something the user doesn’t necessarily Intuit the value being there, if they have to really experience it to get the value. I really struggle with how to balance those kind of features when they’re are maybe technically your best features but also maybe like need to be experience. Like what are your what are your thoughts on that? Totally nothing is stopping you from. At least at the start saying you need to at least start a free trial. Remember, they’re not paying anything. They’re just starting a free trial, starting a free trial to unlock the entire app. And after a week, if they don’t subscribe, send them a push notification and let them know that they can now use half the features for free. Right you know, you could move and jump around between these things to figure out what configuration works best. Also, there’s another this was a really good point that one of the founders of fit bod made, which was if you have a set of features and you decide to, let’s say, let’s say for your app to work, you need to charge $100 a year. Right? if you go the freemium route and you decide that some features are free and some of them cost $100 a year, you’re actually valuing you’re completely devaluing your free features. They’re literally free. And you’re saying that your best features are worth $100 a year, whereas when you say that your entire app is worth $100 a year, you’re not like placing your features above or below one another. And so. There’s no devaluing and there’s no like, you know, maybe these specific features aren’t worth $100 here, but together they’re worth $100 a year. I think it’s an interesting way of thinking about it. Yeah yeah, that is it’s also like if you still want it, like instead of going freemium and feature getting half of your features, something that you can do that. We’ve seen work pretty well, especially in flipboard’s case is it’s great. Like you can give away your entire app for free without starting a free trial through apple, but just make it a usage based limit so you get, you know, photo room might give you three free photos with premium exports available or, you know, fitness say I might give you three free workouts. Definitely don’t do that with the free trial through apple, though, right? Oh, that’s another thing. Usually with a freemium model, we see non-free trial products do better. Whereas if you feature get your entire app, it almost always makes more sense to add a free trial. A time to free trial. Right so third is pricing. So the trick with pricing is showing the right price to the right person. And the reason is because you’ve already paid for the install either through ads or, you know, they found you organically. And it’s best for you to recoup at least any amount of that investment that you can. So if you paid $5 for the install, for example, and the users are only willing to pay $1 or they won’t pay at all, you may as well take the dollar, right, because you already paid for the install. Again, this is considering in a vacuum where, you know, users don’t communicate with each other and all that stuff. But the general principle is with software businesses, there is no costs to actually like there’s no variable costs. So it doesn’t cost you any more to have three installs versus, you know, one install. So you may as well recoup any amount of the investment you can, whereas if you’re selling books, you know, the book might cost $5 to make so you wouldn’t sell it for four. And so something that we’ve been exploring is meeting the user wherever their demand is. So if someone hasn’t if someone downloaded your app but they haven’t purchased it for a few weeks, maybe show them a cheaper price. If you have areas that are low, like they have low conversion rates, show cheaper pricing there. And that’s something that anyone who localizes their app will tell you that they show different prices and, you know, in China and in India and in Europe than they do in the United States. So there are some geographic pricing strategies that you can do. Again, giving discounts over time and offering web checkout experience probably through revenue, CAD and straight. Usually you could split the difference with the user because Apple doesn’t take a cut. And that’s a great way to get to drive the price down. But the key takeaway here is you should be giving discounts to your users over time. That’s like easily a 20% to 30% price increase in revenue by doing that. Other things to test. You want to test huge price drops versus small ones. So if you’re charging 29, 99 a year, don’t test that against 39, 99 a year. It probably wouldn’t be that significant. You should test it against 90 bucks a year and 140 bucks a year. I know it sounds crazy, but it’s your job to figure out how you can generate the most revenue. And then again, nothing is stopping you from charging 60 bucks a year on day one and then offering discounts throughout the lifetime of the user. And I’ll show you guys later how to do that with, you know, my favorite setup and the best tools to use. Another good one is offering lifetime value at 1 and 1/2 to two year LTV. This is great because I mean famously does this at like I think 300 or 400 bucks to unlock it forever. Something worth testing. Definitely offer sales. So Black Friday, Christmas, you know, mother’s day, father’s day, all those good things. And again, notification reminders, testing different trial lengths. And don’t forget to track your refunds. This is huge. You might move to a three day trial from a seven day trial and Jack your price up only to realize that all of your gravy is being lost to refunds because people are unhappy. So keep an eye on that. Another tip only offer free trials on annual products that pushes people towards the annual subscription and you do want to test showing like multiple options like one, two and three products. But it’s really tough to figure out LTV on the paywall and on each specific product so many times 1 annual product does best. It’s the best way to gauge what is the demand for this product, just showing one product, whether it’s annual or monthly. Thanks Yeah. There’s something I really liked on. On that pay wall that you showed us for fitness is that you just had this single, like, start free trial button. And, you know, in my apps and a lot of apps, there’s like, you know, the annual will be highlighted, but then you will have the monthly and the lifetime available as options. And imagine like in paywalls, like further down the funnel, you might show that. But for that first paywall they see just making like taking away the decisions seems like a really great win. I mean, is that how you would typically approach it is to go ahead and in most situations just show a single call to action. And then maybe in other situations where there’s maybe lower intent or you’re offering a discount, then you give them more options to kind of meet them where they’re at. Exactly it’s never clear cut. So we recommend at least starting with just one annual product because it’s a really quick way to figure out what is your LTV, right? So if you’re offering a weekly product or a monthly product, you won’t know what your LTV is for around six months at least with annual you can say that the revenue you’re getting on day one from this annual product, that’s a floor on what your LTVs. Once you’ve established that, then you can start adding multiple products and tracking those LTVs over time and looking at what their attention is. But it only makes sense if you have somewhere like some sort of baseline, an LTV, and the quickest way to get there is with one annual product. And again, there, there, there are many options here. So like, for example, here’s, here’s planters paywall. Hits this usually it’s not templating, right? Because I’m just opening up a new browser. There’s just one button that says subscribe, but they have the other plans button. Right not many people do the other plans, though. Yeah, but but this is a great way to capture that person who isn’t willing to pay, you know, 100 bucks upfront for the year. You know, they will see that other plans button. But if you are willing, it kind of just it fades away. Yeah and you don’t really see it. There are. You know, obviously, if you’re a really good product and you retain really well, the product that we’ll probably win is a monthly or a weekly product because you have a really long tail of revenue that you’re missing out on. But again, another approach is you should think of it as planting seeds. So like something I did in fitness, I allottees and I’ll choose one month out of the year to just show tons of products to random people. So some people would see only $299 a week. Some people would see, you know, 90, 99 a month, basically plant seeds. And then that same time next year, go back and see what happened. So only show those options for a month, then go back to your winning option and then look at that cohort over time to see which product is generating the most revenue and, and, and so forth. Again, you need a lot of skill, though, to be able to pull that off or you need to leave it the experiment running for a long time. That’s something we see, too, that people underestimate. They don’t understand statistical significance and at least so many. But yeah, like experiments can flip-flop so many times. Yeah, that’s great advice. I especially love that other plans. And it’s a really great way to kind of offer a, an offering like a, an alternate choice, but in a subtle enough way that only people who are looking for another choice are going to see it rather than always showing all the choices. Yeah, totally. Really fantastic. Another great idea that I love. I haven’t seen it much in the wild, but financially I does. This is even though our paywall only shows one annual option. With the 90 bucks here. This, by the way, isn’t. This isn’t the pay wall that Apple approved. This isn’t what it says. Usually it says subscribe like start seven days free, then 89, 99 a year. Right over here. The best thing I’ve done that I really like is users can only subscribe to the 90 bucks a year option, but they’re 7 days free, right? A lot of people think I’m not really down to pay 90 bucks here, but it’s seven days for free. At least I’ll give it a shot. Let’s say at the end of their trial, they feel like they’d rather pay monthly or weekly when they go to cancel their subscription inside of Apple. Inside of this subscription group is another option that’s 15 bucks a month, but there’s no way to actually subscribe to that from within fitness. You could only see that if you’re going to cancel. That’s smart. And around 15% of our revenue is on that product, which is pretty substantial. That’s like total like a win back campaign that everyone gets for free. Yeah so this is something, you know, I’ve talked with other developers about because people are like, well, you know, if I’m experimenting with pricing, you know, people are going to go to this page when they go to cancel or manager subscription or whatever, and they’re going to see these other prices. Like, I don’t want them to see these other prices. You know what? If they pick a cheaper plan. And what I tell people is like, why not? Like, why not let them us see the. Because in some ways it’s like kind of a marketing and user psychology thing is like if somebody feels like they’re getting a good deal by kind of working the system and getting this cheaper price by going to cancel, it’s kind of like a built in win back campaign. Absolutely if they’re into me, it’s also it’s a good place to do it in some ways, because I hadn’t thought of doing it in this way. You’re talking about creating a specific subscription group with just these two specific things in it. But it’s a perfect time because they have the option to cancel, like they’re right there with a Cancel button in front of them. So you’re not like tricking them, you’re not using dark patterns, but if they still have intent, if they still want the app, you’re giving them an opportunity to continue paying in a way that’s more convenient for them. So it’s like the perfect way to do that. That’s, that’s like, oh, absolutely. Love that and own up to it to make the product like call it like Easter egg pricing, like surprise, you know, so when they’re going to cancel, they like it catches their eye and they realize they found something. Maybe they’ll share that with a friend. Yeah Yeah. It’s like if people are, if people are telling all their friends, hey, sign up for the 99 a year and then go to cancel and you’ll find that it’s like $15 a month plan or $10 a month. I mean, you could even make it a better deal, like you said before, like, you know, give them better pricing. Yeah if they’re telling their friends, it’s actually great. Yeah, totally. Yeah OK. Fourth is visuals. So which visuals work best? Charts and visuals. What they do is they convey meaning as to what you’re offering, how your trial works. It’s yeah, it’s like the famous a picture says a what is words or whatever. Maybe a million. But yeah, a video is definitely a words. Video’s worth a few billion. So we’ve seen much higher impact from people adding videos to their paywalls than the free trial timeline. Also what a lot of people don’t realize. They don’t realize the process in which Blinkist found the free trial timeline. The way they got there is by adding an exit survey. So if you saw your paywall, if you saw the paywall and you exit out, they would ask you, why didn’t you just start your trial? It’s free. You know, why wouldn’t you start it? And for them, the biggest response was, I didn’t know when I would be charged. And so that’s why the free trial timeline works for them. In apps where the free trial timeline doesn’t work. It’s likely because users aren’t confused about when they’ll be charged. But the real lesson here is exit surveys. That’s what that’s the real lesson. So listen to your users. Understand why they aren’t converting. Produce a video to help show them why. Or produce a chart or a visual that helps convey meaning. Something that we barely see and apps that do that do great are Q&A sections. You know, there’s a reason why, like web paywalls have Q&A sections. I don’t know why many apps don’t do that. Another one is live support. So on fitness ace paywall, you can tap help and message us. Someone once told me, if your support team isn’t paying for itself, then you’re doing something wrong. Your business, you provide solutions. And so if someone’s coming to you with a problem, you should probably find a solution to sell them. So definitely adding life support is a good investment. And try to pull in real reviews from the App Store. You can use the app figures API. You can use a web view to pull live reviews and show users what real people are saying. Another thing for reviews make sure there’s always a name and a date. Otherwise, people. People smell bullshit. Yeah good tip. So videos build trust. Here’s the before and after for an 80% increase in conversion and I think a 50% to 100% increase in revenue. It just shows what the app does. It doesn’t need there’s no audio. There’s no text even. If you have trouble building something like this, you can do it yourself with an app called rotator. They basically let you just take a screen recording and drag it into a 3D model of a phone and it like creates these really cool animations for you. So some tips for visuals. While some designs are for sure game changing, like the blink as timeline or like a video while written text without any graphics can usually outperform like. So you have to assume at least tested everything. So hopefully, hopefully in their experiment, their plaintext did better make use of checklists and big fonts. This is like a big mistake. We see people do all the time. The purchase button should be the only colored button on your screen. So like, you shouldn’t do something like this on the left and you should most definitely opt for something on the right. There’s a famous USC experiment where they did. I hate I hate quoting experiences with experiments without knowing the source. But they wanted to see if a red button, a blue button or a yellow button did better. And they found that for each brand, it’s different. But by far, something better than choosing your best color is making sure that the CTA button is actually the only colored element. That’s where you get your biggest boost from. That’s amazing. OK lastly is the copy. This is really, really important because it’s the cheapest thing to change. So whether you use super wall or have some fancy paywall set up, everyone has access to Remote Config for free. Everyone who uses Remote Config can change the text on their paywall. It’s free, it’s easy, and there’s a huge boost that you can gain from it. So if you’re advertising test the same messaging as your top performing ads, if you don’t know where to start, look at your best reviews that people leave you on the App Store. That is literally, in a user’s words, why they love your app. Use that in your advertising. Keep in mind that humans are emotional decision makers, at least like, you know, there’s the fast part of your brain in the slow part of your brain. The fast part doesn’t work the way the slow part works, meaning it makes quicker decisions based on different based on different metrics and emotions. And the reason building a paywall or writing copy is so it’s so counterintuitive is because when you’re writing, when you’re building the paywall, you’re using the slow part of your brain. So you’re trying to say, look, how like, how worth it this thing is? Like, look at all this thought you could put into figuring out why, you know, fitness makes more sense than a personal trainer. The truth is, that doesn’t work. What works is selling the outcome of you’re going to look your best and feel your best and you’re going to save so much time so that you don’t have to spend that much time at the gym. That works much better than this is 20 times cheaper than a trainer. And we use AI to figure out how many sets, reps and weight you should lift. Those are features, not outcomes. Right so sell the outcomes. Go a step further and tell the user how the outcome will make them feel right. So, like, boost your confidence with a nice smile. Right that’s the toothpaste they are. I don’t know if it’s Colgate or whoever it is. It’s not like, you know, 3x whiter teeth with brushing. And half the time, you know, it’s build confidence with a nicer smile. And again, social proof always does great. So if you can figure out a way to hack it, to show reviews on your pay wall, live reviews that goes a long way. So again, copy is the first thing you should test and at the very least, make use of checklists and big fonts to. Keep your biggest ideas in those places. Any questions on copy? No, that’s great. All righty. So for a two, two, 2 and 1/2 acts increase in your revenue, optimize your placement, your features, your pricing, your visuals and your copy. We’ve seen it work on apps all over the place. In every category. There is no exception. There are tons of possibilities. There’s a lot to test. Brace yourself. It’s not easy, right? If only there was a platform that let you test and iterate on your paywalls remotely. So that’s super. Well, that’s exactly what we built. We’re very early on in our journey, but we’re, we’re getting there and it’d be great to have amazing apps to work for, honestly, and, and getting them to a higher revenue. Yeah I got to tell you, Jake, I’m sort of there. I mean, I’ve seen I mean, I’ll just give, like, very live reactions here. I’ve seen your product. You know, I’ve had the conversations on Twitter and privately about paywalls and how important they are and stuff like that. And like this conversation today has changed my mind on a lot of things, and I’m excited to try these in my own apps because you’re exactly right about so many of the things that I’ve gotten wrong over the years. Like I’m that app that like, you know, only 20% of users see the paywall. You know, I’m not. So I made a huge mistake in my watching our pro app a while back where they didn’t see the paywall in the onboarding. It set these expectations that the app was completely free. And then I pushed a lot of people off because I put the paywall after they had already done a bunch of work. And it was just a terrible user experience where if I would have put it in the onboarding or even the first thing, it at least sets the expectation like, OK, I’m going to use this app, but at some point I might need to pay. And not even showing the paywall ends up being a worse experience when they do get to the paywall versus like giving them some idea that there are going to be paid features. So like I’ve made every mistake in the book and going through this presentation. And then thinking about doing some of this in my own app, like, yeah, you got a new customer today. That’s good to hear. Yeah, building. Go ahead. We’re building out free paywalls for everyone who signs up. We have someone full time making paywalls now. Oh, Wow. We’re including all the best practices that we learned from designing literally hundreds of paywalls and running hundreds of tests. And we’d love to see launch Center Pro in there. Well, all right. And then we are running short on time, so we need to speedrun this. So we have time for questions. But for sure, you did show me this amazing diagram that you just racked up to kind of show the whole workflow. And so I’d love for you to just speed read us through this. Real quick and then we’ll move on to questions for sure. I’ll share this graphic also with you and on Twitter. So that because there’s a lot to digest here. This is the best flow that we’ve ever seen. We haven’t seen any app do all of this. We’ve seen some apps do part of this. This this is basically our road at Super well. So up top, you see how people can find you either on Google through some sort of performance marketing like Facebook ads or organically on the App Store. This shows you exactly like what APIs you need to build for, what layers, what you should be able to, you know, remotely control. But like you can see, this is like a gold standard of something to iterate towards. So like, for example, on app install, you want to go right to a video app paywall. After that you want to get push notification access. Then you want to go through onboarding. Ideally, you have your onboarding questions, data driven by your API so that your web onboarding is the same. After that, you collect your email signup complete. Now, if you’re on the app side, you show them the paywall again after onboarding. If there’s a trial start, the app is unlocked and you’re good to go. If there’s no conversion. They should enter a drip campaign. The drip campaign should send push notifications if the app is installed to get to an app open to open up back to the paywall. If they signed up on the web and you collected their email, but they dropped off, they also enter this drip campaign. They get emails. Now, emails always go to the web paywall. So even if you download it via the app, if you don’t convert, you enter this drip campaign. Push notifications. Open up to the app. Paywall emails. Open up to the web. Paywall right. If there’s a trial start, you know, you can see what there’s a whole, you know, workflow of what you need to have, what you need to build for your drip campaigns. This is like very confusing for tons of people to set up. You have two options mainly. One is to use intercom, which is super expensive, but it pays for itself. If you do it right, I promise it pays for itself. The other option is you can use customer data to design flowcharts like this. And then you can integrate with Firebase for push notifications and segment for emails. And this is sort of the stack that we recommend. I love this scene. And soon. Yes, we’ll do it all soon. Besides, we’re not. We’re not masochists. Like you guys. That revenue somebody somebody somebody’s got to build the basement in the all. Someone’s got to do that. Yeah, Yeah. All right. That’s fantastic. And I feel like we could do a whole other webinar, just. Just talking through that diagram right there. So maybe we’ll have you back in the summer as it’s more fleshed out and maybe especially interesting to start, you know, between Super and revenue cap. We have a ton of data and we now have a data scientist and are working on data projects where we can share more. But it’d be super interesting to be able to look at some of those flows in like have some kind of range of percentage target, but then also look at the whole lifecycle, like some people are going to have lower trial starts, but How’s higher trial conversion? Some people are going to have high trial conversions, but then low retention. And so yeah, over time, I’d love to work with you on, on some data projects to kind of show that whole lifecycle and give some sense of what you should be shooting for in those ranges. Absolutely all right. Let’s go ahead. That’s something that’s something that the industry is missing. Like there’s so many amazing Indies who like open source their metrics, like like, you know, Curtis does this all the time. Zach is actually had does this all the time. It’s so much easier to run a race when you know where the finish line is. Right you know? Yeah Yeah. I’d love to work on that with you guys. Awesome well, let’s jump into questions then for sure. So we’ve got quite a few. We’ll try and get through them all. And I don’t know if you have a heart out, I can maybe take some of these questions in and go past the hour. If we need to. I can’t. You all right. So starting with Jake for the drip campaign, should we send both a push notification and an email if they have both enabled? Or should we prioritize push notifications over email? Both fitness. I used to send an email every single time you viewed the paywall. So every time you viewed the paywall, we would say like, hey, just realize your open finance. I like don’t subscribe there. We have a sale on our website. Check it out. We’re like, oh, wow, that’s this guy so nice. You know, why would he be messaging me about a sale? Right so that’s a great tip. Um, let’s see. I think I was ordering them in reverse order. Um, let’s see, I’ll just. I’ll just kind of get random now from Eddie, how valuable do you find email campaigns with web paywalls in terms of percentage? It’s higher than you’d think. Finishing, I guess like a 20% to 30% open rate on those emails and a few subscribe. But again, your trial to conversion on stripe is much higher. And obviously you take the whole pay rate. So it would for fitness say that accounted for. So like it’d be two it would be it was 2% to 3% of all people who receive these emails. Right so that’s people who haven’t subscribed. So our, our install two paid rate was like 4% to 5% So out of that 5% and an additional two or two, 2% out of the five. So it’s like 20% relatively of revenue. It’s 20% of revenue. If you if you figure it out, that’s great. You know, one thing, as a lot of people talk more and more about, you know, using Stripe and avoiding the Apple tax and everything like that. You know, I’ve talked to some developers who are like, oh, yeah, a paywall convert on the web converts amazingly well, like it’s way better than now. And then I, you know, I kind of follow up and ask more questions and more questions. And what you come to realize is, is that they’re sending like super high intent users there with an incentive up, like a discount or otherwise. And so it’s hard to compare apples to apples when you have your in-app paywall that doesn’t have that same level of intent and and you’re not offering a sale or whatever. These retention emails is a perfect place to send people to the web because you can offer the discount is higher. And if they actually open the email, you’re going to offer them a discount for going to the web. And so it’s like you stack all this intent on top of each other. So of course that, you know, those campaigns are going to perform really well if you run an ad and send people directly to the paywall versus, I mean, through the web versus running an ad and sending people into the app, that’s a whole different experiment that you need to A/B test in isolation against and not include the people that you’re sending back to the paywall on the web where you have all of this kind of internet lined up. Totally Yeah. So much of it is about understanding when the user’s intent for understanding that it’s temporary. And capitalizing on it. Yeah yeah, that’s fantastic. All right. Vlad asks, how do you track conversions since ad network, especially with free trials? That’s is that from. From from paid marketing? Yeah, I assume so. Yeah I mean, like, I don’t know if you’re in the weeds of this. I mean, feel free to, like, say you just, you know, are in the weeds in this anymore but it is sad network has been really tricky to work with from old folks I’ve been talking to. Are you are you in the weeds on that or. No, I’m not. I used to be pre. Yeah pre this whole thing. Well, vlad, stay tuned because we are going to do more webinars later this year on this kind of stuff because it’s something that does come up a lot. You know, I’m not working with it directly, personally, but I’ve talked to a lot of people who are and it’s challenging. So we’ll do another event specifically on that. All right. From Roddy. I tried adding a new monthly subscription to my app hidden to users. But there so that I could offer it in certain occasions. But Apple rejected my app because they couldn’t find this new in-app purchase. How can we appease apple? That’s a great question. You have to add it to your have to change your pricing to the monthly option. Well usually you can get an app approved, an Iep approved without pushing it into your app. Like you don’t need to push an app update to get an app approved. So I would find it pretty rare that they would reject it, especially because you need to get the product approved before you can make the paywall. That’s why they make you upload the screenshot. But I mean, finish it. We have maybe 10 to 15 products in our App Store Connect account. We never had issues with like updates and stuff. The answer here is probably just to keep trying. Like different reviewer. Yeah, Yeah. Submit it standalone. Yeah upload the screenshot of it being displayed in the app. And then, you know, and like we talked about earlier, like maybe there are some products you don’t want to display in the app, but maybe you experiment where it is shown in the app. So you’re not like lying to Apple that it’s ever going to be shown in the app. And maybe experiment with it in a retention wall or something like that. Totally all right. Omar asks, does the paywall pre-onboarding scare users away? That’s a fantastic question. If it does, just make your exit button more apparent, like you can have it right under your seats. Just say, like, not now, but again, you’ve got to test it. You know, not if your video is really captivating. Not if your paywall is an endless scroller and people can just keep scrolling and learning more about your app. It’s a fine line. You know, you got it. You got to just test. Yeah Yeah. That’s I think allowing an easy way out is probably the best thing to do if you’re aggressive with a pay wall. And again, in this conversation, I am convinced to be a little more aggressive with my paywalls. But that’s exactly what I’m about to do, is make sure that there’s an easy out. Generally, when you setting it, when you’re setting up experiments like this, you have to. The first thing you need to do is accept that you’re going to test some things that you’re uncomfortable with. But the reason you’re doing it is to you need to understand what your bounds are. So like an example is show the paywall before onboarding with no exit button. Right that’s the most extreme example. Right you’ll never have a higher bounce rate and use that as a baseline and then make your exit button super obvious and compare the two and try to see if you could find something that you’re comfortable with that leans more towards the number that you’re going for. Yeah all right. Ben asks, what are our best practices for paywall design for a subscription mobile app that has two pricing tiers, both with offering monthly and annual subscriptions. Oh, like, like a light and a pro or like a pro and make a Pro mode. I think he’s asking more what we talked about earlier, where you have multiple options shown at the same time. Is there a design for showing multiple at the same time that performs well versus just having the single call down? I like the other plans to show more. I also like the we actually haven’t run a test. I think if you have three options, horizontal picker does better. If you have two options, a vertical picker does better. And then also we did run a test on. You have two buttons. One of them is buy this product. The other one is buy this product. And then in the other version, you have a selector where you could choose one of the two and then press the bottom button. The one where it’s like a radio button and then you press the Check Out button that does better than just having to purchase buttons where you just click one to purchase that one. Does that make sense? Yeah yeah, that’s correct. It’s because there’s one button it’s easier to understand. They see the CTAs clear, right? All right. Roddy asked for apps that show a paywall before onboarding. Has there been a negative effect on App Store reviews or quick app deletion? I worry that people will see the paywall and then delete the app and leave a one star review without even trying the app. People don’t leave reviews like it’s very rare unless they’re furious, right? I’ve never run a paywall test that significantly affected the rating of the app and you know it. It’s your job to read the reviews. And yeah, if they are bad reviews, then don’t do it. Yeah again, nothing is set in stone. And then ideally, like I loved in. What was it like the third or fourth slide you said, you know, a lot of apps miss out on the last step, which is for retained users. You want to get them to share and you want to get them to review. And so imagine do you send an email follow up asking for reviews specifically? Not sort of. So this is the best way that I’ve found after they’ve fully experienced your app. So with fitness, say it’s after your fifth workout. An example would be after photo room. If you’ve used. If you’ve exported a photo three or four times, send the user an email. It’s pretty. It’s not intrusive. Ask them. First first off, ask them if they love your product. Then ask them how disappointed they’d be if they could no longer use your product. It’s like a famous product market fit survey that the founder of super of superhuman spearheaded and then made popular. And then here’s what you do. If if they say that they love your product, you should ask them, what do you love most about it? If they don’t like your product, ask them what they hate most about your product. Now, for the users who say that they love your product and they wrote feedback for you after the survey, show them what they wrote. And say, hey, ID mean the world to us if you pasted this into the App Store. And so you sort of trick them into writing a review without, you know, and now it’s like I just wrote this review. I may as well go play it on the App Store. That’s that’s great. And so Apple specifically has stuff in the rules against like filtering people before you send them to do a review, which I get, but I now take screenshots of apps that do this. And so many apps do this. Just in the last week, I took screenshots of They have this, like, Smiley face and a frowning face. And do you like our app, Yes or no? I click Yes. And I saw the native prompt. It’s like one of those things that like, yes, it’s a written rule. And like still in my apps, I don’t do the pre filtering, but like everybody does it. And I don’t know if Apple is just going to eventually change the rule or something like that, but like you’re obviously not enforcing it. And it’s a great way to get good reviews. I mean, it really does work. And then part of why I was asking about getting more reviews is that at some point, like with my way weather app, you know, it’s 4.5 stars, it’s got like 11,000 reviews now or somewhere in the close to 10,000 reviews. And once you kind of hit a critical mass of good reviews, it is really hard to bring that average back down. And so if you’re watching these. So if you make a change and you do get a few negative reviews, you want to watch that. But if it’s like three negative reviews over the course of months, it’s not going to impact your overall ratings, especially if you’re doing what you need. If you have a great product and you’re doing what you need to do to get good reviews and you have a critical mass of reviews, then you can experiment with being a little less worried about the impact of these negative reviews. Because I remember early on, it’s like when you only have like, you know, 500 reviews. Well, yeah, like 10 one star reviews can bring it down. I don’t know if a point can’t do the math on that in my head but but you know, it’s like any shift in that is, is going to hit you harder early on. But once you’ve reached a critical mass, you can experiment a little more freely and address the problems that come up and/or that are brought up in the reviews. But you don’t have to be quite like fearful of them. So totally. I think Apple should change it. And yeah, show a rating based on your last 10,000 because right now every review matters less than the one before it. Right the kind of weird, you know. Yeah no, I don’t know though, because I mean, yeah, I used to do that. So, you know, back in the early days of the. Oh they did time you submitted. Yeah Oh you submitted a new update. It would like wipe out your reviews and start over again. And that was, that was really tough. So yeah, it’s kind of changed over the time, over time. And that’s where I feel like in some ways a great product that amasses a large number of good reviews. Then, you know, ideally that’s what you build up. And so then I don’t like the Rolling ideas because then it’s more sensitive to these kind of experiments and it’s like, you know, like, like with my weather up app, we’re going to release a big update this summer and it’s some big changes to the app. And like, you know, those kind of things, I’m a little worried we’ll get some negative reviews, but with, you know, thousands of reviews and 4.5 stars like I know it’s not going to be enough to just, like, totally tank it. So so that really does help. All right. Brendan asks for Britney for a better ba ba ba ba ba productivity abs. Is it better to go with just annual pricing or both annual and monthly? And what about a higher one time pay option as well? Yeah I’m curious. We kind address some of this in as we were talking through it about, you know, offering the monthly just in settings and like, you know, there’s a lot of stuff we already did talk about, but you didn’t talk much about your thoughts on Lifetime option. I’ve really gone back and forth on this. What do you think about lifetime purchases? So I just put this link in the chat. It’s a book I’m in the middle of. It’s called priceless. It’s about, you know, thinking fast and slow. The book. So priceless is a book on. On how. How behavioral decision theory comes into play when deciding if something is worth it. So should I buy this computer or should I buy? Apple is famous for doing this so well. They have products that are perfectly priced in between specific models that just push you over the edge to buy the more, more expensive model. Right I think lifetime can serve a similar purpose. So, for example, if fitness Ii offered you lifetime access for $1,000. Or annual access for 49, 99 a year. People are like, oh, it’s obvious the annual is the move. You know, why would I pay $1,000 lifetime when it’s 50 bucks a year? But it also conveys that like, wow, I can’t believe that’s an option. Like, do people subscribe to that? Like, do people actually stick to this for 20 years? Like, that’s insane. And so it’s the thing. It’s kind of like you can have theories, but you don’t know. There’s no way to, like, systematically prove it. So you got to just test it. So test again. Set the bounds your extremes. So test 50 bucks a year versus $1,000 lifetime. And then test 50 bucks a year versus $60 lifetime. And you know. And you’ll see. You see, Yeah. The two things that I think a lot about with lifetime and have talked about recently is, one, sometimes you’re kind of putting your best users their like your highest intent users who may actually stay subscribed for 10 years, 20 years. You’re giving them this off ramp to actually pay you way less money, even though they’re the ones who are actually going to experience the most value. And then two, it’s like you could say, OK, I process it 3 times my lifetime value. But do you really even know your lifetime value at this point? Because over time we know that the longer somebody stays subscribed, the less likely they are to churn. And so you can calculate your LTV today, but maybe in three years, you realize so much of some of your early cohorts are still around and your LTV is actually two, three times as much. And so you’re really kind of shortchanging yourself when you’re building a product that’s going to be valuable for ten, 20 years. You know, we’re, you know, with subscription apps, we’re just now getting to that point of having, you know, 5 and 10 year cohorts with some of these very early subscription apps. But then you look at things like Netflix. I mean, you know, I subscribed when it was still DVDs in early 2000 and I’m still subscribed yet. I mean, it’s probably been close to 20 years. So I think that’s another thing I think a lot about is we don’t really even lifetime values. We we can calculate on a shorter time horizon what the lifetime values are. But as you stack those like retentive cohorts, your true lifetime value could be way higher than what you’re currently calculating it at. Totally people don’t realize that LTV is and it’s an average. Yeah it’s not a measure of central tendency. You don’t know where most of your users fall. So some people are generating three to four extra LTV. Some users are canceling after a month. And when you introduce a lifetime, like most people, they think they’re so smart and they like me. I’ve tried this and say, Oh my, my LTV is 102, it’s $120 bucks. I’ll add a lifetime offering for 150 bucks. How could my LTV go down? Well, the answer is the people who are subscribing to the 150 buck one time fee are the ones who would have stayed for five years. And so their LTV is much higher. Right? so a lot of times it brings your LTV down. Yeah is it more for price anchoring? Yeah, I feel like. So today we actually just released LTV charts in revenue cat, which we’re super excited about. Yeah yeah, you should go check it out. But I feel like LTV is something we could just have a two hour conversation on because like, you know, how do you. Like you have to set certain parameters to, to determine it. Predicting LTV is incredibly hard. The whole like subscription, what do you call it? The decay of churn is so fascinating. And understanding those curves is that the longer, the longer somebody stays subscribe, the long are they going to stay. Subscribe like so much of that is so fascinating. So we’ll move on, but we’ll put a pin in that and you and I need to chat more about LTV. It’s such a fascinating topic. Totally all right, Zack. Zack shared good friend of Jake’s and of revenue. How powerful is it to remove the free tier of an app and make the entire thing require a subscription? Why is that more valuable? Um, it depends. So if you’re freemium, we just had a customer who they do like 3 to 400k in revenue a month. They are they have a freemium model. So have their most their power features. You have to pay for the other features you get for free. They tested free trial versus not free trial. And no free trial did better. I believe. For apps up front. I mean, it depends on the category. You got to just test whether locking it out or free trial does better. Yeah, it just depends. But usually I think free trial does better, at least if you’re going to show it really early in the onboarding. Gotcha by the way, I did switch the questions to sort by upvote. So any questions you definitely want us to hit because we are going to need to wrap up fairly soon. So I’m just going to continue answering them in order of upvotes. All right. I’m not going to try to pronounce your name. I should probably just not use this because I always run into sound. But this is a great question. When do you know how you’ve gotten to a point where you’re getting diminishing returns from optimizing the paywall and that you should focus on core app value? Yeah balancing engineering resources. That’s a tough thing, especially like early on. It’s like there’s so much that needs to be done in the core product. And ultimately having a really great product is how you’re going to build your business. So how do you balance the scarce precious resources on this optimization versus like product optimization? The way I would the way I ran it is I would work on product, I’d release a new feature. I’d see if my retention changed. If it didn’t or if it did, then I would at some point, you know, you’re like, oh, well, are people even subscribing to even see this new feature? And then you sort of focus on the paywall and then the revenue goes up and then you work on product. And so it’s a balancing act. Yeah, it’s very it’s sprints sort of it’s not like it’s like mining for oil. Most times you find nothing. Sometimes you become a billionaire, you know, like, look at Blinkist. Eight years in eight years into blank as they found the free trial timeline, two years into literally I mean, I spent way more time than I should have optimizing our paywall. Took me three years to realize I started the wrong company, but like two years in, we found out that videos do better. And so. It’s hard to know, but at least devote some resources to it. Yeah cool. none of your business asks. Super wall seems like breaking Apple’s guidelines remotely changing paywall. A very regulated part of the app. Any comment? That’s a good question. It’s a sticky you know, some of these things that we’ve talked about are sticking points with apple, like Apple is sensitive to price discrimination. And like you’ve talked about offering discounts, which I think is in their view, like a good way to run your business. But I think where they get especially sensitive on price discrimination is, is like what Tinder did, like charging women less and charging men more. And like, that’s not limiting. Yeah discriminating versus on those kind of things. And then I think on the pricing as well, there’s also like as you grow, like Netflix would have a really hard time doing some of these more crazy price experiments because people will talk and know. But when you’re, you know, when you’re still doing 100 to 300k a month, like you’re not a household brand where it’s going to be a topic of discussion. And maybe that’s why you do it at that time and do some dramatic tests at that time, and then you kind of more. Settle in on things. But yeah, back to the original question, like for sure. Yeah What are your thoughts on being able to remotely configure this and staying within Apple’s guidelines? Because the one thing to note, they do go back and review apps now. So they look at live apps on the App store, post shop review. I don’t know what frequency, I don’t know what triggers a re review, but it’s been clear for a while now that they are reviewing apps. So if you’re changing your paywall dramatically, you might get caught. So that’s one thing to keep in mind. But yeah, I’ll let you answer the rest of that tech. For sure. Well, first of all, there’s like a question of if something is legal, is it ethical? If something is illegal, is it unethical? That’s not the case. You know, ethics is separate from legality. That’s why I think it’s ethical to ask a user who likes your app for review, even if it’s against Apple guidelines. Obviously, there’s a risk because you’re playing someone else’s game and they can, you know, throw you out. So there’s a dance there. And you have to be smart in terms of I mean, it’s the classic with great power comes great responsibility. People can do this anyway without super war. The New York Times paywall is a webview. If you check it out, you know, like they update it half the time on the New York times, they take you to their web, check out on their website like no one even knows. You know, like they don’t even know what they’re doing. Like someone who doesn’t even realize that, but whatever. So you have to be careful and you have to be. You can’t be an asshole. You know, you can’t circumvent Apple. You’ll get caught. I think the key and what you just said and this should just apply to all things because like you said, it’s like I mean, you know, you could and, you know, swizzle, swizzle private APIs you can’t like there’s so many things that you can do that are against Apple’s rules that you can get away with. And a lot of apps do. The key is the spirit of the law versus the letter of the law. Right like, why does Apple say don’t change your paywall? Because people do things with their paywall that are against the rules and are dark patterns and are tricking people into subscriptions. So if you are changing your paywall remotely, I don’t think that’s technically against the rules to change your paywall remotely. Remote code execution is not allowed. So like right. Remotely yeah, but like remotely configuring and changing. Like, yeah, they had to make peace with aspects of that. So but the point being is like if you’re going to do these experiments, just don’t do don’t do an experiment that uses a dark pattern. Like don’t like continue following the rules and experiment within the rules. And if Apple catches you doing something that is a dark pattern, you know, you could face consequences. You could be pulled from the App Store or whatever. So, see, I think that’s the best way to frame it is. Yeah, exactly. It’s like you play within the rules ethically, even if it’s not always perfectly to the letter of the law. Because so much of this stuff, again, it’s like I mean, you know,, a multi-billion dollar business is doing something that’s like against the letter of the law. And and, you know, Apple doesn’t seem to be policing it. And, like, I don’t know, this is such a tricky subject. I think it’s like you have to decide with your own app, like how you’re going to run it and then what risks you’re willing to take against that. Breaking Apple’s rules over. You used to be illegal. Yeah getting into an Uber was illegal. Yeah so anyways. Yeah, I mean, you know, got to make your own decisions there. But I think, you know, just be ethical, like keep do what’s best for users. And, and that’s what I’ve seen over and over again is like if, if it comes up and you’re actually doing something good for business, good for users like Apple’s not going to hassle you if you’re doing something that’s a dark kind of, yeah, they’re going to hassle you. So like, you know, that’s where if you draw your line there, that’s the most important thing. And by the way, fitness, I once pushed a paywall that Apple didn’t like that made it through a review. And we went through that review process. They just messaged me and said, the paywall isn’t compliant. You have to change it, right? I uploaded a new builds. They said they still didn’t like it. I uploaded it again and they liked it and that was fine. So it’s not like. Yeah and that’s putting to app jail for this. Besides fortnite, of course. Yeah and that’s a good point, too. Is that what you can do is, is submit various paywalls with new uploads. Like if you are experimenting with your paywalls, submit some of your variations as the active paywall when you submit it and you’ll get Apple’s feedback on it and they say it’s not OK, then stop using that as an experiment. All right. Next question. Do you see a difference in optimizing for iOS versus android, perhaps on different characteristics of the user base? We don’t have Android client libraries yet, so I couldn’t tell you. OK introduce really tricky. I’ve talked to a lot of folks about Android. I mean, you know, we see it with revenue cap. We you know, so many of our customers say it it’s in most apps make the majority of their money on iOS. Even Curtis Herbert we had him on the sub club podcast recently and, and he talked about he just recently on Twitter, you know, he’s put a ton of money into Android and gets all these complaints of like, oh, it’s not, there’s not feature parity or whatever. Well, like, he hasn’t even recouped a quarter of the cost to just get it to where it is. It’s and so Yeah. And Android’s a tough, a tough one. So, yeah, I wish we had a better answer for you, but yeah, just, just know that whatever you do, you know, just the demographics are Android are different. The willingness to pay is different, the expectations are different. Probably even where you should price is different on Android. But it also kind of depends like, you know, Android in the US is a different demographic than like Android in Spain. It’s like, you know, in Spain it’s way more common to be on Android and see. So the like the, the financial demographics and other demographics of the people in certain countries on android, they’re going to look different than the US. And so if you’re talking just about the us, that’s going to look really different than a lot of European countries than developing countries and otherwise. So there’s probably not an easy answer to this anyways, but I think a good way to think about the two users are like if you were to compare an iPhone to an Android phone only on paper. Right so looking at price, screen size, what actual features there are Android wins in my opinion. Like yet everyone in this country chooses an iPhone at least most of the money that the spenders. And so why would they buy an iPhone if on paper it looks worse? And I think it’s because they’re. They care about, like their emotional decision makers. I think it’s because they want the shiny, nice Apple products that works well. And they don’t care if it doesn’t have these individual features. People on android, they do care about those individual features. And so for sure, the pitch. The pitch changes, but it’s definitely easier to sell to that category of people who don’t care about individual features. These are people who don’t care about the thread, count on their sheets. You know, they just want you know, that’s like those are iPhone users. Yeah, that’s a great point is when you do start to do these optimizations on Android versus os, don’t just automatically apply your iOS learnings to Android. You really want to probably message differently, highlight different things again, probably even priced differently. It’s just it’s just fundamentally a different market. And then you probably even want to do that. I mean, you know, when you’re small, you’re not going to have the resources to do this. But ideally, as you grow, you could also do some of that differently country by country, depending on the demographics of Android users versus iOS users. It comes down to just understanding your users. All right. So I have a bunch of questions. We’ll wrap up with maybe two or three more questions. If there’s one in there that you really want to see answered, go ahead and upload it now. So let’s see. How do you deal with current subscribers when you increase the subscription price, keep the same price with them or keep them happy by making them? Pay the new price. Well, the way that you do that I the same price to keep them happy. Yeah the grandfathering. Yeah I don’t ever edit the price of a subscription product I have tons when I’m running a price test I have tons of different groups, each one of them with only one product inside of it. And if a new one wins, I just show that product. But people on the old products continue to pay that way. And it’s the same product ID it’s too hard to match to keep track of revenue that has specific products generated over time when you’re adding the variable that the price of the product can actually change. So better to create a new product with a new price and do it that way. So and if you go that route, they’ll always be grandfathered in because you’re never changing the price for a specific product. And then I’ll expand on that answer. Currently, if you raise the price ID, it is an automatic opt out for anybody raise the price on. So I would always advise to grandfathered people in because you’re just creating automatic churn right from your user base because they have to open the app, they have to see the price increase after agree to the price increase in order to continue the subscription. But Apple is piloting a program with Disney Plus. And, you know, maybe they’ll roll this out to more developers. I don’t know if they’ll do it just for like, you know, streaming platforms or, you know, have certain requirements or whatever. But hopefully they’re going to roll this out beyond just Disney. And so with this new flow, the users are automatically opted in to the price increase. And all they see is like if you search on techcrunch, search like apple, Disney price increase. Sarah Perez did a post on this and got Apple to confirm and I helped her get some of the data. Is that on the Disney there’s just an OK button and subscribers are automatically opted into the new price. So if Apple rolls that out more broadly, I think that is an opportunity to raise your prices in a reasonable way if you’re, like, doubling your price. You know, like Jake’s talking about some pretty big price experimentations test $30 and test $60. Well, you do have to think about the experience of users. If you say your $30 subscription is now 60 and one, I don’t think Apple is going to allow that. So, you know, you wouldn’t want to do those kind of price increases. It’s a bad user experience, but if it’s like you’re at 30 and this year, you’re going to increase it to 35 or something like that or, you know, 60, and you go to 80, you do risk churn by doing that and you risk, you know, annoying people. But at some point, if you’re truly delivering value and you’re convinced that it’s worth it and that people will see the value. I’m hoping Apple rolls this out and we’re able to kind of experiment with that more. But it will be it’ll be tricky. And it’s like it’s like this. This will be like a whole new area of learning. Just like we spent an hour talking about paywalls. Like price increases will be tricky and some apps will see a huge amount of churn. So for now, if you’re going to do a price increase, just grandfathered users or create separate SKUs. If Apple rolls this out to everyone, I would even way let other people kind of take the first and see what the best practices are, see how big of a price increase Apple’s going to allow, you know, figure out is that worth it? You know, look at your retention, make sure that you think people are going to stay subscribed. So, yeah, a lot to think about there. But for now, I would just grandfathered people in. Totally all right. Let’s get to more questions. I have an app with high installs and retention, but low subscriptions. How risky is it to paywall features that are currently free? That’s a great question. The greatest risk is to take no risk at all. So you’ve got to test it. Test it? That’s a great answer. Yeah, you got to just test it. That’s an easy one, too. Oh, this will be an easy kind of technical question. Is Flutter are you Flutter compatible with Super Bowl revenue? Has a Flutter sdk? So yes, but super. Well, not yet. Not yet, no. Our biggest request is react native, but only for iOS. So we’re going to release that soon. And then we’ll do the same for Flutter. Awesome that was an easy one. All right, one more. Let me make sure I’m getting the highest upvoted one and then we’ll wrap up. OK I found it quite difficult to find retention numbers. What retention should you aim for? D, one, d, 7 and 30? Oh, this is a great question. Mostly most shitty apps. I started from offending anyone. They do around 15% to 20% year one retention. I don’t I don’t know. Do you want 70/30. And I’m speaking about paying retention, by the way. Yeah, that’s a tricky thing. And a lot of people don’t clarify. So, you know, d1d 73d is probably usage retention. But I think for subscription apps that kind of, you know, day one, retention is probably less meaningful because you’re kind of intentionally filtering to the people who are willing to pay. And so looking at usage and a lot of the benchmarks that you’ll see on the web for d1d7d 30 are from like games and other apps like that. Yeah, you’re usage retention is the KPI. It’s like the metric where subscription apps and what you’re about to talk to is that subscription retention really should be more of your KPI than usage retention. An example I love on this is I use VS the VS app and I only use it when I take photos I really care about. So like when I’m on vacation, when like, you know, doing something fun with the kids or whatever. So I don’t necessarily even use it every month, but I am like zero churn risk because I love the app. I take pictures I care about, I open this, go edit, I love their filters, you know, so, so that’s where subscriber retention isn’t necessarily even indicated in your usage retention metrics and you need to understand the difference between your usage metrics and your subscription retention metrics and look for those patterns like me that aren’t necessarily heavy users. You don’t necessarily need to incentivize heavy usage to in order to drive retention, you need to increase value to drive retention. So now flip to you and let’s talk about some benchmarks for subscription retention, which is what I think we should be more focused on, less subscription apps for sure. Well, Yeah. First off, I mean, look at gyms. Gyms only around 20% of their monthly paying customers use the subscription. So only 20% of people who have a gym membership use the gym. So that’s definitely the case. Well, of course they’re correlated, though. But then another thing is iOS makes it so easy to both subscribe to an app and unsubscribe to an app. So like I fall in the category of I just unsubscribe from stuff when I’m not using it. Like, why not? I haven’t used it in a month. I unsubscribe because I know when I open the app again, I’ll probably either get a discount or I just resubscribe again. Yeah so why not? So that’s for starters. But I know. I think maybe you guys. Yeah revenue would be much better to answer the subscription retention than I would. But I’ve seen 15% to 20% on the low end. I’ve seen 30% or 40% from really good apps. And I like 40, 50, 60% is like so good that it’s almost unheard of. Yeah and not much different from credit card retention. Oh, like, for example, our, our stripe. So all things constant, our subscriber retention on stripe for fitness guy is like 60% year over year and our annual retention is 40 is, is, is 30% year over year. Wow and it’s easier to cancel, right, if you subscribe on stripe because in settings we show you a manager subscription. And you can just cancel within the app. So I can’t explain why. Well, I mean, it goes back to what we were talking about earlier potentially, is that you’re driving some of the more high end tech users and you’re offering, like the people on the stripe subscription probably have a bigger discount. They’re scared they’re going to lose the discount and have to pay full price. Like there’s probably some kind of like. Built in retention hooks. And that’s why it’s a good idea to go ahead and send your higher intent users with a discount to the web because you’re going to create that kind of lock in that they don’t want to lose this great deal that they got. And maybe they are your higher content users anyway. But then to answer this and another question and we’ll wrap up with this. We we are going to release benchmarks at some point in 2022. It’s something I’ve wanted to do since I joined Remnick in 2019. It’s tricky, though, and one of the things I’m thinking a lot about is how we can actually display lifecycle and benchmarks versus individual benchmarks. And we’ve kind of talked about this through the whole conversation today. Is that just because you have a high install to trial store, it doesn’t necessarily follow that you’re going to have a high trial start to trial conversion, which doesn’t necessarily follow leads. You’re going to have a high trial conversion to retention. And so there’s like this entire user lifecycle from opening the app to, you know, like year five retention. And what I anticipate seeing in the data and hope to find a way to visualize is that even great apps are going to look different. So like maybe the average trial start rate is 8%, but you could have a fantastic app where the trial story, it’s only five, but then they’re there, conversion rates like 90% and then their retention is incredible. And so if you over Yeah if you want and if you only look at each of these kind of lifestyle metrics in isolation, even if you’re looking at averages, you’re not understanding the swing. So like if you’re shooting for like I got to get a 10% trial start rate or I’m not in the 50th percentile. And then I need to get a 50% trial conversion rate to be in the 50 percentile. Like if, if you set these individual goals in isolation as your target, you’re not necessarily hitting your own kind of global maxima because you might actually do better. Like, like Jake said earlier, like locking the whole app. So now you’re trial start rates are going to go way up, but then do your trial conversion rates, do your retention and everything else balance out. And so look at each of these individually is probably a fool’s errand compared to really trying to understand the full lifecycle. And that’s what I mean. Now I’m setting a super high bar for us, but that’s what I hope for us to do when we are able to release these benchmarks, is to try and have a better understanding of the full lifecycle versus just benchmarking these in isolation. And then to complicate that even further, you know, category by category, country by country, those things are going to look different as well. And so if you’re looking at a globally blended peak, that’s going to look really different than if you look at just the US or if you look at just, you know, other countries specifically. So it is if you see benchmarks out there taken with a huge grain of salt, because there’s a lot of confounding variables. And I hope that we can produce some benchmarks that at least give context around those confounding variables and try and help you understand the full lifecycle instead of just throwing random benchmarks out for each of these individually. So long answer, but hopefully we’ll will be able to do something meaningful in that area in 2022. Last thing to note, we still have quite a few people here. I do run a private forum for subscription app developers called sub club Jake stone AMS, and they’re just actually completed one last month. So trying to figure out I need to next time we do the webinar, I’ll figure out how to drop a link in earlier in the conversation. But hit me up, David, at revenue and I’ll show you a link to the waitlist. Somebody at revenue hopefully heard me and dropped the link. And if they had a handy but otherwise. Yep there we go. Corey just dropped it in the waitlist. Thank you, Corey. Yeah, this was fantastic, Jake. Such a so knowledgeable on all this stuff. And again, like, I mean, you really change my mind on some things today and I think this was just an incredible exploration of better monetizing. So thank you for your time and I look forward to continue working together. My pleasure. Thanks for having me. And Thanks for all the great questions, everybody. Thank you very much. All right. Bye, everybody. Take care.

Subscribe to newsletters

Watch more

WWDC 2024 Recap
June 13, 2024

WWDC 2024 Recap Webinar

Our team discusses the latest announcements from this year's Worldwide Developers Conference.

RevenueCat Launch Party - April 2024
April 30, 2024

RevenueCat Launch Party #2 – April 2024

Join us as for our second ever Launch Party - a live webinar designed to keep you in the loop with RevenueCat product enhancements and new feature rollouts.

The State of Subscription Apps 2024: Live Roundtable
March 27, 2024

The State of Subscription Apps 2024: Live Roundtable

What does the future look like for subscription apps? Join our roundtable as we dive into the 2024 State of Subscription Apps report, featuring the most comprehensive dataset on subscription app performance globally.

Want to see how RevenueCat can help?

RevenueCat enables us to have one single source of truth for subscriptions and revenue data.

Olivier Lemarié, PhotoroomOlivier Lemarié, Photoroom
Read Case Study