30 Jul 2018  |  Guide

Moving to Subscription: our WWDC 2018 Fireside Chat

38 minute read

Many software companies ask themselves the same question: should I move my product to subscription? And how could I do that? Many indie developers have moved to subscriptions, or have been thinking of doing so - but it’s neither an easy decision, nor an easy rollout. We assembled a panel made of the founders of Sketch, Ulysses, MacPaw, Flixel and Readdle to learn about their experience.

The Paddle team and I recently went to WWDC, Apple’s annual WorldWide Developer Conference in San Jose, California to catch up with the thousands of independent developers present.

One challenge kept being repeated: should I move to subscriptions? And how could I do that? Many indie developers have moved to subscriptions, or have been thinking of doing so - but it’s neither an easy decision, nor an easy rollout.

After chatting with those we knew had gone through this - some our sellers, some not - we took advantage of Altconf and assembled a panel of some of the smartest minds in the space with the founders of Sketch, Ulysses, MacPaw, Flixel and Readdle, with both strong opinions and battle-hardened experience of subscriptions, to answer 2 main questions:

  1. Should your product use recurring billing, and what are the criteria to decide?
  2. If yes, how do you figure out the pricing, what are the mistakes to avoid and do you roll subscriptions out as smoothly as possible?

Key Insights

Some of the key insights we heard:

The decision is absolutely linked to your product, your market, your types of customers… Each panelist had a different approach linked to their unique situation. B2B or B2C, daily use or sporadic intense use, highly differentiated or commodised… All of these factors play a role.

There are many reasons to want to move from a perpetual license to a subscription model, among which:

  • The sustainable cash flow.
  • A greater simplicity in access management (as opposed to handling which features are or aren’t accessible which can become a huge engineering nightmare).
  • Being able to ship features faster and focus on what makes sense, rather than bundling everything and waiting a year to have a compelling upgrade you can sell.

But not everything should be a subscription - especially with subscription fatigue kicking in, the syndrome where everything’s a subscription:

  • Offering recurring value is very important, as a daily use app lends itself really well to a recurring billing.
  • But subscriptions can also work well when there’s a high infrequency in usage and customers may not be convinced instantly, or would not pay a high one-off fee. This offers greater flexibility (pay when you use it, cancel when you don’t, renew when you do again).
  • Professional apps are much easier to offer as a subscription business (hence the rise of SaaS), as the value they provide is much clearer than to a consumer.

The single biggest way to grow your revenue is often to increase pricing. Not better marketing, or nicer features. But by constantly A/B testing by channel, geographically; by increasing the price until you start seeing a significant decrease in conversion. As opposed to lowering the price until you hit the floor and can’t go any lower.

Your revenue will drop significantly when you switch to subscriptions, as people who used to pay everything up-front start paying each month or year. It can take 18 months to get back to the revenue you had previously - although you will be in a much better state financially in the long run. There are ways to switch progressively, by offering licenses in parallel and progressively reducing their exposure until 99% is subscriptions - although that takes more work.

Communication to your existing customers is key: don’t spend too much time explaining why it’s important to you, focus on what that means for them. There was a big debate about whether people should be told they are paying for constant feature updates, or for the product they buy as is. Offer something special to your existing customers who’ve just paid a one-off price.

To read further on the topic, see how Ulysses, Sketch and MacPaw wrote about their move to subscription:

On our side, we have written extensively about determining the right billing model based on our experience accompanying hundreds of software companies as they transition to a different billing model, or iterate on their billing and pricing.

We’ve made it easy for Flixel to move to subscriptions and MacPaw to A/B test their billing model and pricing in any geography or channel they like. If you’d also like to spend your time growing, not billing with our checkout, subscriptions, taxes, licensing, and insights in one unified platform, just get in touch 😃.


Full Panel Transcription

Introductions

I’m Christian, founder and CEO of Paddle - we’re building the platform for software businesses to run and grow, and helping developers focus on building rather than billing with checkout, subscriptions, taxes, licensing, and insights in one unified platform.

My name is Pieter Omvlee, I’m the founder of Sketch. Sketch is a Mac app for growing user interfaces.

I’m Max, cofounder of Ulysses, writing app for Mac and iPhone.

My name is Oleksandr Kosovan, I’m the founder of MacPaw, developer of CleanMyMac, Gemini, Setapp and many other products.

My name is Mark Pavlidis, cofounder and CTO at Flixel who makes Cinemagraph Pro for Mac and iOS and we’ve got a few other apps as well.

My name is Denys Zhadanov, VP at Readdle. You may have heard of us through our quite popular apps like Spark, Scanner Pro, Documents. We’ve been on the App Store since the first day, witnessed the evolution so we’ve got quite a story to share.

Panel

Christian I think quite a few people in this room have been transitioning or considering transitioning to subscriptions for a long time, either in the App Store or outside of the App Store. I think it’s a tough thing to decide on and a harder thing to roll out and communicate to customers. The 2 main things we’re going to touch on today is what kind of application is a subscription model right for, what applications is it right for and which ones perhaps not so right; and if it is the right model, how do you price it, how do you go about communicating it to customers, and grow a sustainable software business on subscriptions.

It’d be good to hear from Mark. They’ve been through basically every billing model, with Cinemagraph, starting with perpetual licensing and progressively moving towards subscriptions. What was the real catalyst to start making that move and what were you monitoring throughout the process?

Mark The original motivation was because the month on month variance in revenue was very high. It wasn’t very consistent. So this was obviously a problem. If you want to be doing any kind of marketing, advertising activities you need to have sustained revenue to know how to predict. We started before you could do it through the app, it was strictly web services.

Christian Max, I think everybody has read the blog post - the novella. Can you take us thought that experience and how you guys came to the decision?

Max It’s similar to what we just heard and there are lots of technical problems if you’re on different platforms - Mac and iOS, users come and ask “why do I have to pay for 2 apps, even if it’s just one thing”. Then there are additional problems with 2 platforms, like people update one of those and have mismatch of features or incompatibility with the other.

Christian Obviously with Sketch it’s been an interesting road from perpetual to this hybrid model in-between. Is there a reluctance to go to subscriptions and why did you settle on this? I know there’s a handful of apps now that have adopted this hybrid model.

Peter First, our motivation to go to subscriptions wasn’t the same as what we heard here. We’re operating within a very competitive market and found ourselves struggling with figuring out which features we were going to build and when. We were coming up with version 4 of Sketch and wanted to ship some high impact features that were going to entice customers and attract them to buy the new version. There were so many big features that we wanted to build, so you start thinking that this big feature can’t go into version 4, so that’s version 5. But if you have to start working on version 4 then version 5 is at least 2-3 years out. And in design, a competitive market, we didn’t want to work like that, we wanted to build features whenever we decided to build them. So we decided that if we stopped going minor / major version completely we don’t have to think about how we’re going to entice users. So if we use a subscription model then we can build the features that make sense. The model that we have is subscriptions in the way that you have magazine subscription: if you cancel your subscription, nobody comes to your house and takes the magazines you’re already received, whereas if you have a subscription with Netlix, if you stop you lose the full access. We felt that the first model worked much better for us because our users create very important documents in Sketch. If they stop paying and stop getting access to these documents, it almost feels like you’re holding them hostage. Of course we need recurring revenue to develop new features, hence that model.

Christian Denys, Readdle has several apps - 7 - some of which you have chosen to go completely free, freemium, perpetual license or subscription. What types of apps lend themselves well to each model. I agree that at the end of they day we are all building software for customers. Yes we have to build a sustainable business but we don’t want to hold them hostage, especially with their data and their files.

Denys I was just going to say that we as developers, many of us think to switch to subscriptions. They write blog posts why, from the perspective of being developers. I think that’s a bit wrong and I think we have to think about the customers. Because our goal is to build great software for the customers. So they can use it and be happy about that and pay for that. And that highly depends on the type of product or service. If you’re using Sketch for work or use Photoshop, you’ll pay for this because that’s the place where you work. Depending on the retention, the usage, and how you use it - depending on the tool, for example if you look at our app Scanner Pro: you wouldn’t use this every day, you’d use it once a week or once a month so there would be no good use for subscriptions there. So I think it really depends on the nature of the product.

Christian How do you balance that with the need to build a software company, where you balance that with the need to pay people and ship updates?

Denys We have 135 people in the company. We have been lucky to be building products that people love buying. PDF Expert is still a revenue driver for the whole company. Right now it’s a one-time purchase but we are exploring ways to add subscriptions onto that product. With Spark, the email client, that’s the place where you spend 2-3h a day, and adding collaborative features where people can share and work on emails together and discuss, we feel that this is enough value that we provide to justify a recurring revenue. So you need to provide recurring value so that you can charge a recurring revenue.

Mark To comment on that a bit further, it also depends on the use case. Some apps you use them every day, like Sketch and Spark. For our app, someone might use it once a month, once every 3 months. So it’s not something you use all the time. So asking people to pay upfront with a big one-time license fee was difficult. With our product, people thought “this is new, I don’t know what this is” so people needed to try, they want to explore it. Sure you can a nice free trial and a watermarked output if you want to get your hands around it but there’s still a huge barrier if there was a high one-time price point. So what does that mean? It either means we have to drop the price low to allow people to try it out. But then you drop the price low, which in our case because we’re a niche product, means that our price is not sustainable. So you need to have that higher. So you lose a part of the market because they’re not willing to do that. So that’s where having a subscription model, at least you can change that pricing because you can get that lifetime value over a longer period of time, as well as introduce monthly pricing or different types of trials where some of these customers can try that out for a brief period of time and maybe it works for them, great, they’ll continue on for a longer period of time. Others won’t. At least that way you get to realize some of that revenue over a couple of months, from those customers who may not be long term customers but would never have paid for a high upfront license.

Christian On pricing, are you optimising for that lifetime value with that knowledge that for some customers, it’s a utility but a seldom used utility?

Mark There is a chunk of our customers that are win-back. They subscribed when we had a monthly pricing, they subscribe for a couple of months, they’ll cancel, and 6 months down the road they’ll subscribe again for a little bit. It’s the nature of the product: I’ve got a project to do, I’ll subscribe to it and then come back. Those customers are great because the follow-up acquisition cost for them is zero. They’re already familiar with it. They’ve already paid. They just don’t need it right now and it gives them the options to pay when they need it as they move forward.

Christian Oleksander, running Setapp which has 117 apps - for people who don’t know Setapp is a subscription service for Mac software: pay once and get access to all of these products. You probably better than most have some interesting insight into what types of products are lending themselves really well to subscriptions - is this the utilities, is that the use every day, at what point does it make sense for these businesses to run a subscription model through Setapp or a similar service versus running their own subscription model and where is the balance between those two things?

Oleksandr First of all our vision was is to deliver continuous incremental updates to the customer because when you build some great features you want customers to have it right away. For example in our previous business model we had major upgrades once per three years so we had to deliver some bug fixes and keep a lot of great functionality for the major upgrade to justify the value of the upgrade. This is where one subscription works best for a lot of products because if you want to continuously develop the product the business model should support that as well. This was the main vision behind Setapp. So we wanted to allow a subscription model to be accessible to a lot of developers. So if you do whatever kind of app you do - it can be one purpose application or everyday use app it can benefit from one subscription. You don’t have to think about the churn rate in the subscription model because this is what the Setapp service takes care of. I just have to care that you deliver continuous value to your customers and you have the good retention rate to your app.

Christian It actually feels sort of a little bit more like a meritocracy: the best and highest quality apps who are delivering the most value on a consistent basis could be the ones that drive revenue for themselves through alternate platforms like this. I think one of the interesting side points on this is around pricing. What we’ve seen and what a few of you have written about over the last few years on the willingness to pay for software, especially at higher price points diminishing significantly over time. Do you think that we’re gonna get to a world where more and more of these products are moving to subscription and we’re gonna have a similar problem, where people can get a subscription fatigue? Do you think there is a need for this sort of more consolidated subscription? Do you think it’s gonna be moving to these very utility everyday apps living more and more in a subscription-based world? It’d be interesting to hear how you guys feel about that and trying to develop a sustainable business with price points going down.

Denys So on pricing. There was a big trend since we’ve launched our first app, back in 2008 on the very first day that of the App Store. The price was $15 and then over the years we saw that developers were competing and then there was this race at the bottom in terms of pricing: you’d then decrease to $15, $10, $5, $3. But back in 2015 we actually saw a trend where the prices were going back up. It was a trend all on the App Store. So I think people are ok paying for great services and apps to use. I also think the big trend is there in our productivity space. There is a major change in terms of the conglomeration if you like, the big companies they become bigger. Microsoft acquired Sunrise the calendar app. They acquire it Acompli, which is now Outlook and then there is Google Suite - all of the products are actually free to use or e.g $5 to use. They also have a different business model where they make money through advertising and getting your data. So you as an independent developer can’t really compete in terms of pricing directly so you have to come up with a different business model or a different product that people will actually go and buy.

Mark We’ve been bucking that trend in terms of instead of going down, going up. We made a few early mistakes of decreasing the price. We started off with a high price point, we got featured on the App Store and won an Apple Design Award 4 years ago. We thought “Oh, lots of people are looking at this, let’s make it more accessible and drop the price”. So yeah that turned into a whole bunch of revenue for those two weeks and that was it. But what it also did, which still comes back to haunt us, is that if you search for our app within a few hits an article comes up on the first page where it’s like hey, this app is $15. No, no, it’s not but people still have that belief that it is so we’ve completely devalued the product then by doing that by dropping that price. That was in retrospect a bad idea, even though it was a short-term windfall of dropping the price. Over time we’ve just continuously increased the price and the single biggest thing to lift our revenue has been increasing the price. All the other kind of marketing activities have helped but the single biggest has been increasing our price.

Denys What’s a fair price?

Mark Well, I guess that it’s one of those things that you got to figure out but I guess in our case the elasticity of demand is critical so it isn’t something where you need to compete on price if you’re providing value to someone. Our use case for example is that we make it really easy and simple to create a cinemagraph in a matter of seconds instead of taking hours or days in Photoshop. So for people whose time is money, there’s a very simple value proposition: “this makes it really simple for me and really easy and speeds up my workflow. I see value there I’m willing to pay for that - how much, I don’t know”. Maybe it’s even more than we’re charging yet. But if you’re just going down, there’s a floor that you can’t go down any further. You’re gonna hit a point. At least if you’re going up, if you keep increasing your prices and revenue keeps going up it means on the demand curve, there’s a point there. That’s the maximum and if you keep raising prices and revenue goes up it means you’re going up that curve. At some point you go over the top and you’re going to start decreasing revenue. The easy fix there is you just roll it back to whatever price point you were. Where you want to be is at the maximum of that curve. So the only way to figure that out is by going up and figuring out where that point is. Then you can always scale that back.

Pieter We started with a similar story. We launched the first version of Sketch version 2 which was the first to get slightly more popular. We were in contact with Apple who said “well, we think you can increase your revenue, if you drop the price and open it up to a wider audience”. So we tried that for a month and that didn’t work. We went from $49 to $39. Ever since then we’ve gone up and we’re now at $99 for our license. Like you said it’s a professional app that people use together to get their work done. So $99 is not that much to ask. We’ve not seen this trend of prices going down at all, or the people being unwilling to pay. We also felt like we couldn’t raise the price higher because while there are some people who would be happy to pay more. We have a long tail of users whose needs are not slightly the same or not that high and for them $99 seems to be the tipping point. We tried higher prices and it didn’t work and we’ve rolled back so we think we found that maximum.

Oleksandr We have a bit of a different approach to pricing. We believe that you never know unless you try. So the pricing needs to be tested. We do a lot of tests, even on different markets. Since we are sold via Paddle it allows us to get a lot of flexibility to sell on different prices even on different countries. From time to time we test different prices and try to optimize the maximum revenue on certain market or even the channel. So if the customers come in from certain channels, like a marketplace or another channel the price could be different for them. It can be discounted for students or some other sources. For example even in Europe we have several different prices, e.g for Germany and for Italy.

Denys So Germans are ready to pay more than Italians?

Oleksandr Yeah :). I think you need to test the price continuously and even when you develop the app further, you add more functionality to get more brand awareness: you can try even in one year to increase the price and see what effect this will have on your revenue. It’s okay to try.

Christian I think Max when you moved from this perpetual license model to subscriptions, there was a dramatic difference in price. How has that gone?

Max Software has always been too cheap. In the recent years at least it was always too cheap and we’re now finally getting back to decent prices. When we switch to subscription, yeah, it was a price increase. We went from I think if you got the iPhone and the iPad app for $70 and the Mac app together for $70. We went to 40 dollars a year for a subscription. I mean you can take the paid one price and calculate how long will a user use it, 1 year 2 year 3 years and you end up with a price. But whatever you do, you probably end up below $40 a year. So that was the increase and I think it’s important to communicate with the users and say “well, here’s the issue. I need to continuously work on the product. You want continuous updates and I provide continuous value. So I need to make a price that just fits that situation”. You need to get them into to the mindset, I think that’s the most important thing about subscriptions, is that people understand that whatever you do is continuous work and providing continuous value and that the only thing that can match that is continuous payment.

Christian One of the things that has come up a few times is about actually making a switch from being a one-time perpetual license app to a subscriptions app and the quite significant change it means in terms of how you would run a software business. Going from having yearly, every two years, every three years these massive massive spikes in revenue, to a much more steady month-to-month (hopefully increased) revenue. Is there anything anyone can share about how they’ve navigated that, how it’s been on their business and what people need to watch out for in terms of planning?

Max So to get us started on the subscription, and to reward regular users, we offered them a perpetual discount. The regular annual price is $40, then for someone who has bought any of our products previously the price is $30. This is perpetual so they’re saving infinite amounts in unit number of years and that really helped us. Making them an offer really gave us a big spike of transition users on day one and the coming weeks, and that really helped us get into subscriptions. Everybody that does subscription knows that there is a stretch, there’s a long ramp of subscription and it’s still there. But having a lot of users and having access to them by a mailing list and within the app itself really helped us get started. I think if we hadn’t done an upgrade pricing we would have had more struggles and more frustrated users.

Denys When you moved from one-time purchase to subscriptions, did you see the increase of revenue or do you see less conversions, how many users actually stayed? That’s what’s interesting for I guess everybody and for us as well, as we’ve never done it before.

Max So the thing is the transition is not yet done. For users that bought just before the switch, we offered them free transition periods up to one and a half years. And the switch is only nine months ago, so we still have nine months to go. But what you will see is a revenue drop, definitely. That just comes from the fact that someone buying it for $70 is now someone subscribing for $5 a month: that first payment is way way less and over time the user will pay the same amount but not in that month and not in the next month. So it has to build up and even if they subscribe yearly for $40 or $30, then it still needs two or three years for you to get back to that same amount. So you really have a drop in revenue: it’s like at least 50%, at least. But but it grows over time and I hope that we’ll have the break-even by the end of the year, which means one and a half years of worse numbers than before.

Mark When we introduced subscriptions, we did it side by side with licenses and we left it that way for about two years. Even the UI was very much side by side. You could choose a license or you could choose a subscription and you had the choice side by side. Over the course of about six months we went from $0 in subscriptions to about 50/50 and it stayed about 50/50 all the way through. So it gave people those options. We dealt with that sudden drop of revenue that you’re going to get, especially if you have a monthly pricing, and were able to bridge that by having the licenses. I would definitely recommend doing that: I don’t know what situation you guys are in but yeah, you’re gonna have a huge cash flow crunch because you’re not getting your lifetime value up front, you’re getting it over a period of time. That becomes very difficult if that cash flow is important. Over time obviously it gets much nicer and it’s much more sustainable. I think our in our case we probably let it be that way for a little bit too long. A subtle change that we did was simply taking it from being side by side to just showing the subscription option and having a button at the bottom and saying “want a license?”. The idea is that if some people really want a license, there is still an option for them. Within a short period of time it went from 50/50 to 90/10. Then most recently in our last major updates we removed the license option completely. You can still get it, if someone really complains we can direct them to the store and they can buy it but we’re just putting everybody into the subscription funnel.

Max So the users who get the paid license, do they get updates forever for free? Is it perpetual like a lifetime subscription, for a one-off price?

Mark So users who bought version 1 licenses: because of the nature of our app we were able to nicely slice it. We’re also on the Mac App Store, and we wanted to do an in-place update. In-place updates are possible with upgrade pricing on the Mac App Store. It’s a pain and it’s difficult for the developer but it is possible. The way we did that was simply by not removing any feature sets: so the features that you had with version one, you still had access to. You could preview version two features and the output would be watermarked. It’d be the free trial kind of mode. And then you could pay for an upgrade at a discounted price and the same thing with version 2 cost of licenses. It will be until maybe version 3. The whole idea behind doing that, was that it was a whole problem, it that was a lot of work and it was difficult and I would rather avoid doing that again. This was the motivation for going on subscription as well for not having to deal with the upgrade pricing and holding those features back and having all of that complexity in the code.

Max Yeah, I can see that. This is a real problem with all those features which are in there. We’ve discussed all the options - really all the options - and when we came to that, we could stop shipping or enabling the features to use us. In my mind this got so complicated. Then you need to update the file format and then you need to opt out some features here and there and everywhere. I wouldn’t recommend that. I’ve not done it, but he said it’s not great.

Mark We’ve done all that, we’ve done App Store subscriptions. We did App Store subscriptions before they had all the new stuff. I’m a masochist I guess, I don’t know.

Pieter So what happened in our case? We were at version 3.4 or something and we were not really consistent in when we would do updates, for example version 4 took us six months to get out and we thought “ok, we’re never going to do that again” because that was too long, it was just a disaster. When we started to think about whether we needed to move to subscriptions, the first thing that was obviously we needed to stop thinking in terms of major and minor versions. You need to get continuous about this. You can’t ask a user to pay you every month if you only have an important update every two years. So the first thing we did was try to get a release going every two months. So 3.4 took us 7 months or something and then 3.5, 3.6 were all in two month intervals. We felt that with that we had both proven to ourselves and our customers that we could get updates out on a regular basis. When we announced that we were going to do this, after version 3.7 we went to version 38. And from then on every two months, versions 39 44 41. We just make it clear we ship every two months. No update is special. We just released version 50, that is not version 5: it’s not special in any way. It’s just a version number and I think this consistent set of updates is important if you are going to ask people to pay a subscription price.

Max I guess we just disagree slightly on that, but I don’t like to put too much focus on those updates. Because what you’re doing is you’re forcing yourself into shipping new features every two months. Okay, here we go. The issue is that with paid apps and paid updates, you need to have an argument for users to make the update so they need to have something that they want to pay for the update. With the Sketch model where you buy a license and after a year don’t get updates, you are basically still in the same model of paid updates because the user will be able to continue to use the version that they have but they will only get the new version if they want the new features that you’re shipping. So you’re forcing yourself into shipping new features. I mean if you have an infinite amount of users and money then you’re not actually forced but the principle itself dictates that users will only upgrade to a new version when they see value in it. And so there must be something different to them so they see value in it and that is only communicable over features.You can’t say “well, this is a major update and we only fix bugs” and everybody will jump on it. That won’t happen.

Pieter If I can just interject, that’s the entire point. There are no major updates. There’s an update coming out every two months because we have 10 developers working on Sketch. There’s always something ready to go out. And for most of our users that update will be free. And so it’s fine, they will just download that update. Of course with every update that we released, they’re always people for who bought just over year ago and that means that they can’t get this update. Maybe in this update there is nothing that appeals to them, no major new feature. Maybe just a few bug fixes, maybe it’s not a bug that they have run into so they don’t care. That’s fine, two months after there’s another update and if you get a few more updates like that, eventually that user is going to think “oh, this is a feature that I like” or “now I’m so many versions behind, it’s time for me to update”. So we never need to convince everybody that this update is interesting, we just need to get updates out and if eventually, we trust that they’ll update. There will be enough change that they will be enticed to. We were saying that you want to continuously provide value to users. Well, that’s what we are doing.

Denys What’s the conversion rate on the updates?

Pieter I don’t have exact numbers for this but many people they renew as soon as there is an update that they can’t have. We find that after three updates most people who are still using the app have gotten the update.

Max What I’m proposing is a slightly different approach to the subscription in that what you’re basically saying is you are paying the subscription for us to continue who to work and make updates and what I’m proposing is you are paying the subscription to be allowed to use the product and I do work on the product and I make sure that it’s good and stable and fine, but I take the freedom to do whatever I want with it and make updates when I wanted and ship whatever I want and the subscription fee you’re paying is for the permission to use it and not for me to guarantee a certain certain number of updates. So that is what I’m proposing.

Denys I have a few comments on this. I think by this approach you’re really limiting the amount of people who can use your product. So what we’re doing with Spark: we’re really inspired by what slack did. It’s free to use, it gets into the team and people start using it. Until somebody from our dev team comes up to us and says “hey guys, we want to get the premium version” and now we’re paying like 9,000 bucks for Slack because everybody’s using it. And I think with this kind of this direction you can get much more share of the market and you can get much more users and that’s what we’re doing with Spark, by providing as much for free even those collaboration features and discussion and email creation… Then give them enough value for some of the teams who need it, to upgrade. I think this opens up much bigger doors, and we can see how big Slack is.

Max But how do you sell the upgrade? Is it because “I’m adding new features every other week”, or are you selling the upgrade because “this is something you can use and I don’t make any commitments about how I’m going to develop it in the future”?

Denys There is no timeline: we have this Premium Package, which they should be happy to pay for and then that’s what we sell.

Max But then that is the permission to use the premium features. They’re not paying you to continue to develop the premium features, they’re paying to be allowed to use the premium features and this is what I’m proposing.

Denys I agree but it’s not permission, it’s giving extra. Wording is very important.

Max Maybe it’s the wrong word.

Pieter You make it sound as if we are forced to release, and we would much rather work on a big update every year: that is not the case at all. What we found in the past when we took too long to release a new update, there were so many things that we changed that we then found out that we actually accidentally introduced a bug here or there. In an extreme case, Facebook has an update out of their app every Friday or something, or they have this continuous police train going and we do something similar because this is how we like to develop the software, not that we are forced to.

Max But I can imagine that not every feature can be done in two months, right? So you have people working on multiple features at the same time and you always have something ready. I mean that’s a great situation to be in, I would want to get there as well. I’m not yet.

Mark To Max’s earlier point. I think it was the framing of it. When you buy a license for a piece of software, it’s basically you’re buying to use that for the current version, you’re buying the thing where “when I buy this, this is what I get” and that’s the perception around it. Whereas, I guess it’s probably to our detriment, you’re saying oh well by going on a subscription model it means we can give you these updates, we can you can do all of those things. But in reality you’re paying for the thing that you’re getting today. You’re not paying for some future feature coming in. Maybe it’s that’s our mistake of communicating in that way. That the subscriptions are for this future thing, when really what you’re unlocking is what you get and will be able to use today. By being able to do that, then we can add features and we can make it better because we have money to be able to afford to do that. But it’s a different framing that maybe gives that perception.

Christian I think it’s all about us just continuously providing this value. In Pieter and Sketch’s case, it’s this aggregate value over time. It builds, it builds and at some point if the renewal doesn’t happen immediately, it’s gonna happen at some point because they’re gonna have this tipping point of additional value. I do think that we have run out of time.

Denys Can I have one note? We’re talking about value here, creation and doing great stuff for people. But let’s be honest people want free stuff. Nobody wants to pay. Let’s be honest and that’s what we’re hearing from everybody.

Oleksandr But is it sustainable? For example, you have a million users on a free version and you have to support the infrastructure.

Mark Maybe if you don’t mind 10 seconds, I can wrap this up. You’ve got a lot of different perspectives on subscriptions here but you’ll note that every one of our businesses or our products are different, have different use cases. So there’s no silver bullet. Whatever you’re considering for your subscription business, identify what fits best for your customers, I think is hopefully what you get out of that.

Oleksandr I also want to share our approach. We’re releasing a new update of Clean My Mac soon which will be on subscriptions. Before even switching to subscription we tested a virtual subscription update we actually ran several weeks on our webstore and we said “hey this current version is on a subscription” and we measured how many customers would buy it. Based on our measurement we only get 15% decrease in revenue, which is great because it was a yearly subscription: it means if just 15% next year upgrade this means that it is okay. So our job is to make sure that more customers will upgrade next year and right now we are working on a full subscription update and everything that is needed to support it. This is how we approached it.

Pieter Do you have any final remarks?

Max My biggest point about subscription is really about principle. When you communicate this switch, don’t try to say “I have this and that technical problem”. Don’t go down that road. Say “I provide value, you have to pay and that’s it”. And don’t get too deep down. Really I mean, it just gets complicated.

Denys What about your blog post?

Max Yeah, I did edit a big blogpost to explain it, but right now when we sell to new customers, we just say this is the thing and no explanation needed, just go.

Pieter I just wanted to basically repeat what you said. It depends on your situation, the market you’re in, your customers, the way your company is setup, the kind of product that you make. It depends if you can do a subscription and what kind of subscription.

Christian Thank you very much, before we get dragged off the stage. Thank you all for coming and big round of applause for you guys.

Question A question I have is have any of you guys encountered subscription fatigue - the theory that if everything is a subscription, people don’t want all of our stuff to be a subscription and we start to get backlash.

Max I get that remark every week. Every week someone comes and says that and my answer is it’s true. There’s only a certain number of subscriptions you can have - but I don’t care. I mean, as long as you have my subscription. I mean why? Why do you blame me for all the other subscriptions, right? And for me, it makes sense for my product. It makes sense, I’m providing value. And this is how it works and either you can get it or you can’t get it. That’s the answer I give. Of course, we are moving away from “I have 10,000 writing apps on my iPhone” to “I just have one”. That’s the trend that’s happening. But I’m not against it.

Pieter I think in most of our cases, we provide a professional product that people need and if you’re providing that kind of value, then if for such a professional product people don’t want to pay on a recurring basis, you have a much deeper problem than subscription fatigue, I think.

Oleksandr Well, our answer is Setapp, we combine all the subscriptions into one place. But yeah, subscription fatigue is already happening because more and more apps go into subscription. And I believe it will affect developers of products that people are used to pay for because every month they will consider “okay, I’m paying already hundred dollars per month for a subscription and I will start kicking out some some apps from my subscription plan”.