When Should Software Be a Subscription Service?

The excellent Mac utility app TextExpander is moving to a subscription pricing model, and this is creating a dialogue about the justification for such a model. From a one-off purchase of $35 – plus upgrades every couple of years, usually at about half that price – Smile Software has moved to a monthly subscription service that would cost $47.52 per year. The company defends this pricing by explaining that they’ve moved the management of the software’s data to a web server, but for most individuals, this service is unnecessary. While there are interesting features for teams – shared text snippets, which can make it easier for employees in a company to stay on message – even this version of the software costs about $100 per year, per employee.

Is there anything in this upgrade to TextExpander that is compelling to individual users? My guess is that Smile has a lot of large business customers who will welcome the shared snippets and team management of their higher-priced offer, but how many individuals actually want to share snippets, or need their snippets to be stored on the web? Dropbox syncing is fine; it allows me to share the same snippets on my iMac and my MacBook.

This has raised the question of when software should be sold as a subscription. Developers may say that they don’t have to hold back features for a major upgrade, and can add new features at any time, since the subscription includes every upgrade to the apps. Consumers, however, see this recurring payment as often stifling innovation; once a company has guaranteed income, and users are locked into an app, they are unlikely to switch, and the developer has little incentive to improve their products.

I don’t contest the fact that developers need income. I have gladly upgraded TextExpander every time there was a new version, even if the changes weren’t important to me. It’s important to support independent developers, to ensure that the Mac and iOS ecosystems have excellent apps. In addition, the race-to-the-bottom pricing that we’re seeing with the App Store model is preventing developers from making enough profits.

The real issue here is not so much that of whether a subscription is good or bad, but of its cost, and its value to users. I have long used TextExpander, and would be willing to pay, say, $20 a year for the app. This is roughly how much I have spent on the app and upgrades, averaged over time. But $47 is just too much to pay for what this app does, especially when there are excellent alternatives, such as TypeIt4Me.

A number of apps and services are sold as subscriptions. Some examples of this are:

  • Microsoft Office, $100 per year for home users, for up to five computers, 5 tablets, and 5 smartphones.
  • Google Apps for Work, $50 a year for email, Google Drive, and a suite of productivity apps.
  • Adobe Creative Cloud, and, in particular, its photography package, which includes Photoshop and Lightroom, $120 per year.
  • Evernote, with clients for just about every platform, offering one free tier and two paid tiers, $25 and $50 per year.
  • Todoist, a task manager, with Mac, Windows, iOS, Android, and web clients, $29 per year.
  • Apple Music or Spotify, $120 per year.

When you look at what these apps offer, and then compare TextExpander’s pricing, you see the problem. TextExpander – which does an important but limited task, that of expanding snippets to longer bits of text – costs nearly as much as Google Apps, or Evernote, services with wide-ranging feature sets.

As I said in my article yesterday, “I really do feel bad to have to say this; I think the people at Smile are great, and they make excellent software. But I think they’ve made a big mistake, essentially increasing the price of this app by more than double.” The move to a subscription model just doesn’t make sense for this type of app, and the increased cost simply isn’t justified.

19 thoughts on “When Should Software Be a Subscription Service?

    • not that it is really connected to the post above, but I disagree. First, mailbutler costs less as proclaimed, plus the guys are offering all possible coupons all the time (just have to check their socials time to time), which I find quite nice of them. Second, calling the plugins needless…Well, it’s really an individual thing, I guess, but I would say a complete opposite thing. Very helpful, especially if you have big volumes of correspondence each day. Helps you handle twice as much as usually. Bottom line, totally worth it.

      • I bought ForgetMeNot 2-3 years ago when it was owned by Chungwasoft. It was a bit pricey, when one considers I had used a free plugin called Attachment Scanner Plugin that unfortunately was end-of-lifed at that point. But FMN was fine and still is. They now will only support it through El Capitan, so beyond that forget it in terms of compatibility. That’s how I found out that they now rolled every one of their plugins into MailButler. All well and good, even though most of it is not really critical or necessary. But to now pay more for that every year to keep the same ForgetMeNot functionality that I paid for one-time (sorry, but giving a 10-second delay for emails or having a read receipt are not things I need to pay around $100/year for) is irrational to me. And while they give 30 free “actions” per month, I found that every time the plugin thought I might be missing an attachment (since their triggers, which are built in and not removable, are understandably broad), that counted against the 30. So pretty easy to quickly and unintentionally use up the quota just on mail attachment checking. For example, go back and forth in an email thread that in the old quoted text mentions an old attachment, and you’re nearly done.

        I gave them feedback. Honest and polite feedback. And never heard more from them. I’d be ok with paying another one-time fee around $20 or less just because I do like to support good developers. But this is nonsense, even extortion. Sure, that is not the perspective of a developer. It is, however, the perspective of a user. So when FMN is no longer compatible with the next version of OS X (or MacOS, whichever they decide to call it), I’ll just have to do like everyone else and realize I might be missing an attachment after sending an email. That happens pretty infrequently, and I’d rather have the extra $100 or more (depending on the exchange rate; the payment is in GBP) every year for several years.

        Does anyone thing a suite of largely unnecessary Mail plugins is worth a total outlay of $500 over five years? I don’t.

        • Exactly. I was interested in their plugin that adds gravatar etc. images, right when they merged everything into MailButler. Tried it, but for such functionality, when I don’t care about anything else in it, to pay a seubscription, and this expensive one at that? No way.

        • Thank you very much for your feedback!

          I have to correct some parts of your comment:

          First, MailButler only consumes an action when you actually confirm that you forgot an attachment by clicking “Select file” in the reminder dialog of MailButler. I would say this is absolutely fair. In case of false alarms, no action is consumed!

          If you only need the attachment reminder functionality, you should be fine with 30 free actions per month, right? I quote: “That happens pretty infrequently”. Consequently, MailButler is completely free for you and there is no reason to complain about.

          And please be precise about the numbers when you write such a post. MailButler costs $81 per year and not $100. And if you are a customer of our legacy plugins, you even get a 25% discount on this price!

  1. Keep the last non-subscription version of the software you paid for. When Photoshop, Illustrator and InDesign CS6 stop working, they’ll be dead. They’ll be replaced by other, more agile apps anyway.

    Have you ever been where threre’s no internet? I have.
    Will you want to use your computer after the grid goes down? I will.

    • True. Though I’d imagine that something like TextExpander just keeps local copies of all your stuff. I don’t know about using, say, the Adobe apps, if they need a live connection to be authorized…

      • Adobe apps work fine without an internet connection. You obviously can’t use the cloud features without a connection, but the app needs to dial home for authorization once a month. If you need longer than that pay annually.

  2. Software development has produced more wealthy companies, and more highly paid workers, than anything else in the last forty years. So developers pleading poverty is unconvincing, at least so far. Yes, things change, and yes, 95%+ of software companies have gone out of business in the same period. But each of these facts is true in every industry.

    Supply and demand are in play. Software development isn’t too difficult to be sustainable, in some ways, it’s too easy. There are tens of thousands of programmers making programs that no one has asked for, in the hopes of capturing some attention and gaining a foothold in the market. That’s a tough challenge, and most fail, but it has produced lots of great software, and has moved the industry forward. The subscription model hurts all of those startups, and squelches innovation.

    I’ve got about 120 pieces of software in my list of programs that I have bought individually, and installed on my Mac. Most of these I paid for, although a few were free. Some I used once, some I use now and then, some I use often, and a few I never used even once. In a small way, I’m supporting 120 companies (minus duplicates), with these purchases. My purchase contribution is small, but so is my usage. Programs that I use seldom, I am less likely to upgrade. But I’ve paid upgrades on the majority, and it isn’t unique for me to only get a use or two from an upgrade version, before I pay for the next one. If each of these companies wanted only a dollar a month in subscription, I would drop the majority of them. If every new piece of software wanted a few dollars a month subscription to try them out, I would try out very few.

    I think the software industry is shooting itself in the foot, in embracing the subscription model.

    • I do think it makes sense for larger companies, especially those selling multi-user licenses to the enterprise market. So Microsoft, Adobe, etc; it works for them and their clients. I think it works for services like Evernote, and for music streaming services. However, when small developers try and hitch their wagon to that star, I think it fails. It’s not the same market, it’s not the same usage. I think they’re just seeking regular income without sufficiently considering how their users consider such a system.

      • I don’t entirely disagree with you, Kirk, but the term “makes sense” can have multiple meanings. It makes sense to rob banks, if you can get away with it, and this comparison with Microsoft and Adobe is not entirely facetious. It is inherent in software, publishing, and most media distribution, that the heavy users pay less (per unit), and are therefore subsidized by the lower-frequency users. If I use Photoshop twice a month, I pay a lot more per use than if I use it ten hours a day. If I listen to Apple Music an hour a day, I pay more per minute than someone who has it on all the time. If my Microsoft Office license is good for five users, I pay more as a single individual than a family does.

        I am not opposed to these differentials. The purchase model makes it more possible for the market price to be acceptable to a range of users. What I am opposed to is the model that requires the low intensity, small user to always pay as though they were a constant user. If and when it works for the companies and the users, fine. But it is quite possible for an industry to destroy its livelihood. The music industry has reduced its income to a fraction, through greed. I’m not sure if the current music model is good for consumers or not. I think it is likely that the software industry, in trying to get too much money and too much control over user choice, will hurt its own income, most of its customers, its growth, and innovation generally. The subscription model is the modern equivalent of the dumb terminal, with all decision-making centralized in some distant, minimally responsive, controlling hub.

        • “What I am opposed to is the model that requires the low intensity, small user to always pay as though they were a constant user.”

          Yes, I agree. I meant more the user who really does use the software. When I got my new camera last year, I tried out the Adobe Creative Cloud photography package. I quickly realized that I wouldn’t use it enough to justify the cost. But I also couldn’t justify buying the software outright.

          Office, for example, is a fairly good deal if you’ve got several people using it, but when you look at it for one person, who may not use it a lot, then it’s not such a bargain. I think there should be more tiers in pricing to take that into account. Take 1Password. They new pricing is $5 a month for a “family plan,” of up to five users. Well, I could count myself, my partner, and my son – even though he doesn’t live with me – but I’m still only using 3/5 of the available slots, and I don’t plan to have any more children. So, yes, it can feel like a rip-off. But also if you have six people and five slots; that doesn’t fit.

  3. Another excellent alternative is Typinator. I have been using it for many years; the developers keep improving it. Yes, like TextExpander they introduce a new version every few years with a discounted upgrade price.

  4. In a way, I’m now kind od glad Smile has made a change in their licensing schema … had they not … I would have never known some other options exist that may better serve my needs.

    I’m quite enjoying researching apps like Typinator and Keyboard Maestro.

    Thanks Smile! I would have never ventured forth without your inspiration to do so.

  5. I’ve been barely using TextExpander for years. It one of a handful of apps I install right away.

    This said, I have maybe 12 snippets.



    It saves me time, but I’m not what you would call a power user by any means.

    But I’ve upgraded whenever they ask because I like supporting good software and I don’t want this to die (or stop). But that’s sort of a ransom.

    I plan to stay with 5 until it dies, and then look for something else. Honestly, at this point I wish I hadn’t upgraded my earlier copies. Maybe they would have stopped working by now, but there’s no features added since I started using the app that we’re already there in the first place. It did what I wanted way back when, so I bought it. I’m not going to pay monthly to make sure it keeps working the way I want it to (I don’t want my snippets on their server regardless).

  6. I haven’t ever tried TypeIt4Me, and I never even bothered evaluating TextExpander because it felt overpriced for my needs back when it was on its old pricing model – let alone this one.

    Personally I use Atext from Tran Ky Nam Software. Costs $4.99. I’m sure it doesn’t do half the things that TextExpander does but it works perfectly for what I require. I think Atext is underpriced, but that’s their choice.

    Ultimately I think developers should be allowed to sell their software for however much they want, in whatever way they want. The market will decide if it’s a fair deal or not.

    Vote with your wallets, people.

  7. I’m glad you included Evernote in your list, as their model matches my own thoughts that the best revenue model is one in which the cost is related to the value received from the software.

    Evernote, like almost every other app, doesn’t provide much value for the vast majority of it’s users. They recognize this, and those users can use the app for free. Evernote monetizes them, but it does so in an open and honest way.

    Some small percentage of Evernote users are heavy users and they pay for the app. The determination of who is a power user is dependent on limits built in to the app. For the most part, the so-called premium features are just an extra. People move to the middle tier when they exceed the 60MB upload limit. They move to the top tier when they exceed the 1 GB upload limit. We don’t know what happens when they exceed the 10GB limit, but I suspect Evernote has some, so-called, ‘whales’ that use a lot more than 10GB and happily pay extra.

    The beauty of this model is two-fold. If applied to the prior discussion, I would be in the top tier for TextExpander, you would be in the middle tier (paying exactly the same as you were with the upgrade model), and most participants in that discussion would be using TextExpander for free. People would be paying in accordance with the value they derive.

    The real benefit with this model, though, is that it aligns Evernote’s incentives with the users. Evernote is incentivized to get users to use the app more. Their goal is to turn casual users into power users. The best way to achieve that goal is to improve the app, which benefits it’s users.

    Contrast that with a one-time purchase model, where Evernote’s goal would be to get more people to purchase the app. That goal might sometimes align with making the app better, but it would sometimes align with adding unnecessary bells & whistles or making the app easier for new users, but not necessarily better.

    Unfortunately, Evernote took a step backwards a while back by putting out a questionable update. That was a mistake, but the revenue model is a good one.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.