On subscription software

On Tuesday of this week, Smile Software announced that their TextExpander software would be switching to a subscription model. As a longtime user and big fan of TextExpander, I have some thoughts on this change.

Another bit of software that I’m a huge fan of, YNAB, made a similar switch at the end of 2015—no more desktop apps sold with traditional licenses, but rather a new web app with a yearly subscription.

They’re hardly the first to abandon the paid-up-front/paid-upgrade business model. And it’s not just a question of small software publishers that trying on a new pricing option to improve cash flow. Most famously, Microsoft and Adobe moved to subscription models for Office and CC, respectively. While Microsoft still sells traditional one-time-purchase licenses for Office, that may not be an option forever—it’s clear that they want you to move to their subscription model.

Some believe that subscription models are the most logical conclusion for sustainable revenue. Certainly, it does make income more predictable: it more-or-less normalizes revenue across the fiscal year, moving away from the typical huge-spike-at-launch (followed by smaller spikes on paid upgrades) sales patterns. When a business can better predict income, it can better plan resource allocation and feature releases.

This is something that web-service businesses have been doing from the outset—monthly fees (with, perhaps, a discounted yearly fee) are the norm. I can only think of one service (Pinboard) that used to charge a one-time, paid-up-front fee (although even that has changed).

But charging a monthly fee does not sustainable revenue make.

At least, I don’t think it does for a large segment of the market. Monthly fees are something that are far easier to justify when they’re mission critical—which typically means software that helps a business improve cash flow. They can look at a cost/benefit ratio and determine that the net result is positive.

The vast majority of software does not fall in this category.

But I don’t think this is the largest impediment for software publishers going to a monthly fee. An even more interesting problem that comes with transitioning to a subscription model is that it takes the renewal-timing decision out of users’ control.

Nerds like myself will probably upgrade to the latest version of software the day it’s released, but smart people will weigh the pros and cons and see if it’s a worthwhile upgrade. Even if it is, a user can decide that they’ll keep using the current version awhile longer, if only because it improves their total cost of ownership.

Sven Fechner illustrates this nicely in his tweet on the TCO of TextExpander Classic vs. TextExpander 6. Taken over a longer period of time, with the traditional pricing model, users could opt to upgrade every eighteen months or even two years (assuming yearly upgrades) and save even more money. Or, at least, improve cash flow a bit by delaying some upgrades so that they don’t all fall around the same time of year.

Instead, with a subscription, what happens now is that I have no influence on when the lump-sum yearly payments will become due (unless there’s an option for a monthly fee, I guess, but that’s usually the more expensive option).

So, instead, I’m forced to create a “subscription software” budget category and start putting aside some amount to cover next year’s fee—or worse, come renewal time, I’m taken by surprise when I see a the fee I forgot about charged to my credit card.

I can tell you this: as that subscription software category starts to grow, some hard choices will be made. YNAB will stay because the money it saves me is at least an order of magnitude over what it costs me. TextExpander? Well, it’s a great utility, but I don’t know that I can justify a yearly fee for it.

I get that it’s tough for smaller/indie developers to make a sustainable income from their apps, but the reality of the market is borne out by the App Store: cheap/free apps are what get the downloads. So, businesses are forced to try new things, so that they can afford to keep making the software that we use and love. At the end of the day, the only people who are really willing to pay sustainable prices for software are software developers.

Paul Haddad tweeted some possible outcomes of TextExpander’s pricing experiment.

While I think that subscription software is a phase that some parts of the industry are trying on for size, it doesn’t have any staying power unless there’s a radical shift in the way we consume software. Something like, for example, an App Store subscription. We’re moving to music and TV/movie subscriptions anyhow; maybe—just maybe—one way to make this work would be to pay Apple a monthly fee for access to any app on the App Stores?¹

An interesting note: Smile have made sure to clarify that TextExpander 5 will continue to be supported for at least one more OS X version. It’s clear that they’re listening to the feedback they’re getting from their users.