Subduction

There's a worrying trend afoot in the world of computer software. After a decade of unsuccessfully pushing the idea of thin client computing, "the cloud" has finally made the idea fashionable enough for both Microsoft and Adobe to do what they've been wanting to do for a long time - launch subscription model products, with Adobe in particular pushing for this to replace their conventional offerings. Pushing the advantages of continual upgrades and access from anywhere through the cloud, Adobe think the public are finally ready to buy in to the new model.

Why am I worried? Ask yourself this: why are Adobe doing it? What is it about subscription software that's so attractive to a publisher?

I've been using Microsoft Office, Word in particular, for the best part of twenty years. From my first copy of Word 2, right up to my current install of Office 2007, I've handed Microsoft a fair amount of money over the years. At the same time I've gone from Windows 3.1 and MS-DOS 6.2 to Windows 7. The two are not unrelated, because the truth is there are few features in Word 2007 that I use and weren't present in Word 2. I am, after all, only writing novels.

This is an issue for software developers. Because any product as it matures will struggle to find new features with which to encourage repeat sales. Sometimes software gets an incentive in the form of new technology, so CD burner software expanded its remit by taking on first DVD and then Blu-ray, video encoders went from AVI to MPEG-4 and so on. Office software, meanwhile, has had few such technological shifts to drive upgrades - there haven't been that many leaps since the development of the printing press - and this makes it difficult to persuade users to hand over more of their hard-earned cash.

Microsoft realised this problem some years ago. Their first answer was to create Microsoft Office - a software suite. The idea was simple, charge people more than they would pay for one product, but less than they'd pay for all the products in the suite and, as long as users see new features they want in one of the products, they'll probably upgrade the whole thing. By selling the notion of "integration" across the packages, you also discourage users from upgrading applications individually on an as needed basis. Word 6 won't integrate well with Excel 2003, that's just the way it is and even if users realise this is deliberate design there's nothing they can do about it. So, Office was born and Visual Studio and a whole host of other suites of related software. Microsoft hit pay-dirt and many a software producer followed suit.

The trouble is that this approach only takes you so far. Sooner or later the whole suite is going to be enough for most users or at least the perceived value of any new features will be less than the actual cost of the software. If you're only buying the suite for one or two packages, this moment will come even sooner. Some companies dealt with this by just adding more and more shovelware to their suites. Nero, for example, has expanded its offering from a simple disc burner to a veritable suite of multimedia apps, most of which you'll never use, but which have often proven enough to encourage people to upgrade.

Microsoft, of course, had another advantage. Windows. During the 1990's the update cycle for Windows became entrenched. Driven by advances in gaming technology, users demanded more and more powerful graphics with a seemingly endless array of new features. DirectX, a feature Microsoft introduced in Windows 95, provided the software needed for developers to target the widest range of graphics hardware. Microsoft were canny enough to make DirectX itself free, thus ensuring its wide acceptance in game development, but by periodically insisting that users needed the latest Windows to use the latest DirectX they managed to ensure that gamers spent at least some of their money on keeping the operating system up to date. This, in turn, encouraged hardware manufacturers to support the latest Windows, which meant they were less likely to offer support to older versions, meaning that users generally had to buy a new version of Windows when they bought new hardware. It was simplicity itself to ensure that changes in Windows also undermined older versions of Office - not every time, of course, that would be too obvious, but every so often there would be some change in architecture which meant that a humble word processor had to be replaced to remain usable. Users were therefore locked into a cycle of upgrading software they were perfectly happy with just to keep using it.

In the end, even this strategy began to run out of road. PC gaming is not what it once was and advances in graphics performance have slowed, making relatively old games look less prehistoric than their forebears did at the same age. DirectX updates have become less frequent and less immediately desirable. Windows itself is no longer a licence to print money, with hostility to new versions like Vista and Windows 8 no longer confined to sections of the geek community. So where did they go from here?

The answer appeared to come from one area of software which has managed to secure a steady revenue: anti-virus software. Originally, anti-virus apps sold as boxed products with a number of years of free updates to keep track with new viruses. At the end of that time you bought a new release of software with a new series of updates. The trouble is that users didn't. Some switched to competitors; others, having had few incidents in that time, decided they simply didn't need anti-virus software. There's only so many viruses that can "escape" from the labs to keep users paranoid enough to regularly dip into their wallets. So the anti-virus companies made a subtle shift. Instead of buying the software, you pay for the service. A regular payment means that not only does your virus database keep up to date, so does your software. People's natural laziness does the rest and the revenue stream is secured. Even reasonable price increases can be factored in without spooking people.

This model is, not unnaturally, attractive to companies like Microsoft and Adobe. After all, if you can't persuade people to pay for new software, why not get them to keep paying for the same software? Throw them a periodic bone and they'll be happy, surely? And the rise of subscription music services like Spotify has added weight to this argument. After all, if people are prepared to keep paying a subscription for something that never changes at all like a song, they must be prepared to do the same for software, right?

But Spotify reveals one of the fundamental problems with subscription. The service is notorious for changing the availability of content. Whilst this is largely the fault of rights holders, that doesn't mean an office suite would be immune. How often has a release of Windows withdrawn a feature people wanted or added one they didn't? When the start button disappeared in Windows 8, you could hear the howls across the internet. Imagine if that had happened automatically, silently vanishing one night after an automatic update, rather than simply being something annoying in a new release that dissuaded you from upgrading. If you aren't the same as what Microsoft perceives as the majority of users, subscription will mean constant niggles about whether your app will do tomorrow what it does today.

For me, however, the key problem with subscription is the one revealed by the motives of the software producers in introducing it. Can it really be moral to force people to pay time and time again for the same product just because you can't persuade them to buy a new one? If a double glazing company demanded you pay a monthly rental for your windows you would quite rightly report it as a protection racket. Why should software be different? Free space on the cloud might sound attractive, but is it really worth a subscription to Adobe to get it? And if a company tries to lure you with attractive pricing, ask yourself how they plan to make money. Will the price shoot up when the customers reach a critical mass or will they make their money another way? We've already seen free services like Facebook monetizing people's content or making data about it available to marketing companies. Anyone who values their privacy or makes a living by their writing, photography, music or movies should be very wary of any easily accessible online storage.

Finally, the fatal flaw in a subscription system is the risk of impermanence. Imagine you got rid of all your music and subscribed to a system like Spotify. Now, imagine in a few years time you had a financial crisis which forced you to cut back. Spotify being a luxury would have to be in the frame for disposal, but to cease payment would be to lose your entire music collection. Had you retained your CDs or digital downloads you would still have all the music you'd ever had but would simply have to restrain the expansion of your collection, but if you're paying a subscription for access you'd have nothing. The same problem could arise if Spotify themselves went bust. No competitor would feel obliged to take on their customer base and give them the same accrued benefits. After all, why should they?

For businesses, the problem is even more acute. If you're reliant on subscription software and your income contracts it could mean choosing between vital services like rent, gas and water or the software on which your business depends. Again, if you'd owned a boxed product you would be able to continue simply by not upgrading unnecessarily. If you're hooked into a subscription, however, you lose it all - which could mean the end of your business. And if the software companies manage to get everybody onto a subscription model there's nothing stopping them simply upping the prices to the point where some people can't afford them. After all, if the majority have no choice but to keep on paying, revenue will keep going up.

As a software engineer myself, I do have concerns about the future of the industry. The continual upgrade model was never going to last forever. A shift to subscription might sound superficially as if it's good for people like me, but the truth is that if you can persuade people to pay a regular tariff for software which changes relatively little, companies will come to the view they need fewer developers. In the same way the publishers of those cookbooks you get in bargain bookshops have managed to continually sell the same few hundred recipes just by changing the formats of the books or combinations within them, the software companies will simply rearrange their software periodically to look new. And marketing will persuade management that this can be done with more people like them and fewer of those long-haired scruffy types in the basement. After all, if nobody needs new features why pay developers to write them? Subscription models are bad for developers and bad for customers. The only way to avoid a world where we all have to give Microsoft a chunk of our monthly paycheque to keep the marketing people swanking around in suits is if enough of us resist now and send a clear signal that we won't be sucked in.