10.06, Wednesday 9 Jan 2008

New software pricing models are always worth looking out for. More and more cleanly defined things are sliding towards plain ol' data (not just media. Home fabbing and local, short run manufacture are turning home electricals and clothes into free data plus effort). And because the concept of commercial software is so new (only since Bill Gates invented software-as-property 30 years ago), its business models aren't as entrenched as, say, music, so it'll blaze the way in finding new ways to be sold. Also software has an inherent lightness, which means it's suitable as a testbed for ideas that'll eventually inform how businesses work around open hardware and other forms of data (though of course it won't be directly translated, just as the Web is a testbed for including social ideas in mobile phones and televisions without direct translation). Software pricing models are try-outs for the future, just like social software on the Web.

I ran across the Celtx script editing/collaboration application [thanks]. It's free, with the money coming from web services built into the interface.

This follows the model of applications like iPhoto, which has integrated photo book publishing, and of course iTunes with its music store. Linotype FontExplorer X follows the iTunes model but is more interesting in that it gives away superb font management software (something that used to be expensive) and, as its commercial play, bundles access to an online store font (and buying fonts was always somewhat tedious).

Until recently Eudora had a paid and ad-supported mode, and Twitterific can be used free of charge, supported by in-line advertising.

Celtx, though, seems to represent something newer than these, something much more mature, considered and integrated as a model:

  • for one, the desktop software doesn't act merely as an enabler or gateway to the extra services; the extra services are there to make the core activity (writing a script) richer.
  • Second, the software has a pleasant gradient toward commercial use: while I can't find any pay-for services yet, my reward for starting a user account (from within the application) was free access to private, off-site back-ups; collaboration and sharing; and PDF generation.
  • And by joining the software with a community website for collaboration and as the home for other Web services, the software has fuzzy edges which gives it room to grow in the direction of user interest--something more like a web application rather than desktop software. Actually, I'll be intrigued to see where Celtx takes this: integration with Lulu for quick+easy self publishing, or the ability to send your script to the sub-editor equivalent of Lazymask? (Lazymask is an in-betweener.)

But what a turnabout! Social software where the software's the commodity and the social comes at a premium!

Where Celtx differs from end-user webapps is that web applications haven't made a great showing with various levels of features for different levels of paying. Flickr offer stats to pro users (and bandwidth), and blogging software has had a tradition of pricing tiers (before Wordpress--who remembers Blogger Pro?), but these are exceptions: in the main webapps are (a) ad supported, and (b) networked: additional pay-for services are offered via non-exclusive affiliate links, with the end provider providing the functionality instead of integrating into the original webapp.

Authenticity and going with the grain:

There's something else that rings true about Celtx, and that's its authenticity. The internet has helped reduce the variable unit price of software close to zero (mod marketing and sales). In part due to improved SDKs, IDEs, and APIs, open source, and the sharing of knowledge on the internet, the barrier to entry and the cost of software development is astoundingly low. It's possible - I don't have figures to back this up - that the cost of developing (or licensing) a web front-end to manage sales, registration and lost licenses for a piece of software like Celtx would exceed its original development costs.

The authenticity comes in because they will charge for what actually costs money per unit, unlike the reproduction of software-as-data: collaboration necessarily involves servers, which somebody has to host, and that costs money. PDF generation involves expertise and money to install, host and scale. And so on.

On top of authenticity is a sense of going with the grain: whereas software is simple to pirate without hard-to-code countermeasures, the use of a service online is just not pirateable. This is the way physical property and consultancy has always been sold, of course - in non-pirateable forms - and paying only for the hard bit is not new. Two interesting twists elsewhere: CentOS is a free Linux distribution, built by the community from Red Hat Enterprise Linux. It's identical but free; with RHEL you pay for support. Perfect for the difference between development and production servers. And with Mint web stats, what you get for purchasing is the latest plug-ins and access to the forum.

So while I agree that the qualities a software pricing methodology must have are, as identified by ASG [pdf], budget predictability, controllability, technological independence, value, flexibility, and simplicity, I'd want to add to that those two above: authenticity (because that provides simplicity and a sense of fairness), and going with the grain (because it's easier, and automatically doesn't promote piracy).

Further questions:

For all this discussion, where do new software pricing models get us? There are a number of areas ripe for a further look.

  • What these new models feel most like is open source hardware: the replicable-for-low-unit-cost part is free (with hardware, that's the design once it's drawn on), and what costs money is where costs are actually incurred--in the manufacture, whether you buy it from the same people hosting the community effort, or produce it yourself. Charge for the plastic, throw open the APIs. Ponoko is intriguing in a similar regard: you, as a designer, upload a specification for laser-cut wood/other material furniture and objects. Then, as a customer, you pay Ponoko - or a local laser cutter who has joined their network - to product the object for you. The designer gets a royalty. Yesterday I was talking with Schulze about publishing the designs for the furniture he makes in the workshop, in a form that anyone could take down to their local builders merchant and get cut for home assembly (D.I.Y.kea, he called it). Same deal. And Tim O'Reilly has talked about the significance of Threadless.com too. Same again. What happens when software fragments along the cost faultlines across its entire life-cycle? What divisions are there?
  • Technology offers new simplicities. Mobile phones tariffs, insurance and mortgages have all become more complex in the name of - let's be generous - giving the consumer a closer fit to their actual usage. But in the case of mobile phones, the tariff ends up altering your behaviour, which in a social situation really isn't great. For a business, new technology (primarily the ability to price adaptively) has the double advantages of easier to understand pricing models, and not having to build the expensive measurement infrastructure demanded by over-complex pricing systems. Car insurance per mile, for example, is more complex to develop as a business (otherwise we'd have seen it before) yet way simpler to understand and as an experience. What new simplicities could there be with software... in a way that the user doesn't feel they're being nickle and dimed, with "pay for me" nags next to every menu item?
  • Start-ups run on the Amazon Elastic Computing Cloud. Amazon S3 is used for Web-accessible storage of user profile images by Twitter, and accessible as a desktop drive through companies like Jungle Disk. Dave Winer has asked whether S3 could be an end-user product: As a developer who has to pay for his users' storage needs I would very much like to see users learn how to use S3 to store their stuff, so I can focus on writing software and fixing bugs instead of paying to store your stuff. I'm inclined to agree! We already pay for our bandwidth separately, so what if we also paid for our computing resources independently from the webapp, and paid webapps for the service they offered rather than for their infrastructure?
  • The Mint web stats package, mentioned earlier, asks for money from people who want to be involved in the community (or ask for support, and there's a quid pro quo there). In the comments threads about the soon-to-be-updated Super Duper Mac backup software, people are offering to pay for a free upgrade because they value the developer so highly. What if this behaviour was taken further: only the keen are asked to pay, for services only the keen value. Could this fund the entire software development? Okay, that's the shareware model, and it doesn't work too well... but what about novel differences? What if some upgrades were flagged as optional? Perhaps only businesses would pay for Microsoft Word, because they want the budget stability, and everyone else gets it for free? To be honest, I'm more interesting in exploring concepts like yield management for variable pricing in a software context. What if I priced my software so that it was a $1,000 per copy for the first 10 copies, $100 each for the next 1,000 and free thereafter? Or if I sold hours of support at some extremely high rate, and let any other company sell as many copies of my software as they felt they profitably could while having to purchase support hours from me (amortised across their whole user base)?
  • What other pay-for services are there, if all software worked like Celtx? What if Adobe had built out Lazymask, and included it in Photoshop... and Photoshop had become a platform like a Wii or an iPod that encouraged a huge secondary market around it? Or rather, what if Photoshop was a platform more like Dubai where taxes are non-existent to encourage growth, and the government makes its money in the same way everyone else does: on real estate and running hotels. What if Adobe sold Photoshop at cost... then distributed plug-ins to its services, competing with everyone else who also makes plug-ins to their services, offering everything from online collaboration tools and source control for images, to integration with local print shops?

My first job was at the Saturday boy in the local ironmongers and once, cleaning the top storage shelves, I found a price label that'd been there decades. Eric, my boss and one of the best men I have ever had the honour to meet, told me about the way products were priced when he was younger. The iron lawn roller he'd sold had the price cast into the body of the machine itself, because the price was as durable as the manufacturing process used to make it. It's not so simple now.