There are two ideas out and about concerning the progression of a technology to adoption, and the consumption of something that has already been adopted. The first, the adoption chasm, is about there being a chasm between early adopters and mass consumption. There is a book. The second, the long-tail, is about there being nearly infinite consumption of items with small demand. There is an article.
The each got a lot of press, so I’m not really going to talk about them as much as I’m trying to understand them in relation to the technology-utility relationship. At first glance, there seems to be a clear correlation between technology and the innovator, the early adopter, and the early mass market. There also seems like a pretty clear correlation between utility and the laggards. So the question becomes what about the early and late majority; the bulk of the distribution diagram? Also still out there is when does the long-tail really take off – with technology or with utility?
My hunch is that there is a competitive advantage to making the technology – utility timeline push the transition period back as close to the chasm as possible. I also think that there is a competitive advantage to achieving the ability to profit from the long-tail as soon as possible, but that the numbers won’t justify the investment until after the chasm is crossed.
I have a question about what are the kind of products and services for which a long-tail exists, and also for which a transition from technology to utility is possible.
The long-tail seems have physical limits; the further down the tail you want to sell, the closer your storage capacity moves towards infinity. The storage capacity issue can itself be off set by delays in availability; on-demand access requires on-hand stock but just-in-time production can effectively emulate infinite storage capacity. This makes selling physical products in the long-tail relatively difficult for anyone not able to solve the storage or production problem. Goods that require specialized production lines – I’m guessing that cars and microchips might fall into this category – in the long-tail can’t support the maintenance of those lines, which is why a potential market for, say, 1969 Ford Mustangs isn’t something that Ford can really support. It’s hard to support a long-tail market for something that can’t be reduced to a digital instruction for a programmable production tool.
The technology to utility transformation is, I think, unconstrained by actual limits, but definitely bounded by market acceptance. Given enough time, and left unmolested by manipulation, every technology will become a utility once it crosses the adoption chasm.
Briefly, the difference between technology and utility is three things:
- specialized knowledge: technology requires it, utility does not
- failure frequency; technology fails often, utility fails seldom
- failure severity; technology is fragile and brittle, utility fails gracefully (or catastrophically, but rarely in between)
If a technology makes it into widespread use, the market demand will drive it towards utility. This motion can be rapid or slow; electricity took something like 50 years to become something reliable in both industry and households, cable television seems to have done it in about 30, and ATM’s have managed it in about 20. On the other hand, internetworked personal computers have been available for about 20 years and remain generally burdened by specialized knowledge, frequent failures, and fragility.
One way to view this is that there is more profit to be made by technologists in delaying the transition than rushing into it because the end result is the commoditization of once high-paying jobs into low-paying ones. Such delay would be considered a manipulation even though it isn’t arranged by any stated agreement or collusion. A hallmark of manipulation is the persistent requirement of significant specialized knowledge to have full use of something that has been adopted by more than 20% of the marketplace.
What does that mean? It means that if you aren’t using a personal computer to do more than browse the web, send email, use an office suite, and maybe an instant messenger, most of the computer’s capability is lost to you. The same could be said of your car. You drive it, but it’s likely that you let someone else service it. The difference there being you car will go for years, even without service, and have nothing go wrong with it but the wearing out of consumables like tires, oil, or brakes while your computer might not go hours before something goes wrong with it that can corrupt your data, degrade its’ performance, or impede its’ ability to function.
A hallmark of utility is a go/no-go status, with perhaps some user replaceable consumables, like light bulbs. Utilities ‘just work’, technologies need technologists to keep it working. The less often you need to see a technologist, the more mature and stable and commoditized your technology is. Eventually that becomes utility, and the need for technologists drops off sharply. I’m not sure there is necessarily any specific point at which you can watch a technology become a utility, but I know it has to happen. Electricity, municipal water and sewer services, and telephones all seem to be entirely utilitarian, maybe even too much. Cable television, cellular telephones, ATMs, VCRs, and DVD players seem to me, more utility than technology; they meet the ‘just works’ criterion pretty reliably. Broadband, game consoles, digital cameras, and HDTV all seem to have not yet crossed the chasm, though some of them are close and when they do, they will ‘just-work’.
PDA’s, and mp3 players seem to me to be special. They don’t require a lot of specialized knowledge to make them work, but they do require some skills to use them, and they are generally disposable; they fail catastrophically, but infrequently enough in normal use to make you feel like you ‘got your money’s worth’ out of them. I still have a perfectly good Palm III that I don’t know what to do with, because it works fine; as an electronic datebook and address book it is very much a utility, but as a hand-held computer, it is still a technology. The iPod is another example; to listen to music it ‘just works’ especially when paired with a Macintosh computer, but to really exploit it’s vast set of capabilities, it remains tied to specialized knowledge.
VoIP and WiFi, on the other hand, need specialized knowledge just to use. They remain in early adoption but services like Vonage are pulling traditional dial tone providers and cable companies into the VoIP market and WiFi is being evangelized strongly by hardware vendors, so they seem poised to cross the chasm soon. The will however, do so as technology and will likely remain a technology well into the saturation of the market.
The point is that many things we use in consumer electronics seem to be crossing the chasm ready for commoditization, whereas most things in information technology do not. Well designed hardware, operating systems, and applications are not the norm. It is easy to blame Microsoft for this, so I will, but there are ample other place for blame to be spread as well. All you have to do is look at what Apple has done with the Macintosh and OS X, what TiVo has done with a Linux appliance, or what Asterix has done with an application to see that good design will yield a product that doesn’t come with strings attached in the form of cradle to grave tech support requirements. These simple, ‘just-works’ focused products have come to the market aimed at the technologist but ready to become utilities as soon as the mass market adopts them.
Additionally, the Macintosh and TiVo are well positioned to take advantage of the long-tail market by being the broker of the long-tail transaction. Apple’s iTunes Music Store and TiVo’s relationships with Netflix and Comcast each are entertainment markets where digital format and broadband connections enable nearly instant consumption of very low demand items for no more cost than for the most popular items.
Microsoft doesn’t do this, in fact they generally do the opposite. The typical Microsoft release is not aimed at technologists, but rather at the mass market. This is because their installed base is so large that they can essentially force feed the marketplace anything they come up with. Because of this, Microsoft software and operating system products – both upgrades and new entries – never have to deal with innovators or early adopters, they are effectively given a free pass over the chasm by Microsoft’s monopoly. The end result of this is that technologists generally dis-like Microsoft products, and consumers generally don’t have much chance to get free of technology in the operating system unless they want to abandon the i386 architecture all together and use a more expensive, less familiar Macintosh that may or may not support their favorite application, game, or device. Because Microsoft software and operating systems aren’t designed from the beginning to be utilities, and compounded by their avoidance of the innovator and early adopter markets, advancing versions of a given product move towards utility at a much slower pace than those with a more progressive design philosophy. Microsoft sees that it ought to control the user interaction with the application or operating system, and this prejudices it towards the status quo where the back office is for technologists and the desktop is for users of technology. Contrast that with the electrical company paradigm of power is for people and it is provided by tradesmen.
Not to pick on Microsoft alone, Sun, HP, and IBM all make similar choices for their enterprise server operating systems and applications. Oracle and SAP do too. The Linux vendors Red Hat and Novell are following similar paths, as are the hardware manufacturers Dell and HP. Dominating a market and using that market domination to avoid the scrutiny of the first 20% of the adoption curve seems to be all the rage. At the same time, it seems to be detrimental to the transition from a technology to a utility, and any competitive advantage that might bring with it.