Steel yourselves - the curve is coming
- Summary:
- The IT industry is in flux and for some vendors it will be disaster. There is a model, the Steel Intensity Curve, which just might help explain what is happening, and why this flux is inevitable.
Is the real 'lever of power' in IT shifting from the technology vendors to the individuals, organisations and businesses that use their products? I would say yes, but without any real evidence except gut-feel.
But some real evidence has come to hand which suggests that this change is inevitable and predictable. What is more, it is the story about Microsoft going versionless for the future of Windows that gave the clue.
Up until now the world's IT industries have been unwittingly operating under the strictures of a 'law', claimed by some to come from Intel's one time CEO, Andy Grove. Though definitive proof of that connection is hard to find, its effect is nevertheless important. It says: that when a business is on the upside of an asymptotic curve, it cannot see the top of it.
In other words, the world looks as though it will go on the same forever, and businesses don't know that their world is changing until it is too late and it has changed.
Couple this with a known reason for the occurrence of such change, the Steel Intensity Curve, and a recipe for great change, coupled with a lack of understanding of that change, is created.
Much of the IT industry - mainly, but not exclusively the hardware side - is heading towards significant, negative change. And yes, many of the software vendors will not escape unscathed.
Intense
The Steel Intensity Curve, is a tool for modelling the development and long term trends for the steel production industries. But in practice it applies to many more business sectors - with IT not amongst the excluded. It proposes that, over time, each percentage point increase in GDP leads to less steel being required to achieve it.
In other words, as far as steel consumption is concerned, as GDP grows, more and more of the reasons for steel consumption is needed are fulfilled. The bridges needed to be built are built, and they last maybe a hundred years and only need maintenance, which requires much less steel.
So the market demand declines. Steel lasts longer than the market needs to sustain its current levels of production and growth.
This suggests that the IT business may be in for something of a negatively trending surprise. Its customers, on the other hand, look to be on the cusp of some powerful developments in capability and business agility.
The Microsoft 'versionless' Windows story provides a clue as to why this should be. Moving to a versionless, continuous delivery approach opens the possibility for even reasonably ancient Win 7 PCs - and possibly older ones - to run the new operating environment.
And as more and more of the applications will be run on servers back at the datacentres, the client device (PC laptop, tablet, smartphone and Tesla car just for starters) will increasingly be just a communications medium between the users and the datacenter services. So the replacement market for such devices will diminish.
More of the workload will, of course, be processed on servers, so there will always be a market for new processor chips, disk drives (rotating as well as solid state) and all the other paraphenalia. But don't forget the key advantage of virtualization. That has increased the utilization of each physical processor significantly, and the trend towards a combo of continuous delivery.
Hyper-converged, smaller servers will not make up for the millions of new PCs that have been bought to sit on desktops doing damn-all most of the time they are on. The more this hardware gets commoditised and, more importantly, fully standardised the more existing server hardware will find opportunities for continued re-use.
So the Steel Intensity Curve certainly applies to the core components of IT hardware. The vendors have created an at least partially-false requirement for ever-faster, ever-more-capacious devices which will now probably come undone. There is an argument, however, that storage devices are perhaps the only ones to not fit the model, as the expectations on data growth suggest it will continue out into infinity.
Application thinking
The same can be said for a lot of applications software. The history here is pretty well charted. Vendors have sold licences to use individual copies of their products, sometimes even tying those licences to individual hardware systems that meet a defined specification. Update the hardware spec and you can find the software licence invalidated. That is definitely the case if you buy a new one. Port the application to new hardware and you pay again.
New versions of applications have, naturally, required new licences and new fees to be paid. It was a model that could invalidate the Steel Intensity Curve rules because this was like the software vendors saying 'I know you already have a steel bridge that works OK, but now you need a new one, whether you really think you do or not'.
But the arrival of the cloud has changed much of this, and the changes will get worse for those vendors that can't or won't adapt. The notion of shared, multi-tenanted use of applications in the cloud runs counter to the mindset of legacy software vendors. The rise of new competitors offering SaaS alternatives to the traditional approach is gathering pace.
In short, the cloud has applied Steel Intensity Curve rules to the software industry whether it likes it or not. At the same time, however, it is also creating a 'Day Zero' start point for a new market sector - Business Information Services (to call it something).
This will throw up new market and brand leaders, who will come from the those able to define and deliver types of service that resonate with consumers and business managers and their specific information management and exploitation problems that require solutions. These brand leaders are unlikely to come from too many of the current brand leaders in IT.
For the users, of course, this is likely to bring several benefits - not least more consistency across services and platforms, making them all ever easier to use together. There will be huge opportunities for users to build specific information service environments where the whole is worth infinity more to them and their business than the cost of the sum of parts.
Indeed, for the early adopters of cloud that get this right, they could become the brand leaders of the future themselves.
My take
This is obviously a subject for much deeper study, but the deep dive taken by the world's steel industries over the last 20 or 30 years can mapped and modelled using the Steel Intensity Curve. And it sure as hell looks like the same rule can be applied to what is happening in and around IT and its use today.