There have been many articles previewing Microsoft's XP operating system. My impression is that the general tenor of these articles has been critical - more expense, more Microsoft lock-in, more features being forced on us. Most of this criticism has focussed on Passport, with its potential for security/privacy problems, and on licensing models that may mean that you pay, and pay, and pay….

This article points out a different hidden cost, that of incompatibility. The author claims that: -

Read the full article at: -

My "Strategic Use of IT" course includes a couple of slides on Disruptive Innovation, and in a recent course we left the students with the question "Can your company support Disruptive Innovation". Somebody else's disruptive innovation may destroy our company, so we'd like to believe "Yes". But few companies can - and perhaps they shouldn't.

A Disruptive Innovation is something that changes the way a need is met, making obsolete the technology (and/or the industry) that has previously met that need. Classical examples are the development of electric lighting, which destroyed the gas lighting industry, refrigeration (which destroyed the ice industry), motor transport (horse transport). A Disruptive Innovation is not an improvement, however radical, in an existing product. Thus if (say) UNISYS had introduced a new mainframe 20 years ago with tenfold better price/performance than IBM's best, it might have changed their competitive position and challenged IBM's hegemony but it would NOT have been a disruptive innovation. Mini computers, and personal computers however WERE disruptive innovations - not only were they much cheaper, but more importantly they invented a whole new paradigm of how computers could be used, and attracted a different market.

Conventional wisdom, well supported by the evidence, is that disruptive innovations rarely, if ever, originate with the industry's leaders. They have too much to loose. Disruptive innovations almost invariably come from outsiders, who see an opportunity to break into a market with a radically different value proposition. For good reasons most innovations within a company are Sustaining Innovations, building on previous strengths to make them even stronger.

It's far from clear what the correct response of the industry incumbents should be to a disruptive innovation. Should Kodak, for example, seek to become a leader in digital image processing, competing with the electronic companies? Embracing the disruptive innovation in this way is a risky strategy. It sends a confused signal to the market, and shifts the struggle to a battlefield of your opponent's choice. Would a general let the enemy choose where and when to fight, if he had any choice? Yet ignoring it is even more perilous.

One approach is to continue to improve the incumbent technology. There have been many alternatives to the petrol engine with reciprocating pistons, yet none has prevailed. Today's petrol engine is cleaner, quieter, more efficient, and cheaper than ever, having seen off a host of radically different designs. Another example: computer disks. I can remember forecasts of the near-term demise of spinning disks, as RAM (then called "Core" or "Main" memory) became cheaper, and technology such as "bubble memory" seemed set to replace electro-mechanical devices. Yet tapes and disks are still with us twenty years later, and alternatives such as flash memory have not replaced them.

What caused companies such as Memorex to fail? Did they stumble, pausing for breath while their competitors went on with the next innovation? Christensen argues that they failed because they were good at things that companies are supposed to be good at - they understood their market and what their customers wanted, and they innovated and developed products accordingly.

The 14 inch disk manufacturers, for example, were oriented to the mainframe market. This market wanted large capacity, fast access, and low cost per byte of storage. By all these measures eight inch disks were inferior. The advantage that they did have - smaller size - had no value in this market. It did have a value in the minicomputer market however, but the established manufacturers had no presence in this market. The pattern was repeated with 5 1/4 disks (personal computers) and 3 1/2 inch disks (notebooks). In each case the new technology was initially slower, lower in capacity, and more expensive per byte than the earlier technology, and it's advantage - smaller size - had little value in the manufacturer's existing market.

It's not as if the manufacturers were blind to the lessons of history. Christensen relates the story of the 1.8 inch drive. By the end of 1993 each of the leading drive makers had developed 1.8 inch models, yet had found no market for them. In the meantime, Honda were buying 1.8 inch drives from a startup. The point is not that the makers of larger drives are not capable of making the smaller ones. The point is that developing a new market with an initial potential of perhaps 1% of the existing revenue is going to receive very little management attention at a time when management attention is focussed on fighting competitive battles and growing the other 99%. Focus is on the main market, where your innovations allow you to manufacture 100MB, 1GB, … disk drives. In the process you loose the ability to manufacture and market 10MB drives economically.

It is no coincidence that successful introductions of disruptive technology almost always require a new organization. Market leaders who attempt to manage disruptive technology within their mainstream business usually fail, whereas those that create a new business, either by splitting off a new company, or investing in a startup, often succeed. Hewlett Packard, for example, had little success with ink jet printers until they set up a new business, geographically and administratively separate from their very successful laser printer business. Established businesses need big markets and big opportunities to sustain their growth, but these don't exist (at first) when the new technology is being established. To establish the new technology requires an organization that can get excited about a $100,000 order. This is not an organization that is looking to grow a $1Bn market. In time the new business may overtake the old, but it won't have been given the chance unless it has the right environment at the beginning. The tale of the Apple Newton is relevant. Widely viewed as a failure, in fact the Newton sold about three times as many units in its first two years than had the Apple II. This would have been a great success, except that it was viewed as a major strategy of a giant company. Perhaps if it had been treated as a startup, today we'd be using Newton-X's instead of Palm-V's.

The essence of Extreme Programming, RAD, Iterative Prototyping, and other agile methodologies is rapid delivery of small increment of function. An example if this approach is a project that I have been involved in which has replaced the IT systems in a building company. The system has been brilliantly successful. It has increased internal efficiencies dramatically: staff now spend less than 1/2 of the time previously spend with their computers to administer their projects. Everybody within the company now works from the same information, so that when a supervisor notes that a task is complete, or an administrator accepts an invoice for payment, this is immediately reflected in the information available to the company's financial managers. Clicking a button produces monthly reports! The system is now being extended over the Internet so that suppliers and subcontractors will also see information relevant to them in real time.

A system like this didn't happen because we just asked the users what they wanted, then went away and built it. We started with a broad vision of what we wanted to achieve, and sorted out which parts of this vision were to be implemented with off-the-shelf software, and which parts we'd build ourselves. However, instead of then preparing a very detailed requirement specification and project plan, we built a prototype of the system. From then on, the development consisted of a rapid iteration of a "Review, change, release" process. After a few iterations the prototype became "Version 1" and was good enough to start replacing the previous software. Progressively branches were brought on to the new system. We then added "Phase 2" features, such as project scheduling. All the time we were learning - things that didn't work, things we'd omitted, things we'd done to plan, but now had a better idea, things that had arisen as new requirements. Typically we'd find that meeting the stated requirements wasn't difficult, but it might take two or three goes for everybody to agree what the REAL requirement was.

The rapid release cycle, with new versions of the software being delivered every week or two, has continued through the entire project life. There is a broad vision of where we wanted to end up, but instead of managing against a project plan management has the simpler task of managing a priority list. What is worth doing THIS MONTH? Are we still delivering benefit with this project?

To help us manage this project I developed a simple MS-Access database. Notes (of bugs, enhancements, ideas for discussion, and so on) are recorded in a database. A note can be assigned a priority and tracked through to delivery in a release if we decide that we want the change, closed if we decide that we don't want it, or left on the list for future consideration.

In the course of filing a patent application (more of that in a future newsletter), I've come across the following sites that allow you to search for patent and trademark information on-line.

The US Patent Office site has more advanced search capabilities (boolean options etc), but the Delphion site seemed to return better results anyway. An advantage of the Delphion site is the cross referencing of patents.

I must say that I found it difficult to find out what I wanted to know from either of these sites, although there may be little that they can do about it. It's often not obvious from the summary what the patent really covers, and downloading the actual patent application itself