Tuesday, July 28, 2009

PC Sales Cycle Has Stopped. Is the Windows era over?

Windows is the crown jewel of Microsoft. It was Windows which made Microsoft one of the richest businesses in the world; thanks to Windows Microsoft was able to penetrate the market with its other desktop and server products. But Windows was even more powerful than that. Up until recently this system was able to dictate the development of the entire “PC ecosystem”. It was the Windows, which helped Microsoft to create so called “PC sales cycle”, which forced users to buy a new computer every 3 to 5 years and which forced any other player on the PC market (both from hardware and software side) to support sales of Microsoft software. This cycle worked reliably since 80’s, has been already repeated six times and up until now has been appreciated by the majority of PC makers and software producers, because it has been bringing new repeated business to all of them. But today, to the surprise of everyone - and mainly Microsoft - the PC cycle has stopped. Vista failed to push enough users to buy the new, stronger hardware. Quite to the contrary, new hardware commodity appeared on the market that ignored increased Vista requirements and – what is even worse – it gained massive success. Whom to blame? The netbooks. And of course the new emerging web applications which ignored the rules that governed the PC world up till now. No doubt about what does this all mean. Microsoft lost its market dominance. Other players don’t need him anymore. It is now only a matter of time when the Windows era is over. The new ruler will not be the operating system – the new computing landscape will be ruled by web based services, and financed by targeted advertisement. Operating system will become again a free companion of hardware, as it was before the Microsoft created its incredible marketing tool called Windows.

How to sell the same thing again

Before we explain the PC Sales Cycle, let us first look at one important business tactics: the so-called phenomenon of moral obsolescence. This is the tactics helping companies to get repeating business from existing customers even on markets, on which this would be otherwise very difficult or nearly impossible.

Every company needs to earn revenues on an ongoing basis. It seeks to have recurring revenue, because it needs to gradually finance its operations and growth. For companies selling services this requirement is met automatically. The more successful is my service, the bigger is my users’ base and the more money flows into my business. For companies which operate in the commercial software market it's however more complicated. Let us illustrate the problem by the following example.

Existing legislation gives all buyers of commercial software the right to use such software for a life-time. This represents a significant problem for software manufacturers. If I have a car, it will rod over time, the engine will stop working, so sooner or later I will have to buy a new one. The same holds in principle for any tangible goods - washing machines, refrigerators, vacuum cleaners, etc.

In contrast, for example a word processor is "eternal". It cannot rod; it has nothing what could break. It will always work as well as (or as poorly as) in the beginning. In the worst case, its users can really use it for their life-time. In that case they will unfortunately never buy any other word processor. And if all customers behave this way, the market of word processors will disappear.

This is obviously unacceptable for the software manufacturers. If they are unable to sell to existing customers, they cannot get the “low hanging fruit”, the easy revenues. This is even worse for the manufacturers that control significant market share of their particular market, because they have even smaller “free” market where they can grow. There must be a way to sell the same user the word processor again!

And indeed there is such a way – it is called moral obsolescence. The manufacturer comes up with a better product than the existing one, puts the new product on the market and offers it to its existing users. If customers consider that this new product (in our case a word processor) is really better, they will purchase it and pay for it to the manufacturer again. Up to this point this strategy is reasonable and fair to all participants.

The Problems Begin

Problems occur when the product has already undergone several such improvements, and therefore there are no features that this product would significantly lack. Users of the product are essentially satisfied. It sounds very good, but such a situation is in fact very threatening for the software manufacturer. In the worst case scenario these satisfied users will use their products (in accordance with the license) until the end of their lives, because the manufacturer is unable to offer them anything better. This would mean that these users will never pay the manufacturer again. The software manufacturer however desperately needs revenue from his existing users! He needs this revenue the more, the bigger market share he owns. The bigger market share he has, the smaller space for growth is available to him.

The solution is to create a new product even when the original product is good enough.

The manufacturer must therefore find new features at any price, because without them he cannot introduce a new version to the market. This principle can however lead only to two results.

Result 1: Growth of complexity of user interface

In the best case scenario the manufacturer adds features that are not important for most users, but still are somewhat useful. Apparently, this should cause no problem at all - why couldn’t my word processor have some built in features that I might make use sometimes in the future?

The reality is different. In fact, any new unnecessary feature (slightly) complicates control, and thus the usefulness of the whole product (again slightly) deteriorates. The more functions are added over the time, the more complex and confusing the work with the product becomes. Even the most advanced modern aircraft doesn’t have several hundred buttons and displays in its cockpit. A "modern" word processor, unfortunately, does have such a complexity. As a result, effectiveness of usage of such a word processor gets worse and worse.

Result 2: Re-making features and UI

In the worst case scenario the manufacturer is unable to find any reasonable function that is missing, and so it starts to re-invent the existing functions. This, however, has even worse implications to the users, because things suddenly work differently than the users were used to. The worst idea the manufacturer can have is to remake the user interface.
What can be a “modernization” of UI compared to? Imagine you bought a new car. With high expectations, you sit down to your new car, but at the same time you have to start looking for where the manufacturer put this time the steering wheel, or where the manufacturer has newly decided to hide brake and gas. Later, when driving the car, you are many times stuck because you suddenly could not cope with situations that were easy for you even one day ago, in your old car. And on top of that you are forced to constantly read and swallow the new marketing handbook in which "expert" explains to you that this change is actually the best for your efficiency.

In fact, users do well know how quickly and effectively they work in a familiar environment, where they do not have to think about how to find things and how to do tasks. Worst on the contrary is a work in an environment where every three years someone completely reorganizes things on our table. This means a real harm to any effectiveness, regardless of what the marketing experts claim.

Moral obsolescence

The conclusion is very simple. When a product develops to a certain stage, any further modification of it only reduces its practical usability and the efficiency of its users. Software companies are forced to do these “surplus” modifications, because they are vitally dependent on revenue from repeat sales and they have no other option how to sell the same product again. This dilemma results in a situation that can be called as "swelling of commercial software."

Two development stages of mass market products

To sum up, every commercial software (but this holds for basically every product on the mass market) develops through two stages:

  1. Development phase – in this phase, innovation is natural and valuable.
  2. Destruction phase – in this phase, innovation is artificial. The product is already beyond the stage when innovation makes sense. The innovation is now motivated only with the need to sell new product to the same customers. In this phase, we are facing the phenomenon of “artificial moral obsolescence”.

The specific of commercial software is that software producers are forced to develop their products beyond reasonable innovation for as long time as possible. The longer they are able to stay in the “destruction phase”, the bigger revenue they get. Every additional sales cycle counts in the company revenue. Companies in, say, consumer electronics market can do “destructive” innovation, too (see e.g., some models of Nokia phones), but they are not so desperate because they have also other options – to create a really brand new innovative product and thus return into the “development” stage (this is not possible in the, say, word processor market). In addition, everyone in the consumer electronics market knows that new innovative products (e.g., iPhone) can earn the company much more money than even the smartest and longest “destructive” innovation.

Swelling of operating system

This problem of “software swelling” is built-in into any commercial software. Operating system is no exception here. Also operating system must bring repeated revenue to its manufacturer; also this product is after some period of innovation in fact good enough and reasonably usable. So which are the specific consequences of “software swelling” for the operating system?

The operating system originally served as a simple tool to enable comfortable control of the hardware. Openly speaking, it should have stayed in this area. In particular, its role should be to provide a relatively stable platform for other applications and abstract the application software from the diversity of hardware (through the HAL - Hardware Abstraction Layer). Everything else should be left for the application software. (Naturally, Google Chrome goes right into this area of a simple lightweight OS – it doesn’t reinvent the wheel.)

Well, this is not how Windows looks like these days. As a result of the need to sell Windows to existing customers again and again, Windows gradually developed into a complex bulky system that includes more and more applications and services on top of a relatively complex user interface. Vista includes a browser, media player, DRM system, indexing system, etc. It is of course only the decision of operating system manufacturer what he would like to add to its operating system. Microsoft in particular, quite logically, adds components that support its other applications and services.

Emergence of the PC sales cycle

To be fair to Microsoft, we have to say that the "swelling" is an essential feature that is built into any commercial software, Microsoft cannot do anything too much about this. Microsoft depends on the sale of the operating system with a large part of its income. It does not have really any other option than to publish every three (if possible, preferred) to five years a new, "improved" system, to milk the market for some additional money. In the recent years, however, it became very difficult to find "improvements" that were still not built-in in the OS functionality (and which would not break the backwards compatibility). This problem became particularly strong during Vista development. Many bloggers and authors commented that existing users are already satisfied with Windows XP and don’t see any innovation opportunity here (and thus any compelling reason to migrate to the new system). Hesitant adoption of Vista by the market confirms reasonability of these opinions.

The influencer

The operating system has a unique position in the entire ecosystem of commercial software: every commercial application depends on it, and if the position of operating system is strong, even PC manufacturers have to closely watch its development and adjust their hardware. Any change in the operating system has therefore implications for the entire computer market.

Microsoft was able to use this dependence of the market with a real mastership – it created in fact a strong dependency of all application software manufacturers and PC makers on the success of its operating systems, and hence on its own success. It was a sophisticated mechanism, called "PC sales cycle."

PC Sales cycle

PC sales cycle is 3 to 5 year long cycle during which the PC users are forced again and again to buy new hardware, new operating system and new application software. The cycle begins with the creation of a new operating system, which has higher hardware requirements than the previous one, and therefore it is not running (or it is running only under very limiting restrictions) on the existing hardware. Hardware manufacturers are however already prepared to resolve this “problem of the users” and are coming with their offer of readily prepared new hardware that is certified right for this new system. Also manufacturers of application software do not hesitate to use this opportunity and bring promptly new versions of their products, which are able to make better use of the new hardware and new features of the operating system.

The main role in the sales cycle is played by two companies - Microsoft and Intel. For this reason, people sometimes speak about the Wintel (Windows + Intel) alliance. Of course, no formal alliance ever existed, however the better this system worked! Every new version of Windows operating system forced users to buy a new, more powerful hardware, and then the new software which made a better use of this hardware.

Even until today companies calculate the "moral" life of corporate notebooks and PCs as 3 to 4 years. They take it as a fact and incorporate it into their IT budgets. No one asks questions like, for example, why the computer must be disposed of after only 3-4 years of usage, while a TV set (a similarly complex product) would keep operating without major problems for 15 to 20 years.

Corporations also automatically allocate in their budgets money to purchase new versions of the operating system. Again, no question asked. On the monopoly market, you don’t have many choices, do you?

Microsoft however helps its customers to make such migration decisions much easier. Promptly, without unnecessary delays it notifies its customers that it is going to cease support for previous versions of its operating system, so everyone who decides not to migrate should understand he will not be supported.

End of the PC sales cycle

Vista was no exception to this proven strategy and shortly after Windows Vista launched, Microsoft announced that it would "not support Windows XP too much beyond 2008." This plan however never materialized. Around the time when Microsoft wanted to stop selling Windows XP, something unexpected occurred, that prevented Microsoft to make this step.

Lightweight devices called “netbooks” entered the market and started to gain popularity. The problem was that netbooks were unable to run Vista. In full compliance with the PC sales cycle Vista had to be more hardware hungry than XP. Unfortunately, netbooks are lightweight devices and their priority is portability, not performance. This is also what makes them appealing to more and more people.

Netbooks – the breach of the plan

Netbooks are a clear violation of the original plan: first time in the PC history, Microsoft was dislocated from the hands-on control of the sales cycle. Vista has inevitably higher hardware requirements than XP, to allow manufacturers to make money on their new hardware. Netbooks are outside of the will of Microsoft. This means very likely the end of the Wintel alliance. Its end will be as informal as was its beginning.

From our point of view it is rather surprising how long the PC sales cycle was running. It has repeated in total six times. The explanation is however easy: all participants on the PC market were happy, because the cycle yielded to the PC industry much more money than if the industry was left to the free market competition. In the situation when one major player was "conducting" the PC market, everyone was able to sell every 3 to 4 years to the same customers again, and therefore everyone benefited from new versions of Windows (Microsoft made the other PC players dependant on its own success - see this IDC analysis [PDF, 147 KB] of December 2006 - for every dollar earned by Microsoft from the sale of Vista the other companies in the PC ecosystem will make $ 18).

Maneuvering space of Microsoft

Microsoft had to respond to the success of netbooks. The company decided to keep selling Windows XP and accelerated to the maximum extent its work on Windows 7, which will again run on netbooks.

Here we should stop for a minute. This is a real landmark moment. For the first time since its dominance on the PC market, Microsoft has again to adapt to the market developments. This is the first version of Windows ever that is less hardware hungry than the previous one. After a long time it is not Microsoft who determines the evolution of the market. This puts Microsoft into a new, unfamiliar position.

In addition, his maneuvering options are very limited.

Microsoft Limitations

Microsoft, for example, cannot afford itself to develop a lightweight system with limited functionality. Microsoft must convince users that the OEM price of about $ 10 to $ 50 for one netbook brings visible functionality. If the operating system was light, almost invisible (however effective), it could hardly be distinguished from other systems, which are also almost invisible and also work, but are free.

Especially in the netbook market Microsoft will have a very difficult situation. When the price of these devices come down to $ 100 to $ 200, every dollar spent for operating system starts to be really visible on this very competitive market.

Microsoft, in addition, can not waive its backward compatibility. The main competitive advantage of Windows is that users can continue to use all their software, which they are used to and which they previously purchased. This is the main reason that keeps the users loyal to Windows. But it is also a brake on further development. Microsoft cannot thus go to a fundamentally different concept that would address issues such as viruses.

When will Windows die?

In summary, the position of Windows 7 will be very difficult, specifically in the netbooks segment. Unlike Google’s Chrome or Intel’s Moblin, Windows will start more slowly, due to backward compatibility they will allow to run not only all current programs, but also all existing "malware" and viruses. They will also mean a significant price increase on otherwise cheap netbooks. Any growth of netbooks popularity further undermines Microsoft's position in its key operating systems business.

And what will be the position of Microsoft in the traditional desktop? I have no doubt that it will remain very good. The question however is, how long will this market survive. Example of netbooks teaches us that users are already aware that there are alternative approaches how to do things. The fact that the sales cycle ended is very significant warning that things are not going to be as they were only few years ago.

The question therefore is as follows: how long it will take before the mainstream functionality moves from the desktop to the web. Once users start to look for functionality on the Internet, they will have no more the reason to look for it, and therefore to pay for it, on the desktop. The battle for the customer will then finally move from the desktop to the Internet.

At the same time this will end the era of commercial software, as well as the business tactics of artificial moral obsolescence. I am sure nobody will miss this particular feature of software.

Labels: , , , , , ,