Analysis of IT news

Tuesday, June 27, 2006

Case Study: What markets can Microsoft conquer?

With Bill Gates preparing his exit from Microsoft, let's have a look where the giant can go, and what markets it can conquer.

Microsoft has an impressive track record of capturing markets and kicking out incumbent giants: IBM (the PC), Lotus (spreadsheet), Wordperfect (word processor), Novell (network operating system), Netscape (Web browser), Palm (PDA).

On some markets, Microsoft hasn't gained supremacy yet, but the incumbent giants aren't in a great situation either: AOL (Internet portal), Sony (videogame console).

In some markets finally, the incumbent giants are still the undisputed #1: Google (search engine) and Apple (online music store and MP3 player).

So far, not once has Microsoft lost a market after it has conquered it. It doesn't mean however that any new victory is a major step forward for Redmond. A majority of its divisions are indeed losing money. But its policy has always been to invest in every pot to make sure to pick the winning one. That's how, in the early PC days, Microsoft was selling MS-DOS while at the same time co-developing OS/2 with IBM, developing Windows on its own, and owning licencing rights to a Unix for PCs.

Now, let's have a look at the various markets Microsoft is interested in:
  • The Web: the jury is still out there, as Google represents a very strong competition. It is however worthy to note that winning the Web browser war was an empty victory for Microsoft. The main purpose of Internet Explorer was indeed to control Web standards to tie the Web to Windows. This hasn't happened despite Internet Explorer's huge market share. So here Redmond has spent billions on a product which hasn't brought them any revenue and weakened Windows security. Worse, it could have cost Microsoft big if the Department of Justice had won its antitrust lawsuit.
  • The PC (client side): Microsoft two main cash cows are the operating system and the office suite. The big risk for those two markets is commoditization, as it's getting harder and harder to convince customers to upgrade. This is why Microsoft is trying to push for a rental model, where customers pay a monthly fee (you stop paying? your computer doesn't boot). Now, individual customers and corporate customers have different needs so will behave differently.
    • Individual customers: these are so over-served that there is a significant risk of a disruptive innovation slowly gnawing away this market from Windows. There might indeed come a time where an Internet appliance might be good enough for a lot of people. When that time comes, customers with small needs who are sick of all the problems with Windows will be tempted to switch.
    • Corporate customers: this is probably Microsoft strongest market, with no sign of change anywhere in sight. Windows, MS-Office and Outlook are here to stay on the corporate PCs. Whether corporations will upgrade as fast as Microsoft want is another story. The only thorn on Microsoft's side would be Linux. Not because it represents some real competition, but because it has forced Redmond to sweeten its deals with some large customers.
  • The PC (server side): Microsoft has been riding the PC wave on the server side, steadily gaining some market share at the expense of the Unix manufacturers (Sun's future isn't to bright). Icing on the cake, this is a market with wider margins than the client side. The main threat here is Linux as it is riding on the same wave as Microsoft: more and more powerful PCs.
  • The PDA: the market hasn't been that hot these last few years, but who knows? Maybe one day PDAs will be powerful enough so that road warriors prefer using them exclusively rather than carrying a laptop around.
  • The embedded market: Microsoft tries real hard to be into any consumer electronics device. Cell phones, MP3 players, video games, etc. Redmond is here trying to get a foothold in our living rooms. Once again, the jury is out there to determine whether Redmond will be successful.
Microsoft has several assets to win those markets: it has very deep pockets, it is ready to lose a lot of money if need be for a long term win (e.g. the Xbox), and it knows how to work with partners to enforce its products (e.g. the PC or PDA markets). However, it now has so many fronts that it can lose its focus. For instance, Bill Gates acknowledged that the focus on Internet Exporer made Redmond forget about the rise of Google.

Wednesday, June 21, 2006

General Theory: Disruptive innovations

The term was coined by Harvard Business School professor Clayton Christensen in his book The Innovator's Dilemma (which I highly recommend).

Some innovations (called sustaining) improve existing products. They make faster processors, increase efficiency, or add new features to existing products.

Some innovations are called disruptive because they do not fit in the established market model. They're not suited to improve existing products. They lead to the creation of new products which cannot replace the legacy, at least in the short term. Here's how it works:
  1. Disruptive products are generally under-performing (e.g. the early PCs, solar energy). This is because any new technology cannot be as efficient as legacy technologies which have been improved for decades.
  2. The disruptive products need to find a niche market where legacy products are not suited. This niche market allows them to survive, and get more and more per formant with time. That's how the PC found a place on the desktop, where the legacy computers (Unix and mainframe) were just not suited because they were too expensive. Similarly, solar energy is better suited in places far from any electric outlet.
  3. In the meantime, legacy products continue to get better and better. There comes a point however where it doesn't matter. Customer needs constantly increase but products performance/features increase at a much higher rate. There comes a point where the legacy product just over-delivers in terms of performance and/or feature (current office suites would be a good example).
  4. When this happens, the disruptive product doesn't need to match legacy products performance or offering, it needs to be good enough. When it was powerful enough, the PC started to attack the server market, and has been gnawing market shares ever since.
  5. The legacy products then tend to focus on high end customers who have needs the disruptive products still cannot meet. But this is a niche market that can only get smaller as time goes by.
A good example would be Hotmail. The online email service initially was bare bone, but it allowed to check one's emails anywhere provided one has an Internet connection, something Outlook cannot offer. As time went by, online email services got good enough to replace fat clients. Note that Outlook is still heavily used in the corporate world, where the needs are different. But there might be one day where online email might be good enough for companies.

See also:

Case Study: Music and movies online

In the mind of music and movie executives, the Internet is probably associated with piracy more than anything else. But new technologies in general might represent a far greater danger than piracy. It might end their control on their industries.

Music majors and movie studios currently control their respective industries because they have a strong grip on three key elements of the value chain: production, marketing and distribution. There are plenty of independent talents out there, but without at least marketing and distribution, they can't go anywhere.

Now, let's see how new technologies could legally affect the industry by allowing independent artists to bypass completely the current entertainment giants.

Production: the music majors are here at a disadvantage. Nowadays, independent bands can easily come up with a tape of their songs. Producing a movie is of course much more expensive but new technologies have helped, even if they're not completely there yet. The first one was the rise of digital cameras which allow shooting at a much cheaper cost (the cost of film is very expensive when you have a shoestring budget). But the big innovation is CGI (Computer Generated Imagery) movies. You can nowadays find lots of CGI talents all over the world on the Internet, a lot of whom are much cheaper than their U.S. counterparts. Independent directors can assemble virtual teams of people without even meeting them. That's how some movies such as Hoodwinked were created offshore, at a fraction of the cost of a traditional CGI movie. Sure, Hoodwinked isn't nearly as fancy as a Pixar movie, but it's good enough as long as it's backed by a solid story, and the quality of such movies can only go up. As a matter of facts, some people who produced CGI movies on their own have started to negotiate - and obtain - from movies studios a part of the royalties from their movies. Something unheard of before.

Marketing: once you got a hit song or movie, how to let the public know about it? There are plenty of traditional recipes: have the song played on the radio or on VH1. Buy commercial time to air the movie trailer. Use the fame of the featured artists to create some buzz (hey, that's why they're paid big bucks). But all this costs a lot of money, and independent artists seldom have that kind of deep pockets. On top of that, for every hidden talent out there there are hundreds of talentless people who believe they're geniuses. How can the mainstream customer sort out the real talents from the crap? Word of mouth can be effective, but the Internet can help too. What is currently missing is a Web site where people can discover hidden talents. One large enough so that mainstream customers can actually find some independent artists they like.

Distribution: the music majors have deals with all the main music retailers, and the movie studios have deals with movie theaters. Something an independent artist can hardly get. But songs just like movies are something that can be easily downloaded from the Internet, as file sharing networks proved. Online music stores like iTunes showed there was a market for legal music download, and there should be such a market for a lot of movies should the price be right. And iTunes has proved that it has a strong negotiating power over the music majors. Apple CEO Steve Jobs has indeed successfully managed to prevent any price rise, despite the major's strong push. On top of that, even though that's not iTunes main current use, independent artists could bypass music majors altogether by striking a deal with iTunes directly. The advantage for them would be much bigger royalties. And sites such as Google Movies already allow independent artists to upload their movies and set a purchase price.

Currently, the independent artists still have to go through the entertainment giants if they want to go mainstream. But new technologies could change that one day.

Now, one shouldn't expect a complete change overnight. A lot of innovations like CGI movies are disruptive innovations, so won't replace traditional movies instantly. But they have already much conquered the children movie segment, and could move to other markets as they get cheaper and more realistic. By the same token, some movies really gain in being watched on a large screen at a theater. But then again, the rise of super large TV screens and entertainment centers weakens this argument. And watching a movie at home has a lot of upsides (no driving, no line, no overpriced popcorn).

Tuesday, June 20, 2006

General Theory: Competitive Advantage

A key concept in business is competitive advantage. That is, providing something that differentiate you from the competition. Something that your competitors cannot copy, and that customers really care about.

When a given market has not much competitive advantage, customers mostly based on price. This leads to razor thin margins as the only way to stay alive is to keep the lowest price possible.

Coming up with a competitive advantage is not easy. But coming up with a sustainable competitive advantage is even more difficult. Something that gives you an edge in the long run.

One area where competitive advantage is so scarce is the Internet. Just as it's possible to create your website and get heavy traffic overnight, it's possible for other people to create their competing website. Sure, you can come up with a great new feature. But it's very easy for your rivals to copy that feature.'s 1-click purchase idea was stolen by rival Barnes & Noble despite being patended. There are however a few:
  • Features difficult to implement (so difficult to copy): Google's powerful web search is not only about software, it's also about a very difficult optimization. Amazon's feature allowing customers to read a few pages of a book or search for books is hard to copy, as it meant dealing with a lot of publishing companies to get this content. If this can protect a Website against small competitors, it can hardly make against a 500 lb gorilla like Microsoft who has decided to attack your market.
  • Brand name: and eBay are now household names. This is a huge competitive advantage. Another website can have lower prices, customers will still think Amazon when they want to purchase.
  • Customer base: when a Website is about connecting its customers (eBay, Craigslist,, the size of its customer base matters. This is probably the most powerful competitive advantage one can have on the Web. A new guy on the block can come up with the best auction Website around, but it would be extremely difficult to overthrow eBay because of the typical catch-22: no seller so no bidder, no bidder so no seller. The only exception would be when the value of the Website is mostly due to hipness or coolness. That's indeed how MySpace was able to overthrow Friendster as the cool networking site.

Case Study: Dell

Although much younger than a lot of established PC manufacturers, Dell was able to carve its place in a market with cutthroat margins.

Dell's success comes mainly from the fact that they understood and embraced the horizontal model. Dell spends very little on Research & Development, and lets its suppliers take care of the R&D. It indeed understood there isn't much product innovation a PC assembler can come up with that will result in real competitive advantage.

Instead, Dell focuses on other ways to come up with a competitive advantage. When a PC with 1 GB of RAM is a PC with 1 GB of RAM no matter what the brand, customers look at other factors to make their decisions:
  • Price: one of Dell's focuses is operational excellence: how to assemble faster and cheaper than the competition. That's where the Internet comes into play by bypassing the retailers who take their cut of the final price.
  • Convenience: Dell's Website allows customers to customize their PC. Once again, the key is operational excellence to assemble and deliver a customized PC as fast as possible.
  • Customers service: a PC might be a PC, but it's still far from being problem-ridden, so customers (especially the corporate ones) care more and more about service, one area Dell has been trying to emphasize.

Thursday, June 15, 2006

General theory: how technology influences the value chain

Technology influences the evolution of markets by having an impact on the value chain.

Now, what's a value chain? Consider the Personal Computer. PCs are assembled and sold by PC manufacturers (Dell, IBM, HP, etc.). But those companies do not create all the components themselves. They purchase them from 3rd party vendors (the operating system from Microsoft, the processor from Intel or AMD, etc.). Those vendors also rely on suppliers. On top of that, the PC manufacturers seldom sell directly. They often sell through retailers (e.g. CompUSA), service companies, etc. Each element of the value chain is a market, which has more or less weight on the end product.

Faced with this whole value chain, customers theoretically face countless choices. What hard drive? What processor? What PC vendor? What distributor? In real life, customers will only make a few key choices, and will let those choices dictate the rest. One customer might want the best PC for Windows, whoever sells it and wherever. Another might chose to buy HP because she trusts the brand, and find the first store that sells them. Another might go to CompUSA and buy what the sales rep tells him to buy.

At the end of the day, the customer looks at the elements of the chain that have the biggest impact on the end product. If a given element of the chain doesn't deliver much added value (e.g. only assembling parts), the customer doesn't care. That's why the main things customers look at for a PC vendor is the service. On top of that, it's also about competitive added value. If all the brands selling a given components are comparable, why bothering. That's why customers will care more about the graphic chip of a PC (as it can influence a lot performance) than the brand of the hard disk.

The constant evolution in technology thus can trigger changes in the decision process:
  • Standards: standards are a great thing for users, because one decision (the choice of the standard) includes a lot of underlying choices. When chosing VHS over Betamax, customers chose a standard over a VCR brand and simplifies their lives. The downside for lots of vendors is that a standard will turn some manufacturers into mere assemblers (see Vertical Integration vs. Horizontal Integration). The added value is then either moved upstream if some subcomponents can affect performance (that's how PC gamers care a lot about the graphic chip of their computers), and/or downstream as users will start to care about service.
  • A push towards what users really need: users don't need a computer per say, they need to edit their documents, access the Internet, etc. So if they can discard any decision involving technical details and focus on what really matters to them, they will.
  • Feature/performance overkill: There comes a point where components offer more performance and/or features than users need. Remember the time where we never had enough disk space? On the desktop, this era is gone. When this happens, customers barely care about the component anymore.
  • Sometime, an element of the chain can be replaced. That's how Dell was able to use the Internet to replace the network or retail stores.
Some non-technical aspects can also influence the purchase decision process:
  • Force of habit: that's how Microsoft Office is still the #1 office suite, even though there are some free office suites available, and even though MS-Office offers more features than anybody will ever use. But habits can die hard.
  • Hipness: Steve Jobs was able to come up with a very successful product with the iMac and its cool look. Apple was however able to pull that one off only after customers stopped to care that much about computers per say ("A computer is a computer").
See also:

Wednesday, June 14, 2006

Case Study: Apple, iTunes and iPod

News: several European countries are trying to force Apple to open its online music service iTunes to other MP3 players than the iPod.

Analysis: when a company sells two tiers of a solution (like the MP3 player and the online music service), it can favor one at the expense at the other. If you indeed open tier #1 to work with other third party tiers #1, this is at the expense of tier #2. Tier #1 becomes more popular because it can work with more counterparts, but tier #2 becomes more marginalized.

But Apple has decided to go another way: you can only use iTunes with the iPod, and vice versa. Because both products are leaders in their categories, this is a virtuous cicle as they reinforce each other. iPod users are forced to use iTunes if they want to buy music online. And once they do, they are forced to stay on iPod if they want to be able to listen to the songs they purchased. Note that Apple is probably following this strategy not because it's sound, but because it's in its culture to try to sell us all the components of a solution (the first Macintosh could only work with an Apple printer).

Now, if Apple was forced to abandon this model, it can favor one tier at the expense of the other:

Open iTunes to other MP3 players (at the expense of the iPod): one solution is to allow other MP3 players to use its proprietary AAC music format. But it can also let iTunes sell music in other formats. The latter option would allow to keep the iPod market captive.

Open the iPod to other online services (at the expense of iTunes): this can mean by either having the iPod recognize other song formats, but it can also mean allowing other online services to sell music using Apple's AAC format (Real Networks tried to). The latter option once again would allow to keep iPod users captive (you bought a song while you had an iPod? You can't play it anywhere else). One the one hand, people might be more enclined to buy iPods because they wouldn't feel so captive. But on the other hand, Apple WANTS its iPod users to be captive.

But even with legal constraints aside, Apple's current strategy might be difficult to sustain in the long term. Competition is trying very hard to catch up, and Apple may not be #1 in both the MP3 player market and the online music purchase market.

Microsoft has defined its own copyright-protected music format, and is making sure that pretty much any player and any online music service. They applied the exact same strategy on the PDA market, and see where is Palm now... Sure, Apple has an edge on the competition right now. But keeping that edge forever might be difficult.

Also, the music majors would like very much to increase the price of some songs. Apple CEO Steve Jobs didn't let them because his company's market share gives him a strong negotiating power, but chances are they will try again to get their way and will ditch Jobs as soon as they have the opportunity.

Now, if tomorrow Apple is forced to focus on only one of the two products, which one would it be? My guess would be the iPod, for several reasons:
  • Apple is an engineering company at heart, and coming up with devices that can't be easily copied is what they do best.
  • If it were not being the only online music store for the iPod, iTunes doesn't have a strong competitive advantage (no online store has).
  • Even if iTunes is a strong revenue generator, the iPod enjoys very comfortable margins: around 50% for an iPod Nano when sold in an Apple store, more for other models.

Saturday, June 10, 2006

General Theory: Vertical Integration vs. Horizontal Integration

Potentially any market is affected by this: the switch from a vertical integration (where each vendor build most of the components of their products) to a horizontal integration (where vendors specialize in one component and one vendor merely assemble and market them).

When a market is new, vertical integration is paramount. Products at that stage generally need all the performance they can get, so vendors have an incentive to come up with their own components specially tuned for these products. This happens not only in the IT world, but in every market. The Wright brothers, when designing the first plane, built their own engine because no engine on the market was light enough for their need.

But as time goes, performance gets better, and there comes a point where flexibility prevails as complexity rises. When IBM realized it needed to get a foot on the microcomputing market in the early 80's, it didn't have the time to design everything from scratch. So it quickly assembled a computer with existing off-the-shelf components (the operating system from Microsoft, the microprocessor from Intel, the hard disk by Seagate, etc.). The result wasn't particularly performing but it was good enough, and delivered in a record time.

But by doing so, IBM undercut its added value. If they could assemble a PC in no time, so could anybody else by buying from the same suppliers. That's how the people who reaped all the benefits of the PC were those who provided a real added value: Microsoft and Intel. Once a market has shifted to a horizontal model, it's very difficult to fight it. Compaq for example spent millions in research and developpement in the early 90's to come up with motherboards that would provide them with a competitive advantage. Unfortunately for them, Intel began selling their own motherboards. As a result, any PC vendor could buy a motherboard as good (if not better) than Compaq's without having to spend a cent on R&D, and probably at a cheaper price. Compaq just couldn't compete with Intel's huge economies of scale.

And we can observe the same behavior today. Google for example tries to provide a complete suite of online services on its own instead of leveraging existing third party services, which can backfire. It might indeed be because Google started competing with eBay with Google Base that the auction giant struck a partnership with Yahoo! And several people have questioned whether Microsoft capture of the Web browser market and obsession to tie it to the desktop was a good move. It didn't indeed bring any revenue for the company, and created a lot of security problem for Windows (not counting problems with the Department of Justice).

But switching from a vertical logic to a horizontal logic is more difficult than it seems. Relying on suppliers instead of coming up with your own components is a tough choice to make, because it erodes your competitive advantage (just like IBM experienced the hard way). Only a few companies like Dell who excel at operational excellence succeed.

Specializing is even harder, because it means changing your core target customers, which is a thing companies never do. Ford might now rely on thousands of suppliers, it still sells cars. Selling only engines would mean selling to car companies instead of selling to individuals, which would be too huge a cultural shift. Sometimes, a company makes more money with side products. American Airlines with its Sabre reservation system it licenced to all the other airline companies. And car manufacturers make a lot of their margins with car leases. But they still see themselves respectively as an airline company and car manufacturers. Second, companies integrate the vertical logic into their corporate cultures. Google was successful by reinventing the wheel. Web search existed years before Google, yet they succeeded by providing a better and faster service. Why would they want to stop reinventing the wheel?

See also:

Case Study: Google Spreadsheet

News Item: Google released an online spreadsheet, a moved seen by many as a threat to Microsoft and a potential MS-Office killer.

Analysis: Google Spreadsheet is a disrupting innovation, not competing against MS-Office while competing against it at the same time.

Google Spreadsheet (and online office services like online word processors in general) aren't seriously competing with MS-Office in the corporate world anytime soon. They just don't provide many of the features companies are using. And I'm not even taking into account migrating all the legacy spreadsheet.

But traditional office suites have reached a point where they are threatened by commoditization due to an overabundance of features. How did this happen? My guess here is that, paradoxically, Microsoft listened to their customers. 1% of the customers would like feature X? Let's implement it! 2% of the customers would like feature Y? Let's add it! Some customers complained that MS-Office was too difficult to use? Let's add the infamous Office assistants, hated by a huge majority of Office users.

As a result, most users indeed don't use 10% of the features of Word or Excel. Under those conditions, office suites are more and more seen as commodities where an office suite is an office suite. Microsoft is struggling to find new features that will excite its customers, but it's getting harder and harder. There are sometimes when less is more.

And this point is definitely crossed for individual users who have much simpler needs than their corporate counterparts. They will however switch only if a new software provides something the legacy software cannot replicate.

Here, Google Spreadsheet has several key competitive advantages. First of all, it's free. And second, it's available online. You can access your spreadsheets from anywhere and share them friends or family. Here, Google Spreadsheet doesn't need to be as powerful as Excel. It just needs to be powerful enough for its users needs.

Note that the same scenario happened with email. Online email services such as Hotmail didn't replace Outlook inside the workplace, but individuals embraced this new type of application because it was good enough for their needs and allowed them to check their emails everywhere.

News Item: Sun co-founder Scott McNealy stepped down

News: Sun co-founder Scott McNealy stepped down (

Analysis: unless they radically change, Sun is eventually doomed. The reason is simple: Sun is currently stuck in their vertical integration logic. They indeed have always lived by their culture of designing everything themselves (from the microprocessor to the operating system) instead of trying to specialize in a particular piece of a computer.

The problem is that if this logic makes sense at the beginning, horizontal integration eventually takes over. And when this happens, a company stuck in a vertical logic cannot compete against the economies of scale a horizontal model provides. That's how Sun's SPARC chips aren't outperforming the Intel chips as they used to (as a matter of fact, they're now even using Intel Itanium chips on some of their machines). But on the other hand, relying on third party vendors become mere assemblers like anybody else.

So Sun is in a situation where they're damned if they do and damned if they don't. By sticking to their current logic, they see their competitive advantage getting weaker every day, up to a point where there will just be no point in buying Sun hardware. But by switching to a horizontal logic, they lose most if not any competitive advantage, along with the fat margins their proprietary architecture enjoys (ever tried to buy spare parts for a Sun machine?)

They had the opportunity a few years ago to focus on Solaris Intel. But their lack of focus left the room open for Linux.

Maybe Solaris for Itanium will gain some market share against Linux, but it will be an uphill battle - especially if they always want to sell it with Sun hardware.