Jack Welch used to say, “Be Number 1 or Number 2 (or else get out of the market).” The operating principle of this was that the Number 1 company set the direction; the Number two company continuously challenged the leader; and everyone else was a reactive “me too” follower. Does this same “Magic Number” apply in the information technology and software world (where innovation is continuous and new markets emerge every 1-2 years thank to Moore’s Law)?
The Case for Four Market Leaders
When I first through about this, I said to myself, “In tech, the ‘magic number’ is four.” Just take a look at the “Four Horsemen of the Internet” (in the 1990s); the Browser Wars (IE, Mozilla, Chrome and Safari); mobile platforms (Android, iOS, RIM and Windows – with PalmOS left out in the cold); or servers (Dell, HP, Sun-Oracle and IBM).
But then I thought about other tech product categories and wondered about…
The Case for Three Market Leaders
Perhaps the case for the number of tech market leaders if three. In the Browser wars you could argue that Safari is a special case (Mac-focused) and the war is between IE, Mozilla and Chrome. In social media we have Facebook, Twitter and LinkedIN (with many wannabees). In the business applications space you also have IBM and Oracle buying every business vertical leader in sight to fill out a three-way competition with Microsoft. In search you have Google against Bing and Yahoo! (apologies to my old employer, AOL).
But are these really just pre-cursors to real ways between two leaders?
Jack’s Case: Two Market Leaders
Maybe Jack was right (he was about many, many things) and it really comes down to “Number 1 vs. Number 2 (with everyone else on the sidelines). In the database world this is Oracle vs. Microsoft. In the OS world it is Mac vs. Windows (in PCs) and Linux vs. Windows (in Servers). In the chip world it is Intel vs. AMD. You have Java vs. .NET in computer programming…
So What is the Answer?
I think the answer is one of life cycle and level. New markets can support four leaders. As they mature and settle out they will move to three for the “higher-level” items like applications (the speed of innovation will keep this from setting down to two – just look at what Salesforce is doing in CRM). If they are more “fundamental” like platforms or computing languages (things that require enormous capital and training investments to change) they will settle down to two (just like Jack said).