Can you believe it? K&R classic on C is on-line. Go and read it as I did so many years ago.

   

Just finished reading half of the book "Opening Xbox". Interestingly all the characters from "Renegades of the empire" are briefly mentioned (since they did DirectX, technology that was later used in Xbox). But the main thing I got from the book is that right know there are many people working on things that later become a history. There are people and technologies about whom books will be written, just like a book has been written about Xbox, DirectX or eBay. The question is: are you one of those people? Are you making the history? If not, why not?

   

If you've read fairly acclaimed book "The 22 Laws Of Marketing" you know that it stresses the importance of first mover advantage. As applied to marketing and economy, first mover advantage means that if you're the first to bring some product to market, you have almost unbeatable advantage over all other companies that have similar product but will come after you. Conversely it means that if you want to compete with players already present in the market, you're doomed. The book authors have plenty of examples to support their claim about how important first mover advantage is.

Not everyone agrees with that. For one, there are plenty of examples in computer industry that show the exact opposite. First spread-sheet is not the most popular spread-sheet, neither is first word processor, first internet search engine, first presentation package, first GUI operating system, first PDA etc. Alan Cooper in his book "The inmates are running the asylum" destroys this claim by saying that it doesn't matter if your software is first to market if it sucks and you're still extremely vulnerable to competition that will produce a better software.

Those are opposing views, which one is right?

The answer I found so far the most plausible is buried inside the book "Linked - the new science of networks". A network is just a set of nodes connected with links. Web sites obviously form a network - a node is a web page and links are, well, links between them. Products, companies and customers also form a network. Products and customers are nodes, while links represent the fact that a given customer bought a given product. So the success of the product is the number of links it has to customers. One insight of the book is that networks are not static but dynamic, constantly changing and growing. This maps well to real world: we know that markets evolve, some products die, new products are created, some products gain popularity, some loose. Another insight of the book is that links in the network are not created randomly. Observe how the network evolves from the very beginning. Let's start with one node. When second node arrives, forcibly the only link it can form is with the first node. When the third node arrives, it can form a link with one of two nodes. Even if probability of choosing a node to link to is the same for each node, we can see that nodes created first will get more links, just because they are older and had more opportunites to form links. So this support the first mover advantage.

But the probability is not the same. Imagine you have 10 friends, 9 of them drink coke and 1 of them drinks pepsi. When you're going to try soda, you're more likely to try coke because you were exposed to it more often. This "rich-get-richer" effect further strengthens first mover advantage. So how could late comer ever win (i.e. create a lot of links) in the network? It's because all nodes are not created equal. A web site with interesting content is more attractive than an empty web page thus it's more likely that people will link to it. People prefer to drink pepsi over water even though the water has been here longer than we are and pepsi is (relatively) new invention. Google took over AltaVista because it's simply much better search engine. We can model this effect by assigning different probabilities of forming links to a given node. The important insight here is that this intrinsic quality of the node is much more important than the other effects and that's why ultimately first mover advantage is not such a big advantage. A product that is twice as good as the product that is twice as popular will quickly become more popular because with time it'll get more links.

Of course don't foolishly disregard first mover advantage because even in this more elaborate model, if you have a product only marginally better than a competition and the competition has a majority of the market, you'll never catch up with them. You can beat first mover advantage but only with a vastly superior product.