Sure, all businesses prefer to keep the customers who don't care about price; but you know what? there is rather a lot of competition for those customers, bec

Sure, all businesses prefer to keep the customers who don't care about price; but you know what? there is rather a lot of competition for those customers, because all businesses prefer those customers. This means that instead of spending on infrastructure, you will spend on sales. I mean, depending on how good you are at sales, you might still come out ahead, or you might not. All I'm saying is that the top end of every market I've seen is just as competitive as the bottom end, and you are constantly defending yourself from newcomers. Now, in my experience, this manifests itself on the top end as being very difficult to break in to the market. PG has some quote about how you might be able to out-engineer Oracle, but you will never be able to out-sales Oracle. Of course, if you do break in to the top end, it's usually a plum place to be; even after paying exorbitant sales commissions, you are usually sitting on a large margin. But damn, it's hard to break in. On the bottom end of the market? My experience has been that it's much easier to break in. I mean, yeah, margins are much lower, but marketing and sales become almost trivialities if you are dramatically cheaper than the other guys. A traditional strategy, especially in the infrastructure world, is to break in to the market on the low end; build credibility and a customer base, and slowly float in to higher-end markets. This is especially easy with infrastructure because your costs are always falling. Simply double the capacity you provide per dollar less often than once every two and a half years, and you are raising your prices. Really, I think it's a fairly good strategy.