One of the biggest problems I've seen with servers is setting up too little server. On the other hand, it can go the other way quite easily too.

A three-person Real Estate office can probably get by with a very lightweight server, maybe even a single CPU, so long as it can serve up JPGs of the properties and such relatively well.

A thirty-person marketing company, with hundreds of clients and constant need for flashy pages, will need a lot more than that.

(Of course, I'm only talking about small businesses here. Larger need much more powerful servers.)

As the article mentions, it comes down to figuring out what you want the server to do now, what you will definitely want it to do in the near/medium/distant future, and the best guesses as to what you *might* want it to do, but aren't sure yet. The better you can do on figuring all that out, the better decisions on hardware will be.

I see a lot of under-resourcing. It usually happens because someone with financial authority goes with the cheapest hardware that will do the immediate job. Then, once the server is up, people start having ideas of "wouldn't it be cool if the server could also...".

It's cheaper to buy a $5,000 server today than to buy a $500 server today and a $1,000 server next month, and a $2,000 server a couple of months after that, and then another $2,000 server "to run parallel with the other one", and then to finally end up with the $5,000 server that could have been purchased at the beginning of the year.

Anticipate needs/wants. Anticipate Return-on-Investment. Plan, plan and plan some more. Then buy/lease.