As I write this, there are thousands of designers and developers making websites. They are hard at work with coding, graphical work, and object creation. And they are wasting their time, because a large part of the work they're doing will eventually have to be changed.

Question: What is the single most important tool for a web designer?

Answer: A pencil.

The BIGGEST mistake in web design (and probably the most common too) is a lack of planning. It's a very hard urge to fight - you get excited about a new site or great idea, your client is itching to get started, and your schedule is finally clear... so you launch into it. A flurry of designs and database code later, you've got a good solid version of your site finished. And then your client says the scariest words a programmer ever hears:

"I was wondering how hard it would be to add [insert feature no one has thought of yet]?"

O-o-o-o-o-p-s. Adding that feature will require reworking three database tables, two layers of application code, and the web page itself. The whole process will take 3 days, whereas it would have only taken one day if the designer had known about it from the beginning. This pattern eventually leads to another popular phrase, but this one will come from the designer:

"This client has NO idea what he/she wants and keeps changing the project specs. I'm redoing things constantly!"

Who's fault is this? It's the designer's fault. Well, it may be the project manager's fault at a bigger firm. But SOMEBODY messed up, because there is an awful lot of programming going on without the proper blueprints, prototyping, and specifications. Programming a website shouldn't begin until the following has been accomplished:

1. A thorough analysis of the client's business and needs.
2. A diagram of the entire website, including all links, menus, and functionality.
3. A graphical mockup of AT LEAST 2-3 main pages in the site (all the pages would be better).
4. A complete class and database diagram, if the site has with significant application features.
5. An analysis of the future of the website and features. Will it need a mobile browser or web service? Integration with backend accounting system?

Anything less and you should be asking questions (whether you are a programmer or a client). As an example, a lot of web design firms instead use this model when designing websites:

1. Pick a domain name and register it.
2. Sign up for a hosting plan.
3. Approve one of these three site designs we made.
4. Write your content and give it to us.
5. Pay us, we're done.

After this process, you will probably have a website. However, over the next few months you will also probably have some problems with the site - new features you want, graphic ideas that no longer fit, pages you didn't think of, etc. All of these issues are pretty straightforward, they could have been discovered with some forethought and planning. Instead, they'll take 3x the amount of time they should because the designer must now destruct what he/she built and then construct the new functionality in place. Imagine building a house and then trying to change the height of the ceiling in the living room. Does that sound ideal?

Buy a good pencil and eraser, they'll be the best web design tools you own.