1. Less how. More why.

Too much time is spent discussing how something will be accomplished. That isn't a problem to ignore, obviously, but the more important question (and the one often shoved aside) is "Why are we doing this?"

From a design perspective, that means to make a decision quickly about whether spending any time about implementation is going to make sense if the only reason you're putting that there is because it's pretty. Your audience doesn't care about pretty. You think MySpace and YouTube worry about pretty? Worry about why you're doing something before you worry about how you're going to do it.

2. Not you. The user.

First off, I don't like the term "user," either. But it fit so nicely in the scheme of things that I elected to use it. I beg forgiveness.

What you won't be forgiven for is forgetting whom you're making your web site for. It's not being built for you, it's being built for your existing and potential customers. When you look at a designed interface, imagine it from the point of view of someone who hasn't been reviewing it and tweaking it and perfecting it for weeks. Think of it from the point of view of someone who just clicked on a link and found your page.

What are you providing that helps them get what they want or understand you and your company better? Anything? Nothing? Take a step back, then take another step back.

This isn't about you and your needs. It's about them. And their needs.

3. Now. Not later.

While a web site is (or should be) a constantly evolving entity, do not jump the gun to launch a site that isn't ready. If there are questions left unanswered, i.e. "What happens if the customer _________?" then answer them now. You know that old saying about not having a second chance to make a first impression? Here's another one: If you have an opportunity to screw something up, don't. If you know there are problems, fix them now.

Plus, you get double bonus points on this bullet because it has another meaning just as pertinent. When someone hits the screen, don't hide the information from them down the page because you want to show off your vocabulary and provide an elaborate explanation about what you're just about to tell them.

Tell them Now, elaborate Later.

4. Forget sticky. Think action.

"Stickiness" is about how long a customer stays on a page or a site. Stickiness isn't important. You can't create stickiness. You can create actions.

Give the visitor something to do, or something to explore. Make it relevant, make it succinct, make it simple. Allow them to control the experience, because they're going to do it anyway. Provide opportunities to explore (e.g. Amazon's "More like this" or YouTube's related videos) and they'll lose track of the reason they came there in the first place.

5. Clean. Not empty.

White space is important. Typography is important. Layout is important. Color is important. Content is important. Navigation is important. Everything is important.

Don't provide pages that don't provide use. Splash pages, interstitial ad pages and pop-up pages are all great excuses for your customers to leave. Now. And never come back.

Provide a clean and simple interface that honors your brand and takes care of the customer. But don't start inserting emptiness because it looks good.

6. Organize first. Design second.

My job is the hardest when the landscape keeps changing. Before the first pixel is set on a blank Photoshop canvas, I have to know exactly how much I have to account for in all aspects of the design.

What is the navigation structure? How big are the ads? Where do they have to go? What elements are absolutely necessary? How big does your logo need to be? How much content is likely to appear on any given page?

We all know that a web site is never truly done, and that there are lots of chefs trying to add their ingredients to the final dish. But the more involved I am in every decision, the easier (and faster (and cheaper)) the entire project is likely to be.

I actually like meetings, provided there's an agenda, goals are recorded, deliverables are assigned and it ends with clear agreements.

7. Expect accidents.

Plan for everything, even the unplanned.

When I am designing, the process is a bit like baking. Baking requires exact measures or things turn out less than perfect. In cooking, a little extra salt won't necessarily hurt the sauce. In baking, if you're a quarter-teaspoon off, your bread doesn't rise.

If I'm a pixel off, sometimes that screws up the entire interface. I have to be precise — but I also have to be aware of what's happening while I'm attempting to reach that goal of perfection. I have learned to incorporate the mistakes and they help me expand the realm of what's possible. Painting oneself into a corner is never a good idea.

Similarly, a web site has to fit inside a rather rigid structure, but the very nature of the webs highly interconnectedness suggests that it would be foolish to think that everything will work out exactly the way you've planned it. Expect accidents, and use them to your advantage.

8. Never assume. Always ask.

There's one step in the design and development of a web site that is often given little or no attention, and it's the one that needs and deserves the most: User testing.

Along the path toward the finished site, there are a lot of assumptions that get made with the usual two excuses: Not enough time, or not enough money. "We can't test every aspect of this design, let's just assume our audience doesn't need to know that piece." "We're too close to the launch to backtrack now, let's launch first and fix later." Any design project will benefit from a minimal amount of user testing. Simply asking a few questions and observing the results can tell you more in 15 minutes than it would take in two months of meetings trying to guess what the problems are.

When in doubt, always test.