Usability is about good design, and that goes beyond just the user needs!
There are a wide variety of stakeholders in any given design. Most obviously, a design must suit the business needs of the organization creating it, but even this is less obvious than at first glance. Within any organization there are people with multiple perspectives of what's good for the organization, what this particular design is supposed to achieve, and how it will impact them. The design and development team will also have goals they need to accomplish in the process of development: being true to the values of their discipline, their own personal values and quality standards, and the relationship of the project to their own career goals. To all this complexity, we add the fact that users themselves are not one simple, homogeneous collection of people but a diverse set of people with sometimes conflicting needs.
A usability expert is a people person, a mediator and a diplomat. You'll never be able to find a solution that pleases everyone, but your goal is to be aware of all the perspectives and find a design solution that works as well as possible for all of them.
Is there something wrong with all these points of view? Isn't one of these views the right one? "The user is always right", right? Yes, of course, but on the other hand no. A design that delights users but loses money for your business is a failed design. A design that sounds good but that can't actually be implemented, for technical or financial reasons, is a failed design. Nevertheless, all else being equal, the user's perspective is pivotal.
In addition, there is more than one user. Users include people who have to administer the information in an online database (e.g. manage inventory and orders for an e-commerce system). "Indirect users" include people who are affected by the design even if they're not actually using the computer. If the shipping or billing departments get printed reports from a computer system, and the reports don't make sense, then you've got a usability problem. Everybody matters.
Multi-disciplinary Strategies
Design involves tradeoffs among so many concerns that it's impossible to make good design decisions without knowledge of how human beings think and work, business and marketing, technology, graphic design, and, well, just about everything.
It's extremely rare (if even possible) to find anyone with all the relevant skills, so achieving usability is really all about forming effective teams of people from a wide variety of disciplines and getting them focused on common goals.
Visual Design and Usability
You'll occasionally hear the claim that achieving the best visual design or marketing message is in conflict with usability. This is absurd. Effective communication has always been a central tenet of graphic design, and marketing is pointless without it.
"Attractiveness", insofar as it is intended to make a pleasant experience for the user, is an essential part of usability. Usability is attractive and appealing. A design is not completely usable if it doesn't satisfy the emotional and motivational needs of the users. On the other hand, those who would sacrifice usability for attractiveness are simply designers who want to avoid a hard design problem. You can make a design that is beautiful yet hard to use, but what's the point? Designing both compelling and easy to use systems is a harder challenge, but one worth rising to.
"Branding" is all about communicating the audience, values, and intent of a design. An effective brand is one that helps establish clear expectations and consistency for users. Branding is all about usability.
Ego-free Design and Decision-making
Ego-free design recommends that designers not become over-committed to their own design preferences, that they stand back and solve the design problem from a perspective of resolving all the design criteria, especially suitability to the end-user. This means actually finding out what the user needs are, and being willing to give up on a design direction when you see that it's causing trouble for the users.
Many people are building websites for themselves and not their customers. They decide what looks good, sounds good, and makes sense for themselves. This is the first step toward a failed website. They should be saying, "I don't care how this works. What do my customers want?"
Of course, the final design must reflect a balance between the real needs of the business and the users, but no one should confuse their own design preferences for business needs.
Tradeoff Analysis
No design is perfect and no design can be perfect. Design involves complex tradeoffs. In a world of constantly evolving needs and evolving technologies, those tradeoffs will evolve, and better design is always possible but only through a thoughtful design process that explores alternatives and provides appropriate rationale for resolving tradeoffs.
Tradeoff analysis can be done at a very formal level to optimize decisions, but the essence of it is this: When you encounter a disagreement or unclear design decision, analyze the tradeoffs. Write down the criteria for deciding among alternatives, the pros and cons of each option. If necessary, weight the pros and cons, and choose the option with the most factors favoring it. If it's a 50/50 decision, then pick one and keep going.
The crucial piece is not to get emotionally committed to an alternative. Your initial preference will often be wrong, and that's okay. Be fair. List every pro and con that anyone feels belongs on the list. Then be happy with the decision that gets made. It's the best decision you could make, and the best decision will always involve some compromise.
Exposing the pros and cons is often enlightening. It reveals a wider set of concerns than most people anticipate, and it helps balance the views of people with very different backgrounds.