I work in graphic design, which is a highly subjective area. In almost every project, we create a range of designs. Usually, a 1-2 of these designs become lead contenders. Out of these designs, there can only be 1 selected as final. This is where things get interesting.
Design is all about communicating information, transferring knowledge that can be visually decoded and understood. Some designs do this more effectively than others. Some designs are more pleasing and look prettier than others. But beneath it all, design must be judged in how the audience "reads" the design, and if they read what we want them to (this looks delicious, this will clean my teeth, I want to wear this).
So how do we figure out which option works the best? We do consumer research to see what a broad swath of the public think. We look to the design experts and marketers (that's me!) to use their knowledge of the market and experience to give a recommendation. From this, it's still questionable whether the final design selected is "the best" design.
More often than not, clients go for a safe design that isn't too far off from what is already in the market or what they currently do. We push them to take chances and get big returns, but they usually like to go safe for the small but guaranteed win. So which option is "best"?
A colleague of mine at work said, "the best design is the one that everyone agrees upon." This is absolutely true. You can debate 1 design versus another for week, but sometimes the most important part of the whole process is to actually make a decision.
It can be a tough lesson to learn, particularly for designers that get emotionally attached to one design or feel that a particular concept works better on many levels. But this is the awful truth: The "best" design is the one that allows the team to move past the question stage and get on with the project.