A Website is a Hypothesis

1. Behind every website there is a hypothesis

1.1. Alternatively, a website is a diagnosis

1.2. With a website, you've posited a solution to some problem, whether your conscious of that or not

1.3. You make a hypothesis as to how you can solve that problem

1.4. You test your hypothesis by launching a site and measuring the results

1.5. You adjust based on new knowledge

2. Process doesn't ensure quality

2.1. With these insights, it also becomes clear that the steps in your process are a means to an end (making a better hypothesis, putting a better experiment into the world)

2.2. Process and methodology do not ensure a good product, they just wards off the overwheling malaise by details and reduces the coordination costs among team members by creating more shared context and expectations

2.3. The things in your process should be used to help you better understand your problem space so that you can make a more clear hypothesis and then to run a better experiment

2.4. In the end though, your site lives and dies by the insight of your hypothesis, your ability to execute your experiment, and your willingness to iterate as you learned

3. This view offers insights into process

3.1. If a website is a hypothesis about how to solve a problem people have, it motivates a couple behaviors

3.2. You want to understand the problem well

3.3. You want to be able to measure your results

3.4. You should expect to iterate

3.4.1. If your site is a hypothesis or a diagnosis, then you would expect to refine that

4. Measurement becomes simpler to think about with this view

4.1. Once you've accepted that there is some hypothesis behind your site, it becomes more obvious what you should be measuring

4.2. Vanity metrics become less interesting because you suddenly are more specific