Design and optimization process usually looks like this:
While all is fine and good at point 1, everything falls apart at point 2.
It’s good to know best practices, and to have common sense, and to know from experience what might work, and what might probably not. But nothing replaces the research phase. However well you think you know the particular industry.
People are peculiar creatures. And for each industry, for each niche, for each website their peculiarity is different. A thorough research process uncovers that. Everything else is built on top of that.
Once you know in what way the visitors are peculiar, you build to accommodate them. Armed with this knowledge it’s much easier to not give in to number 7, all the changes and requests from other stakeholders.
While everyone usually has only the best interests, when all is said and done, after point 7 is completed, you’re left with a Frankenstein monster. With a thoroughly executed point #1 you gain a shield to protect yourself from all the butchery.
Just like in the sports movies, when you see the end effect and don’t see all the hard training (because it’s boring and unsexy), it’s the same with the optimization process. Before you start changing, deleting, adding and moving the shiny objects on the page, feeling like a hero, you first have to go through the unsexy.
Nail the #1 (discovery, user research, strategy, feedback, testing) and you’ll never have a problem with #8 (stakeholders complaining “it doesn’t work”) and more importantly #4 (questioning your life choices).
PS. Just like grueling sports training, research is a process you hate while you’re doing it, but are happy you did it when it’s done.