Executive Support, or more generically, management support up the ladder to the top. There are subcategories on this one, of course, because support manifests in different ways as needed and where needed. The most important is in recognizing the need for design as a value of the company, and investing in it. This means funding the right teams and instigating the right processes to allow it to happen. "Processes" are a bit vague as a concept, but essentially entails setting up a context in which "design is being done by the designers," i.e., they are enabled to do it, and other people who want to do it or aren't skilled at it are disabled from doing it. Of course, this cashes out in the daily nitty gritty of development meetings in which everyone may have an opinion about design (they will), where there must be a culture and a recognition of the fact that someone in the room has a better track record at being right in their judgments. Even if they make it look trivial and easy. In fact, especially if it looks trivial and easy.
Hiring well is the second factor, most important when you are trying to make the case for good design in a culture-change way (a pretty common experience for most of us, although getting less common, thankfully). I used to think this was a matter of just finding good designers, but that's only a small part of it. You also have to hire people who can work with good designers, which is not usually a factor considered by hiring managers in the rest of the organization. The more design "thinking" pervades the culture (from management messaging) the likelier it will subtly influence hiring in the non-designer roles, I believe. (Trivially, for instance, it's less likely that a product manager or engineer will be hired on the grounds that they are also "good at making icons" if there is someone on staff understood to handle that job already.) There are a lot of other personality and skill factors that influence doing successful design work, though. During the hands-on, day-to-day interactions in which design discussions happen and decisions are made to do one thing or another, the person on the ground has to be able to deliver when the context is good in which to produce and other people are listening.
(Hiring good designers is shockingly hard. Most UI designers can only list a handful of others they think are good. The "why" of that is another topic.)
The third item I'd add, which may seem obvious, is Development of good designs must be possible. This could conceivably fall under the management of processes, but I think it's worth calling out as a first class item. It does you no good to have designers who are good in a company that theoretically wants good design to happen, if there is no way to execute technically. The code itself and the timetables for work must allow design to be implemented; not just features, but designed features. I think Don Norman has a riff on this somewhere, and I know people who live in UI standards "police" roles wrestle with it all the time -- the toolkits used should enforce easy compliance with standards, not make it hard to honor them, etc.
And that's my third, for this six months of reflecting on the professional topic I care most about.