Perhaps even more so than for other startups, our choices are 1) go fast or 2) go out of business.
As we grow, that gets harder.
I've been adopting a couple of new strategies to help with this.
One is well-known, most recently described by Jeff Bezos: make decisions with about 70% of the information you'd like to have. Reverse them quickly if you're wrong.
The other is something I call Default Go. I normally introduce it with "Unless someone stops me... <what I intend to do> <when I intend to do it>". This is a way of cueing the rest of the team that I'd value their feedback but that I won't wait long for it.
It's only a default. Anyone can stop me, turning the default-go into a larger discussion around what we're doing and when we should do it. But if no one does, I have all the permission I need and I'm moving.
I've seen a lot of companies - in fact, even this one - using Default Wait. A project or task stalls until it's had everyone's buy-in - after all, that's only fair, plus it's likely to produce better decisions (and it diffuses blame for bad ones). But good people are always busy, priorities vary, communications aren't 100%, and this shifts the project's delivery date (or sometimes its start date!) to somewhere between "next week" and "never".
After some experimentation, I've been persuaded to use a minimum 24-hour window for larger actions. That is still quite aggressive; someone who's very busy or off ill can miss any chance to comment. I think that, at least for us, once the window is more than a few days wide, the damage from waiting exceeds the possible benefits from more input.
For small actions - for example, getting a second set of eyes to read a mailout before sending it to 10000 people - I've used a window as small as ten minutes. That looks something like posting the mailer to slack with "I'm sending this when I've finished making a cuppa".
Like everything else, this is experimental, but it seems to work. Unless anyone stops me, I'm doing this.