This means that you have to do things very, very cheaply. Think of each growth hack as a single tactic. Execute in < 10 hours of work. Push into the wild.
You cannot know what will work, before you try it. What worked for others won't necessarily work for you.
How do you live in a world where any one thing you do is 90% likely to fail? You do a lot of 'one things', very cheaply, and when one of them hits, you exploit it.
Growth is a process, one very much dependent on speed. Think in terms of a growth system, not growth tactic or hack.
(this is what I call a strategy)