This also describes my experience in the online advertising industry. Come up with an idea that has a chance of increasing traffic or revenue, implement it within a few days, throw it into prod. Then you analyze the bejeezus out of it while tweaking/iterating (usually multiple times per day), but abandon it after a few weeks or months if it turns out to be a dud. Then move on to the next big idea.

It was actually really exhilarating, if you could overlook the fact that the work you're doing offered absolutely zero value to anyone involved except the bosses whose pockets you were lining if you succeeded. I don't work at a large global corp now, but at a small company (can't really call it a startup anymore, though that's how I would have described it a year or two ago), where there is still a relatively quick development cycle, but it's glacial compared to what I did at that last job. We are also trending toward much slower turn around times as we grow, which there was no danger whatsoever of at my ad industry job.

It's unfortunate that these developer-candy-store development cultures largely seem to be limited to get-rich-quick industries which can be unattractive from an ethical standpoint, or new startups which can be unattractive from a job security standpoint.