Well, that’s a bit of a broad question. As a PhD student, I’m currently undergoing training to be a researcher, ie, for my whole life. If you’re talking about individual projects, then that’s a little different. A product life cycle might be research->engineering->manufacturing->sales, the research portion of which can range hugely in the amount of time it takes. Researching the best way to build a plane of the future is likely to take more time than researching whether or not a new table design needs three or four legs.