I think what he's getting at is that a fairly large problem with starting a start-up is that you don't know if your product or service has a market in the first place.

If you start with a fresh idea, then you have to test and discover if what you are offering is viable. By taking an existing product or service, and improving upon it, that entire first phase of discovery is done for you.

That's why most new businesses aren't building something novel (ie: opening up a restaurant, retail store, software consulting, etc...).

That discovery phase is expensive. So if you're boot-strapping or limited on capital and time, then you're in a situation where creating something new is extremely risky. Even large companies with lots of resources often fail when launching a novel product.

Not every company is like Apple.

And conversely, many successful companies built on existing solutions. Google wasn't the first search engine. Facebook wasn't the first social network. Microsoft Word wasn't the first word processor.

I'll agree that "It's better to create a new market." But only IF you have the resources to take on that kind of risk.