Absolutely.
The opposite approach has equal risks: you start with a saleable idea that just isn't possible (or is possible but the costs are more than the sales will ever cover).
Either way it is critical have points where you decide if further effort is worth the investment. You have to make some investment in time/effort/money to find that out, if the answer is the product isn't worth the investment all your efforts to this point are wasted.
This starts from the (potentially flawed) assumption that your great idea is actually saleable. If you build something and prove that there is at least some market then great. But you might build something and find that there is no market.
If instead you start with the market, then you are unlikely to be pointlessly building something.