You can, but first validate the problem, customer demand, finances and your ability to execute. A small pilot is often safer than investing heavily from day one.