I'd say that paying for work product instead of time and materials for very small projects to test things out is smart way to begin. But longer term I've found

I'd say that paying for work product instead of time and materials for very small projects to test things out is smart way to begin. But longer term I've found that serious professionals tend to skew more to billing for time rather than per-project (I'm not saying there aren't exceptions), because they've at some stage come across scenarios where clients scope a huge project with them, and all the friction is backloaded towards the due-date as the client and team begin arguing about whether the spec is delivered or not.