The way I set a minimum is to work out what a truck rolling out the door costs per hour once you include drive time, fuel, insurance, and the admin time to quote and invoice, then make the floor cover the smallest realistic visit plus that overhead. For most one-truck shops that ends up being roughly a two-hour block even if the task takes forty minutes. On estimates, I use satellite or photos to pre-screen and weed out the jobs that were never going to pencil, but anything with unknown condition, tight access, or multiple structures gets an in-person look. The scope creep problem is usually less about the estimate method and more about what the quote says: listing what is explicitly not included has saved me more arguments than any pricing change.
