Andrew -
I spent almost 3 years as the PM for an automobile plant full-scale automation project that tied together all of the 72 acre plant with the order system, tracked cars through, etc. I learned a lot about the auto business and how they work things.
The scenario you suggest relative to internet commerce and direct orders from the factory sounds pretty good. Aside from the fundamental problem of customers not being able to touch and experience their car, I see another problem. The difficulty is the backlog of orders that these plants have. It varies, but usually it is around 6 weeks and they like it to be bigger.
In theory, it is possible to have all internet orders bubble to the top and get scheduled into the plant immediately. But that has huge implications for parts delivery, and tremendously impacts the materials planning process. This will translate into higher costs, since contracts are let based on projected needs.
For an initial test-the-waters type of system, I think a marketing department could get away with hiding the real cost of supporting the special delivery parts requests for these vehicles. (For example: If a car gets scheduled with a deisel engine, but the plant runs out, lets say because the truck overturned, then a special load will get flown in to the airstrip next to the plant. Whether this happens is a function of how many deisels are already in the plant and whether it is cheaper to sideline them (a real pain) or keep them on. i.e. Once the manufacture is committed, the cost equation changes.) So a marketing department could lose these costs for special build orders as long as volume is low in an experimental program. But the plant manager will start see this stuff and figure it out and either squawk or start back-charging marketing for it.)
Build smoothing is a high art in this type of manufacturing, and there are pretty well paid engineers who do nothing but work on tightening the slop between actual work for each station on the line and the available manpower. Build mix control was one of the major things that we were able to improve.
A secondary problem is that for cars manufactured outside the USA you have to add shipping time. Then there are always a certain number of damaged vehicles by the time the car gets to the customer. This number goes up more or less linearly with time in transit.
And last, what do you do with refusals? Consumers are notoriously fickle. For big ticket items like cars, you are going to have some problems with that, particularly for cars bought sight unseen.
In addition to that, some of these really big dealerships make an incredibly small margin on a per-vehicle basis. I think that is going to be hard to match. You would be surprised. They'll move some iron on a $30K vehicle to make $200 on the transaction. It's worth it if they do it fast enough. And when you add to that the float they can accumulate, (collect early, pay late) a really big dealership can easily fly for years in a deficit bubble as long as they can keep volume up or increasing. |
