Poisson not Poison.
One of the top skills that differentiate excellent consultants is the ability to quickly recognize patterns. Reincarnating Rorschach isn’t necessary. The ability to identify inefficient patterns of behavior is based on your knowledge and experience.
The recent Google Wave® application platform changed the collaboration model by allowing multi person editing or real time chatting. Real time chatting is a game changer. I can see what you’re typing instead of waiting for you to hit enter. Google® recognized that the currently acceptable interface of seeing “Someone is typing” is inefficient.
I was recently ate at Arby’s® and noticed a rather inefficient pattern. Three individuals were taking orders and giving out the receipts with the appropriate numbers on them. There was the usual snaking pathway to reach the registers. The line moved quickly; which was good. However, there were only two line cooks who were unable to keep up with the amount of orders. After being properly routed through the ordering process, the customers were expected to create a melee at the end of the routing to collect their order.
It would make more sense to extend the line waiting time, an ordered process, to minimize the unordered process of food collection. Add to the consternation was the soda machines was at the beginning of the melee.
I don’t need to be an expert at the Poisson distribution or queuing theory to recognize the idea that maximizing known, ordered events is more efficient that relying on unknown, unordered events.
Do I have the best answer for Arby’s®?
Probably not, but here is what comes to mind. Intuitively, the order would have at least 2 statuses; In Progress, Complete. The metrics should be available to say the optimal amount of orders, perhaps even to the specific line cook. If the total of the “In Progress” orders exceeds the ability for the line cooks to efficiently keep up then new orders should not be accepted or at least entry should be slowed.When my Grilled Chicken Bacon & Swiss Sandwich and curly fries finally arrived it was delicious.
What inefficiencies have you observed?
There's a reason the order taking process is fast and the food prep is slow. People will leave an arby's if the line is long and they haven't ordered anything but if they've ordered something, they won't leave.
It's optimized for Arby's, not you. :(
Now McDonalds on the other hand, has implemented an orderly way for you to sit and wait for your new food in their newer prototypes. Not necessarily faster, but minimal melee.
Posted by: Patrick St. Dennis | 12/10/2009 at 04:25 PM
That is a valid point. Thanks for commenting.
Customer commitment after purchase would be nearly 100%. I'd posit that the commitment to purchase occurs once entry has been made into the queue with the level of commitment increasing relative to the proximity of order taker. An individual who has waited in the line is unlikely to decline to purchase after the time investment has been made to wait. As for the optimization, I’m sure Arby’s and most other fast food services have figured out the maximum average time a customer expects to wait and allows the melee to last only that amount of time.
I wasn't aware of the new McDonald's in-store food delivery model. I wonder how this will be optimized so it will not affect prices and ensure a reasonable response time.
Posted by: David Crussel | 12/10/2009 at 10:18 PM