An email exchange between myself and Greg Carson.
My initial email:
Imagine this.
You're the owner of a small pizza place. A new corporate place opens across the street. Their drivers are better trained and have faster cars. Plus, they have more cooks and faster ovens so the pizzas come out quicker.
Traditionally, you thought pizza delivery was all about the speed of the driver. But you can no longer compete on that level.What can you do to compete?
The obvious answer is actual reality; the quickness of the run is affected mostly by; 1.) the quickest path taken between orders on the run and 2.) the sorting and allocating of orders for particular runs.
To put it simply, 1.) most people know the quickest route from Montclair Village to Snake and Skyline, but do they know the quickest route from Snake and Skyline to Mountain and Avoca? and 2.) If you take Snake with Colton and leave the Piedmont order for the other guy, the Piedmont order will get there quicker than if you take Snake and the Piedmont order and leave Colton for the other guy.That seems too obvious to include as a revelation in the movie. So I was wondering if you could think of something else.
But I don't think I'll go that route anyway, because although it might be interesting in theory and on paper (trying to create the most efficient pizza delivery model... hey, that could be your Master's
thesis) it doesn't necessarily translate well to film.I was just thinking about it because I'm reading Moneyball and that's what it's all about.
Actually, the way the screenplay is now, is quite the opposite of Moneyball. Whereas Moneyball favors strict analysis over "hunches", the film is more about listening to the voice inside (both in pizza delivery and love, as well as life in general)... you know, "use the force Luke" kind of stuff
The little pizza place that could up against the corporate chain who's drivers have GPS
Another idea I have, though, is instead of the corporate place moving across the street, I have the owner of the original pizza place hire a separate driving firm and outsource his drives jobs to that firm, which has much better technology, etc...
Greg's reply:
There is a hard core pizza case I have seen, but it was about leadership problems; the son of donato's owner was in the class ahead of me. But I have a much more elegant solution and one that you will like better.. Sabermetrics of pizza delivery if you will.. you will be hard pressed not to be tempted to use my strategery with your company..but you certainly will with pizza drivers.. but don't confuse the readers.. lol
It is all about the bottlenecks in the system and the marketing uniqueness.
I might focus my marketing in the more profitable or easier areas
The easiest way to compete though is the bottlenecks. Constraint theory: the drivers are usually the botleneck: getting faster smarter drivers is one way to do it, but then you have to rely on them. You could easily compete at a slightly smaller profit margin by having an extra 2 to 3 drivers in peak times.. Eliminating thew advantage. Since they are effectively var costs and not fixed there is little risk. You augment one or two good ones with a less experienced 2 young ones. You give them all of the easy close ones that are off cycle. Here I can figure it out for you real quick.. you'll love this.
This eliminates the build up. Buffer ahead of the pickups.. Here is the equation for pizza wait time (in the buffer) T= time from order to delivery = expected Cooking time + Tq(waiting time)+p(delivery time
Tq= (p/m)*((u^(sqrt(2*(m+1)-1))/(1-u))*((CVa^2+CVp^2)/2)
U= utilization rate = p/am
p= Average Process time of a pizza delivery
a= average arrival time of a pizza from the cooks (take # of pizzas done by cooks in 1 hour divide divide into 60, and that is a)
m= number of pizza drivers.CV = Coefficient of Variability, it is simply the Std dev/mean
Of CVa is Coef. Of Var for arrival times: you need to make a best guess about how much variability is in a (I would guess it is pretty high, which makes CVa pretty low)
CVp is the Co. of Var for delivery times, I would guess this is moderate: I think you know what a std. deviation is, if you don't I can explain. But it is vary important to the equation, it is why adding drivers is SOOO important to the business in rush times.There you go..
Set T = 25-30 min.. and you win the game.. and you will be more profitable too. .. the reason is you will have increased your throughput.. this increases your profit per hour and increases customer satisfaction. It has other effects too.. like decreasing inventory levels and allowing for better managerial control: meaning better resource allocation:
Basically, if your bosses paid you a consulting fee (and you should get them too in ownership shares or something :) ) I would do this spread sheet for you, and you would know the optimum driver numbers per night.. and more importantly per hour per night.
Also: I would hire people who could double. E.g. a back up cook or phone or manager that would act as a driver when the queue gets to be too long: U= too big..
There you go: Sabermetrics for pizza drivers.. see what 100K and a good GMAT score will get you!.
Gregory
By the way, ask any Harvard or Standford MBA this question and you'll get some lame ass leadership or innovation answer..
Pss you could add a sorter like you said.. but it wouldn't be as effective, because it is hard job and doesn't cover the variability problems, which is your true issue.



Recent Comments