| |
So what do you do? You had to experiment to find out what makes
a mighty tasty pizza. In business terms, this is known as Causes and
Effects (in the world of pizza: it is called ingredients and taste).
Sometimes it is easy to link the a cause to an effect - you run up
25 flights of stairs: you get out of breath. Other times things are
a bit more mysterious: the blue screen of death on your computer seems
to appear for no apparent reason. In the world of pizza, we know that
tomato sauce and cheese usually go well together, but we are not too
sure whether kumquats and chickpeas will work out. So you experiment...
Experiments allow you to create (or recreate) situations in a controlled
fashion. It is a situation that you put together to study and analyse
the problem in a controlled environment. Sometimes these are called
scenarios, role playing or simulations. It is an area of business loaded
with jargon, but we reckon it is actually nearer to pizza making than
rocket science. There are several steps to this procedure.
Experimenting
in Business
|
|
Recreating
the Pizza
|
 |
1. Define Scope
Before you start, you define the limit of your investigation. This scope
therefore gives you a set of practical limits to work with.
|
|
1. In
the case of a pizza experiment, you can set limits: vegetarian
or meat and vegetables, fresh herbs or dried herbs and so on.
|
2. Search for Causes
You then gather your information, look for causes, reasonable causes,
probable causes or even possible causes. After collecting a mass of data,
you are about to start investigating.
|
|
2. In the pizza experiment, you gather
reasonable / probable / possible ingredients.
Reasonable: tomato, basil, cheese...
Possible: seaweed, pine nuts...
|
3. Testing Assumptions
You analyse by testing your assumptions. The assumptions will allow you
to categorise the issues broadly, look at the data and try to make some
sense.
|
|
3. For example, you remember the pizza
being savoury, so you assume no sweet ingredients. Or you
remember the pizza was red- so you look through your red
ingredients: tomatoes, peppers, chiles etc.
|
4. Relating Causes to Effects
You link the assumptions to any observations such as: "A will lead to
B". The extent that this actually happens is known as "Correlation".
For example, oil on the walkway is likely to be a cause of
someone slipping.
|
|
4. Tomato with basil combination gives
good pizza taste: a "perfect fit" has a correlation
of "1".
Turnips with pickled cabbage combination is less desirable and unlikely
to work as a good pizza : the "no-fit" has a correlation of "0".
|
5. Designing Experiments
Up to now, it has been assumptions and data, you are now ready to carry
out your experiment.
"Design of Experiments" is a term used in analysis to describe
a technique where groups of factors are combined to minimise
the number of experiments.
|
|
5. Making a pizza is a good example of
Design of Experiments - you can try high correlation combinations
of ingredients: tomato and cheese with basil or sausage
with peppers and mushrooms on pizzas, rather than trying
a much large number of pizzas each with one ingredient
only. (See table below).
|
In a business situation, you need to list the relevant variables that
you derived from your analysis. These are then assessed as to the causes
and effects. You then analyse them for the correlation to each other.
From there you will have a structured set of factors to conduct the
necessary experiments.
Example of Design
of Experiments - Designing the Ultimate Pizza
|
2 Causes: (tomatoes, olives)
|
A
|
B
|
Experiments: 8
|
2 Observations: (cheese, peppers)
|
X
|
Y
|
Combinations
A.X.H B.X.H
A.X.L B.X.L
A.Y.H B.Y.H
A.Y.L B.Y.L
|
2 Levels: (sausages, mushrooms)
|
H
|
L
|
In pizza experimenting, we would recommend taking the same approach but with
the addition of having some indigestion medicine ready to hand if your Design
of Experiments analysis suggests anything more than 2-3 pizzas - Enjoy!
|
|
