Models Aren't Just for Airplanes
April 2010
The strange thing about so many businesses
is that the successful and unsuccessful both start with the same entrepreneurial
zeal to look for something unique in the way of an opportunity. The truly persistent
(and, they'll admit, lucky) win not only
because they focus on the one thing that makes them successful, but also by
accumulating and protecting profitability.
by Adam
Webster
Marcil Technology
Group, Inc.
Shoring up gains after getting them is part of all growth. It also worked
on D-day and countless other parallels outside the world of business. But as
risk managers, we know that protecting the brand, the asset(s), the revenue
pipeline, that's the unequivocal mandate.
The innovators come up with the idea, but the risk mitigators figure out
how to protect the success of such ideas from icebergs, meteors, and just plain
old inevitable human behavior.
Doing It Backward
But what if the risk mitigator could be involved in the beginning, playing
the role of the polite but welcome devil's advocate? Every now and then, aviation,
manufacturing and even consumer services businesses are born via an idea that
is tested in simulation. Google, while not a strict parallel, shows how an idea
("Hey, wouldn't it be great if ...") goes into the lab for lots of work and
only gets released after its value and scalability are proven. (As an aside,
Google's original name was "Backrub" just to drive home how casual it might
have been on day one.)
The purpose of modeling is that you can have a nice $50,000 or $100,000 catastrophe
with the lab, before delving into a $6 million investment. Virtually every investment
in private aviation has tanked in a spectacular way. Why? Think of every fractional,
jet card, charter broker, and air taxi startup. Close to 100 percent fail out
of the gate, yet investors (or sometimes just desperate aircraft manufacturers)
keep lining up for another flogging.
Airplane people, fueled by emotions and characters, rarely have time to stop
and take a look at things that keep airplanes in the air—money. Even in the
cases of initially successful models, once things do get going, there is the
temptation to change the plan and order a new (second) kind of airplane, enter
a new market or do something far outside their zone of safety, experience and
... profitability. (Compare Jetblue with Southwest's story to see how mutually
smart beginnings can lead to divergent paths.)
Even the mighty Berkshire Hathaway has suffered under this dynamic and continues
to struggle
today. Tens of billions of dollars later, there is a $157 million cumulative
loss on the books. That is an unusually bad story. Even more unusual when you
consider who is at the helm—the king of all risk managers.
So, before capital investments and other substantial commitments are made,
why not run a random system test with computer simulation? Chaos theory geeks
call this modeling. And when you see what it is and how
it works, you realize that you really can't tell the story of a new business
on a spreadsheet.
Start with a Model
Here are some upsides.
#1 Everything, All at Once: Computers today
can model personalities, weather, machines, behavioral conditions ...
all at the same time. You have 1,000 customers,
7 of which will sue for cold vegetables in their meals, 100 of which will help
open up new opportunities, and the bulk of whom will pay their bills on time
and exhibit fairly predictable behavior. A well-run, autonomous agent bidding
system can give you meaningful data about what you need to be charging. This,
incidentally, is the central tenant to using chaos theory to better business.
#2 Adapt and Live: You can feed real data
back into the beast once you actually start a new profit center or business.
This data then allows you to tweak (subtly or dramatically) your original assumptions.
Risk mitigation means knowing where you've been, where you are now, and where
you're going. If a pilot ever loses track of any combination of those three
things, they typically die and kill the passengers too. Why not run a business
with the same zeal for security?
#3 Gee Whiz Bang Effect: Running the
simulation can be done in text only (boring) or have cool visual effects to
dazzle your superiors with since they won't read anything anyway. Nothing makes
your day like showing a flash animation or YouTube video of something that doesn't
yet exist making money for the gang that will ultimately control the purse strings.
Here is the only downside: You'll have a sobriety check before you launch
the business, thereby saving millions but learning you will be unable to launch
your dream. (Sobriety, to entrepreneurs, can be a downer.) The problem with
most entrepreneurs is that they are creative, but also stubborn individuals
who "know they are right" and when they "feel that something will work," they
will move mountains (and manipulate data to the board) to make their point.
Beware of such a beast that has a hard time listening.
What Now?
How to integrate this friendly rant into your day as risk manager? Print
it, keep it handy, and repeat to yourself: Sobriety is okay. Testing is good.
Computers are our friends. Spending a little before we lose a lot won't make
me a hero, but it will certainly make me more trustworthy.
Opinions expressed in Expert Commentary articles are those of the author and are
not necessarily held by the author's employer or IRMI. Expert Commentary articles
and other IRMI Online content do not purport to provide legal, accounting, or other
professional advice or opinion. If such advice is needed, consult with your attorney,
accountant, or other qualified adviser.