envy. He led a group that produced software used in large
commercial aircraft. At first it seemed his team’s graph might
as well have been a connect-the-dots drawing of a cloud.
Over time, with some adjustments to their methodology, that cloud collapsed into a relatively straight line. If the
project team said it would take a certain amount of time or
money, it typically delivered within 5 to 10 percent.
The customer displayed the virtue of kindness and spent
most of a day explaining his company’s process-improvement
plan. We incorporated much of it into our own methodology.
When it became clear we could not meet a goal we had
committed to, we used to simply shift our commitment
to a target we could reach. Rather than be up-front about
our ineffectiveness and use our metrics as an opportunity
to improve, we rebaselined to make ourselves look good. At
the end of each year, we were able to say we had made 90
percent of our targets.
A dead giveaway of such misleading measurements is
that you’re looking at things from your point of view and
not your customers’.
In the Inferno, the proud were forced to walk with stone
slabs bearing down on their backs to induce feelings of humility. Our stone slabs were that our estimates routinely were
embarrassingly off-target. So we’ve reworked our estimating
process several times. At last count, we were on version 15. 1.
This sin actually derives from the Latin luxuria, meaning
“an unrestrained excess.” You want something and you
must have it.
V IE WPOIN TS
We met a woman at a conference who was complaining that her manager insisted her team meet every delivery deadline without exception. His end-of-year reviews
were tied directly to success in meeting schedules—and
the project team did indeed meet every deadline. What’s
wrong with that?
The woman told us her team would ship a product on
time—no matter how low its quality was.
Our data show that we can stamp out bugs for as little as
US$60 each if caught early in the process. Once in the field,
we have spent as much as US$15,000 to fix a single defect.
VOICES ON PROJECT MANAGEMENT
Like most people, sometimes we are sinners. Sometimes
we are saints. In both roles, we’ve learned that a good
estimate can mean the difference between a project being
heaven or hell. PM
Howard Smallowitz, PMP,
is an IT architect on outsourcing accounts at IBM,
Austin, Texas, USA. Previously, he was the estimation review board chair for
George Stark is a senior
technical staff member on
the integrated technology
delivery team at IBM in Austin, Texas, USA. He is also a
member of the IBM estimating community core team.
RAISE YOUR VOICE No one knows project management better than you, the practitioners “in the trenches.” So PM Network launched its Voices on Project Management column.
Every month, project managers will share ideas, experiences and opinions on everything from
sustainability to talent management, and all points in between. If you’re interested in contributing,
please send your idea to