In the future how will agile methods be
remembered by the project management community?
It seems history has a way of distorting the facts and simplifying
concepts out of context. Here are a few examples:
1) In the original Waterfall Software Development Process
paper written by Winston Royce in 1970, after presenting the lifecycle diagram
on page two, the author states “I believe
in this concept, but the implementation described above is risky and invites
failure”. Royce then spends the
remaining nine pages outlining feedback loops and “Do It Twice” recommendations
since there would be things missed in the first pass through. Read in its
entirety, it outlines a fairly robust, risk tolerant approach to building
systems that features multiple iterations and opportunities for learning and
adaptation.
Yet, waterfall
is thought by many to be a single pass lifecycle with all the associated
problems. It is as if the project management community latched onto the
lifecycle diagram depicted on page two and chose to ignore all the more
difficult to implement yet critical steps described in pages 2-11. Read the original Waterfall paper here.
2) Henry Gantt’s project management research and
work actually focussed on retrospectives, diagnostics and optimizing work flow.
Yet people remember him for the Gantt chart. The funny thing is that he did not
even invent what we call the Gantt chart today; that was Joseph Priestley 100
years earlier.
Gantt’s charts
were far more complex diagrams that were created after the work was completed
and then used to identify waste and improve the process. While he started with
scientific management of Fredric Taylor, his research went far beyond charting
and optimizing labor. He was a harsh critic of ineffective management and
promoted many Lean and Theory of Constraint like values such as “Increased production not through speeding
up workmen but by removing the obstacles which prevent them from doing their
work”, “Reduced costs, because of the elimination of idleness and waste as well
as improvements in process”, “Men interested in their work not only because of
the wages but because they have an opportunity to increase their knowledge and
improve their skills.
We often
characterize Gantt with tracking charts, but that’s a faulty summary of a
talented systems thinker; for more information on Henry Gantt, his real charts
and work see Henry
L Gantt 1861-1919 Debunking the Myths.
3)
The Manhattan project is often attributed as the
origin of modern project management and phase gated approaches, however it
actually pursued concurrent development. The Manhattan project to develop a
nuclear bomb in the 1940s, certainly displayed the modern project management
principles of organization and planning, but also high rates of trial-and-error
and multiple parallel streams attempting to solve the same problem.
There was just
so much uncertainty surrounding how to create a bomb, knowledge was theoretical
and incomplete. The project manager, Leslie Groves, said: “The whole endeavour was founded on possibilities rather than
probabilities. Of theory there was a great deal, of proven knowledge, not much.
There was simply no ready solution to the problem we faced.” So, Groves and
his steering committee decided to explore and implement different solutions in
parallel. This approach (well multiple approaches) and willingness to modify
and add solutions mid-course, led to technical breakthroughs that had been
thought impossible by most three years before.
This was 30
years prior to Toyota’s “Set based engineering” but very similar in its pursuit
of multiple parallel approaches. Yet today we most often hear of the Manhattan
project as the birth of the phase gate approach, this is too bad, they did so
much more.
4)
Lastly, the Polaris project that was attributed
as the origin of Critical Path Method (CPM) and PERT was actually about gaining
First-To-Market intellectual property share.
The Polaris project developed the first submarine-launched ballistic
missile (SLBM) carrying nuclear warheads. These offensive weapons, almost
impossible to track and attack, became a key element of nuclear deterrence.
The Polaris
project is today credited with developing the “scientific approach to project
management” with the first large scale application of computerized planning
techniques, particularly CPM and PERT. A big part of the project was about
“getting a share of the ballistic missile pie” away from the newly formed US
Air Force that was receiving most of the Pentogon’s money. Admiral Burke
astutely believed that “The first service
that demonstrates a capability for this is very likely to continue the project
and others may very well drop out”. The result was a clear prioritization
of schedule over cost and specification; and a willingness to experiment and
change specifications over the course of the project.
Trial and error
led to two deployed tests in the early 1960’s and PERT served “…less for improving project control than
for offering technological pizzazz that was valuable in selling the project. (Since) The image of efficiency helped the project.
It mattered not whether parts of the system functioned or even existed, it
mattered only that certain people for a certain period of time believed they
did.” For more information on the fascinating truth of the concurrent
development, experimentation, iteration, and adaptation that really underpinned
the Manhattan and Polaris projects see Lost
Roots: How Project Management Settled on the Phased Approach (and lost its
ability to lead change in modern enterprises).
So, given waterfall was iterative, Gantt
focused more on the theory of constraints not Gantt charts, the Manhattan
project was Lean not Phase Gated, and the Polaris project was about rapidly
gaining mindshare through iteration, I don’t really hold
out much hope for agile methods to be remembered accurately in the future.
The glass half-full view of these history
lessons shows that smart, resourceful people have been tackling complex project
problems for centuries and our ideas such as lean-startup, Kanban, behaviour
driven development, etc are likely not new.
The glass half-empty view is that agile
methods will be erroneously summarized to tangential concepts, such as “individuals without tools”, or
“leave no documentation”. However, smart people will continue to be successful
in managing challenging, audacious projects and terms are really just labels
that matter less than the methods they describe or results they enable.
Maybe we have better collective knowledge
these days and we will not repeat the same erroneous summarization and labeling
of take away ideas. Lenfle and Loch, the authors of the “Lost Roots…” paper,
assert that the loss of trial-and-error and strategy-making focus from popular
project management, that was clearly present on these early projects, restricts
the usefulness of project management today. Agile and lean approaches
rediscovered this science and are attempting to merge it back into mainstream
project management, but are up against a generation of resistance from those
who were taught projects can plan out variability.
From an evolutionary perspective it is an encouraging
sign that techniques that were lost through poor reporting have re-emerged
elsewhere to exploit project environments that have high levels of uncertainty.
Regardless of their names or origins, let’s hope they persist and get
incorporated in mainstream project management, not to be lost again.
This article first appeared in ProjectMangement.com here. Bio: Mike Griffiths is a
project manager experienced in traditional and agile methods who reads about
project management much too much - you really don’t want to get stuck with him
at a party!