Getting major projects right: 5 recommendations for organizations of all kinds (Part 2)
Last week (Part 1), I presented the research findings of Atif Ansar and Bent Flyvbjerg from the University of Oxford. The authors argue and prove that major challenges can be implemented more efficiently with a platform strategy than with one-off projects.
What would a paper be without concrete recommendations? First of all, the authors confirm what was already clear at the beginning: platform projects are faster, better, cheaper and less risky than conventional large-scale projects. More importantly, the authors prove that their findings are transferable to other industries.
In the following, I present the authors' recommendations, which are equally applicable to private companies, corporations, authorities or governments.
Update: What is a major project?
To Part 1 of this article I received the feedback that "major project" is a rather vague term. The I have subsequently concretized.
The "magic triangle" is invalid
Platforms challenge the traditional myth of the Magic Triangle in project management, which states that scope, cost and schedule are linked in a zero-sum game. Platforms like SpaceX are disproving this conventional thinking by delivering greater scope (volume and variety of space cargo) at a tenfold lower cost factor and twice the pace of traditional methods.
Quantity and variety
Conventional thinking envisions a trade-off between scale and diversity. NASA's one-off missions were highly diverse - the doomed Mars Observer carried eight instruments for nearly $500 million worth of scientific discoveries. Even if the Mars Observer had been successful, its continuation would have taken just as much time and money to redesign each individual element. In contrast, platforms are designed for repeatability through their modularity.
The initial advantage of this repeatability is the increase in homogeneous volume. Once this repeatability is predictable, diversity follows at scale by combining modules in different ways. In a virtuous cycle, diversity releases more volume. For example, SpaceX first achieved repeatable and reliable launches of its Falcon 9 rocket, which led to significant cost reductions, which in turn enabled lower prices, which boosted demand, e.g. from owners of small, low-cost satellites. Then SpaceX began to bundle payloads of different sizes to further reduce costs through the aforementioned "rideshares". The authors have visualized this:
Short iterations and rapid failure
The authors emphasize the importance of platforms in the space and how they offer exponential growth benefits. The focus is initially on the learning effect, which is more important than scaling.
Especially in the development phase, failures are not only normal, but an important part of the learning process. Platforms are able to consolidate and utilize this knowledge in a structured form.
SpaceX demonstrates how experimental learning can be used effectively. The company started with basic hypotheses and tested them in planned experiments.
Failures are used to test hypotheses and make improvements. This approach leads to positive learning curves in which activities become faster and more cost-effective the more often they are carried out.
The NASA Mars Observer Mission is a counter-example. Instead of taking several steps with milestones, the project consisted of a single, 17-year-long step - which failed. Authorities are often afraid of small mistakes and therefore prefer to focus on big, slow progress. But this can lead to greater failure.
The authors praise NASA for the "Faster, Better, Cheaper (FBC)"approach. However, fear of failure has led the organization to revert to old patterns. But here's the surprising twist: SpaceX's success is ultimately also a credit to NASA, as the space agency has supported their approach, which allows for experimental learning.
The role of systems engineering
The recommendations presented here are aimed at decision-makers and the management level of organizations. However, systems engineering plays an essential role in the implementation of these recommendations. In particular: Without systems engineering, these recommendations cannot be implemented anyway.
In particular: Correctly practiced systems engineering makes it possible to process what has been learned in the right places in development. This works particularly well when Agility with more traditional concepts such as Traceability be united.
Conclusion
Neither the findings nor the recommendations in this paper are really new to readers of SE-Trends. But this paper is nevertheless important: it gives decision-makers a scientifically sound basis for making decisions about introducing a platform strategy in their organizations. These are strategic, far-reaching changes for many organizations. It also makes recommendations that decision-makers can understand. We, the systems engineers, are then responsible for implementation.
