Skip to content


TODO: Summary

The standard approach for large projects, such as nuclear plants, ERP implementations, data-warehouses and subway systems, is to build monolithical bespoke systems. Subsequently, large investments need to be taken to even start the project and the project won’t deliver, in most cases, any value until completion. Even a project that is 95% complete, is often of no use to anyone until it is delivered, such as in the case of a nuclear plant. This problem also increases the risk that the project may get out of hand by requiring more time and budget while also failing on fulfilling the original project expectations. Additionally, this approach offers little to no scalability potential.

As a far more compelling approach, such large projects should, according to Flyvbjerg, be divided in smaller subprojects that can be tackled separately and where each should deliver value independently of the rest.

Prof. Bent Flyvbjerg

Flyvbjerg Prof. Bent Flyvbjerg
Professor Bent Flyvbjerg is the Executive Chairman of Oxford Global Projects. He has 30+ years of experience as an advisor to government and business, including the US, UK, and Chinese governments, Fortune 500 companies, and top consultancies. Flyvbjerg is the most cited scholar in the world on project management. He is the first BT Professor and Inaugural Chair of Major Programme Management at Oxford University and holds the Villum Kann Rasmussen Professorship and Chair at the IT University of Copenhagen. He has received numerous honors and awards for his work on better project management, including a knighthood and two Fulbright Scholarships.

Flyvbjerg uses Madrid’s subway system as a mere example to illustrate the value of this approach. Manuel Melis Maynar was the person responsible for Madrid subway extension which added 131 km of rail, 79 new stations. This project started in 1995 and ended in 2003. Nonetheless, this project was and is still considered as a success story as it set records in terms of its speed and costs compared to other similar projects took up to 10 or 20 years and cost much more that initially expected. This radical difference stems from a number of decisions that Melis took before starting this project:

  • No monuments: Monuments and other forms of signature architecture can be found in many subway stations but they are also associated with cost overruns and delays. Therefore, Manuel decided not to include any in its 79 stations and each station may follow a predefined replicable modular design based on well-proved construction methods.
  • No new technology: Manuel avoided relying on new construction techniques, tools or methods as they could add risk to the project. Instead, he relied on well-proved techniques, tools and methods that had proved to deliver results fastly, cheaply, safely, and at a high level of quality.
  • Speed: Time is like a window as the bigger it is, the more bad stuff can fly through it. Manuel understood this as he determined that the project needed to be delivered in incremental steps of 3-6 kilometers. This may enable workers and him to get feedback after each step and to foster positive learning. In other projects, this may also make possible to deliver value to customers as soon as possible.

Using this mega-project and other successful ones such Tesla’s Gigafactory, Flyvbjerg identified the principles behind scaling up in a smart way. By smart way, he means starting from small beginning to a large project through experimentation and relying on essentially identical modules. The opposite of this is to start on a massive scale with a project that mainly uses bespoke components. According to Flyvberg, the principles behind smart scale-ups are the following:

  • Speed: As stated by former Google’s CEO, Eric Schmidt; “Create a product, ship it, see how it does, design and implement improvements, and push it back out. Ship and iterate. The companies that are the fastest at this process will win”. Speed is critical for companies not only to create momentum but also to create a feedback loop between the company and its environment to foster positive learningpositive learning within companies. Positive learning is defined by Flyvbjerg as; “learning to do things faster, better, and cheaper by doing the same thing over and over, through replicated modules”. Additionally, tighter time frames result in more controllable chunks of work whose budget and other estimates are easier to calculate and are more reliable.
  • Replicable modularity: By relying on rather high-quality identical modules that can be assembled into a larger project, project managers can foster positive learning that will result in performance increases, lower project management risks and in meeting the project expectations. Additionally, working incremental steps also offers delivering customer value without having to complete the whole project first. For example, Tesla’s Gigafactory consists of different blocks that can be operated once completed and that can start producing batteries long before the whole Gigafactory has been finished. Lastly, relying on modules, whose value and workings have been tested before, significantly lowers the risks associated to bespoke solutions.

Although these concepts are not new to many engineering fields, they are rather new in the business world. However, they have been gaining attention recently by consulting companies such as Gartner with the so-called Composable Business which means; “creating an organization made from interchangeable building blocks.” As stated by Gartner; “A composable business is architected for real-time adaptability and resilience in the face of uncertainty.” In turn, other companies such as WalkMe, Elastic Path and Venture Leap consider packaged business capabilities as the building blocks of composable businesses.


Author: Carlos Tubbax