NASA is clearly slower to market and to iterate than SpaceX. — By Atif Ansar and Bent Flyvbjerg In terms of speed-to-market and iterative updates, SpaceX is on average nearly twice as fast as NASA. In terms of actual delivery time, SpaceX projects have a shorter actual duration from the decision to build to completion.† On average SpaceX projects took 49.2 months…

Now that Boris Johnson is leaving office it’s time to think back and find our favorite stories from his premiership. Here’s mine. What’s yours? — I’m Danish but have worked in Britain for the past 14 years. I was there when Covid-19 hit the country in early 2020. I decided to stay, but resolved to take my cues on how to deal with the pandemic from Denmark.* Smart move, if you ask me. On March…

Elon Musk decided long ago that he cannot afford long projects and therefore also cannot afford to use conventional design and delivery methods, which are notorious for being sluggish. We illustrate Musk’s pursuit of speed by contrasting SpaceX with NASA. Tesla also gets mention. — By Atif Ansar and Bent Flyvbjerg In terms of speed, although SpaceX is significantly quicker in absolute speed-to-market, it is just as optimistic about its schedules as NASA. The figure below shows optimism measured by schedule overruns (delays)of NASA missions compared against SpaceX.

SpaceX is fast, cheap, and scalable. NASA is slow, expensive, and bespoke. — By Atif Ansar and Bent Flyvbjerg What Is Scalability? We define scalability as the ability to handle an increased volume and variety of workload without crashing or needing a disproportionate number of additional resources. The concept is borrowed from computer science and is distinct from conventional notions of economies of scale in economics…

Dumb scale-up is bad. Forced scale-up is worse because it wastes resources faster. Fumble scale-up, and you die. — I covered smart scale-up and dumb scale-up in two previous articles. Here I focus on two further types of scale-up: Forced and fumbled. Forced scale-up designates the combination of a one-off, bespoke design and its high-speed delivery. This will typically result in a product of low quality, because it is…

SpaceX is 10X cheaper with 30X lower cost overrun than NASA in lifting payload into space. Why? Because SpaceX is platform-based, NASA not. — By Atif Ansar and Bent Flyvbjerg “The space race is dominated by new contenders,” claims The Economist (18 October 2018).† Chief among them is SpaceX that has gone from a ridiculed upstart in 2002, when Elon Musk founded it, to capturing nearly two-thirds of the global commercial launch market by…

Platforms make services faster, better, cheaper, and more omnipresent. Platforms are based on modularity and are the antithesis to big, one-off, bespoke capital investments. — By Atif Ansar and Bent Flyvbjerg Platforms start small but become big by growing incrementally, i.e., with no big leaps involved. It is not that platforms are less ambitious than quantum leaps. In fact, once a platform matures, it may easily get exponentially bigger than anything thought possible by a…

How does one best deliver big, ambitious projects in business and government? — Atif Ansar and Bent Flyvbjerg The public sector, even in market economies such as the US or the UK, accounts for nearly half the share of national income (Offer, 2003).† Public spending has grown over the past half a century, and so have major public projects. Public spending on long-term…

Dumb scale-up is one-off, bespoke, and slow. It correlates with delays, cost overruns, and benefit shortfalls. Any business, government, or NGO that depends on dumb scale-up is unlikely to be successful. Smart scale-up is the answer. — Nuclear power is an obvious example of dumb scale-up, almost embarrassingly so. This is because both the technology and the economics of nuclear are so evidently flawed that the deficiencies are impossible to overlook.