Swedish Lessons

On 10 August 1628, the Vasa set sail from Stockholm. Sweden’s most expensive, heavily armed warship sank on its maiden voyage after less than a nautical mile due to elementary design flaws. Salvaged in 1961, it is now displayed in an impressive museum in the Swedish capital I visited over Easter.

If the Vasa represents a stunning fiasco, it is hardly an isolated one. Examples of disastrous project execution include the 1976 Montreal Olympics, California's high-speed rail, the Scottish parliament building, the Berlin Brandenburg airport, and the Sydney Opera House. In a corporate context, the sorry pattern repeats itself, only more privately, across projects in IT, R&D, or post-merger integration.

In 'How Big Things Get Done' (2023), Bent Flyvbjerg, an Oxford professor, draws on a database of more than 16,000 large projects across the world. The numbers are sobering: less than 10% of projects come in on budget and on time. Less than 1% deliver on cost, schedule, and promised benefits.

Large undertakings are derailed by a familiar cocktail of psychology and politics: action bias (failure to consider better alternatives), planning fallacy (misestimating parameters), escalation of commitment (doubling down on ill-fated plans), strategic misrepresentation (presenting palatable rather than realistic budgets to win approval), the ego-driven desire to pursue first-of-a-kind projects, and the selection of project managers whose confidence exceeds their competence.

Flyvbjerg’s remedies sound obvious, but the trillions of dollars wasted annually through poor design and execution suggest otherwise. They include taking the time to think deliberately, planning backwards from the desired outcome (i.e., ‘backcasting’), iterating extensively while experimentation is cheap, and appointing project managers with proven expertise and relevant experience.

On the same island in Stockholm where the Vasa rests, the ABBA Museum highlights the power of a successfully engineered megaproject: a global, multi-billion cultural enterprise built on iteration, precision, consistency, timing, discipline, and operational intelligence.

The lessons from Mr. Flyvbjerg and ABBA carry growing relevance for executives: changing a firm through projects is becoming increasingly critical relative to operating it efficiently, as argued in ‘The Project Economy Has Arrived’ (2021). In that context, enterprise AI programs are a new class of megaproject, with many of their implementation challenges illustrated in a new Stanford study (2026).

Unfortunately, cost overruns, delays, and subpar benefits are often treated as natural features of modern organizational life. Once failure is normalized, reckless ambition finds fertile ground.

The Swedes have a word embedded in their cultural operating system: ‘lagom’. It means ‘not too much, not too little.’ In project management, this translates into the right scope, the right ambition, the right number of iterations, and the right speed.

In a world increasingly shaped by large and complex undertakings, lagom represents a way to make project failure the exception rather than the norm.