LEGO: You cannot subtract what you cannot see.
How LEGO named the crisis, built visibility, removed complexity - and only then returned to growth.
This case study examines LEGO's recovery from near insolvency through four leadership behaviours performed in sequence: naming the position, making value visible, subtracting decisively and rebuilding only once the base was stable.
The company that could stop.
Read this one alongside Case Study 04. Kids Company could always find more and could never do less. LEGO faced the same structural problem - commitments accumulating faster than the capacity to carry them - and did the opposite.
In January 2004, LEGO announced the worst result in its history. It had not recorded a loss between 1932 and 1998. By 2003 it was losing roughly DKK 1.8 billion in a year, with sales down around a quarter, and it was close to running out of cash.
What followed is one of the clearest documented examples of subtraction under genuine pressure: parks were sold, product lines were closed, a headquarters was vacated and the number of distinct components was roughly halved.
Growth followed subtraction. It did not replace it.
The method, stated before we start
We are judging the reasoning, not the recovery.
LEGO went on to become the largest toy company in the world, which makes every decision taken in 2004 look inevitable. It was not. The same programme, with a different Christmas or a different licensing deal, could have become a footnote.
What we can examine is what the leadership did, in what order, and on what information - because that is the only part another organisation can copy.
Years without recording a loss before the crisis.
Fall in net sales in a single year.
Age of the chief executive appointed in October 2004.
Non-family chief executive in the company's history.
What happened.
Watch the order. Naming, then seeing, then cutting - and only then building.
Expansion in every direction at once.
Theme parks, clothing, watches, video games, television, action figures and a rapidly growing catalogue of new components. Each initiative was individually defensible. Together they built an organisation carrying far more than its core business could support.
Somebody says it plainly.
A young head of strategic development, recruited from consulting two years earlier, told the board that the company was on a burning platform and might not survive. He was not senior, and the board included the owning family whose strategy he was describing.
The position is made public.
The company announced its worst-ever result and a liquidity problem. There was no attempt to present the year as a blip.
The instrument is built.
A new finance director joined and established something the company had not previously possessed: profitability analysis by product line. Until this point there were accounts by country, but no reliable view of which products actually made money.
The messenger is promoted.
The person who delivered the assessment became chief executive, aged 35 - the first from outside the family. The founder's grandson stepped back to become chairman.
The subtraction.
A controlling stake in the theme parks was sold. The clothing business closed. Video games moved to licensing rather than in-house development. Publishing was shut. Management left the headquarters building for offices at a factory. The component count was cut by roughly forty per cent.
Then, and only then, growth.
With the base stable, the company expanded again through licensed franchises, films and new formats. By the middle of the next decade, LEGO was the largest toy company in the world by revenue.
The growth everybody remembers began after two years in which the dominant activity was closing things down.
Most turnaround plans invert this, because subtraction is unpopular and new initiatives are easier to announce.
The fact that reframes the whole case.
The popular account is that LEGO diversified too far. True, but incomplete - and the incomplete version leads to the wrong lesson.
LEGO could not tell which of its products made money. It had profit and loss accounts by country. It did not have reliable profitability analysis by product or by line.
The organisation could therefore see that it was losing money in total, while remaining unable to see where the value was being created or consumed.
What the organisation could and could not see before 2004
One institution. Two very different views.
Profit and loss accounts were maintained.
LEGO could see overall performance geographically. It knew where revenue and costs sat across markets and could see that the company was losing money in aggregate.
VisibleWhich lines earned and which consumed was not visible.
Products, components and activities could not be assessed reliably according to whether they created value or absorbed it. Proposed cuts therefore remained matters of argument and opinion.
Not visibleAn organisation in this position cannot subtract intelligently, because every proposed cut is a matter of opinion. Building the measurement was therefore not an accounting exercise. It was the precondition for every decision that followed.
Why this matters more than the diversification
Complexity is only fatal when it is invisible.
Plenty of companies run theme parks, clothing lines and video games profitably. What LEGO had was not too many activities in the abstract - it had more activities than its instruments could account for.
The failure was not simply the expansion. It was expanding past the edge of what the organisation could measure, and then continuing.
You cannot make disciplined choices about what to remove until the institution can see what it is carrying.
The diagnosis
Four behaviours. The order mattered.
Each of these behaviours is ordinary on its own. The sequence is what makes the case worth studying.
Name it
Say the position plainly.
See it
Build the measurement.
Cut it
Subtract, visibly.
Then build
Only once the business was stable.
What most organisations do instead
Soften the position, announce a growth initiative and defer the cutting until the measurement problem has been solved by someone else, later.
Naming it first is what makes the rest permissible. Nobody accepts two years of subtraction on the basis of a difficult quarter.
The first behaviour is the hardest and the cheapest. Describing the position accurately costs nothing financially and everything reputationally, particularly when the strategy being described belongs to the people you are describing it to.
He told the board the company might not survive. They made him chief executive.
The diagnosis, continued
What actually went
Subtraction is easy to endorse and hard to do. This is the scale of it.
The theme parks
The controlling stake was sold in 2005, with a minority holding retained.
The clothing business
Closed.
In-house video games
Replaced by licensing to a specialist.
Publishing operations
Shut down.
The headquarters
The building was vacated and management moved to a factory site.
Around 40% of components
Roughly 12,400 distinct elements were reduced to about 7,000.
They subtracted from the thing they loved
Selling theme parks is a financial decision. Halving the number of distinct bricks is a decision about the product itself, taken against the instincts of every designer in the building. It constrained what could be created.
Most organisations cut the periphery and protect the core. LEGO applied the discipline to the core as well - which is what made the difference, because the complexity lived there.
They centralised to do it
The turnaround was run tightly from the top. Designers lost latitude they had previously enjoyed, and decisions moved upward rather than outward.
That sits awkwardly with Distributed Ownership, and it should be stated rather than smoothed over. A crisis appears to be one of the few conditions under which concentrating decisions is the right move - provided it is temporary, and provided somebody says so at the time.
Against the model
What was present.
The Shortboard model describes twelve attributes built across three cumulative waves. A multi-year programme evidences more than a single decision can. In LEGO's record, three Wave One attributes are clearly visible. The fourth was deliberately suspended - which is a finding rather than an omission.
Lean
dependability
Scrappy resourcefulness
Healthy conflict
Ruthless consolidation
Distributed ownership
New growth
Continuous reinvention
Structural fluidity
Strategic optionality
Platform thinking
Perpetual relevance
Pioneer sanctuaries
The awe-driven mindset
The pioneer's leap
Supply-driven optionality
The sequencing point
Wave One was rebuilt before Wave Two resumed
The expansion that made LEGO the largest toy company in the world came after the floor was repaired, not instead of repairing it. The licensed franchises, films and new formats all sit on a base that had been deliberately simplified first.
This is the cumulative logic running forwards rather than failing backwards - and it is the only case in this series where we can watch it happen in the right order.
Against the model, continued
What each one looked like in practice.
Described as behaviours rather than qualities, because behaviours are what can be copied.
Ruthless consolidation
Not one dramatic cut but a sustained programme of removal across two years, applied to the periphery and the core alike. Crucially, each decision rested on the new line-level view rather than on seniority or sentiment. The parks were sold while still successful, because success is not the test - fit is.
Healthy conflict
A relatively junior person told an owner-family board that their strategy had brought the company close to failure, and was promoted for it. Compare Case Study 12, where the person who documented an uncomfortable finding was dismissed. Both organisations set their culture with a single visible decision.
Scrappy resourcefulness
Building product-line profitability where none had existed, then using it. Moving management out of a headquarters building into a factory. Licensing video game development rather than funding it. Each is a choice to do more with a narrower base rather than to spend into the problem.
Distributed ownership
Decisions moved upward and designers lost autonomy. It was later restored as the business stabilised, but during the turnaround this attribute was consciously traded away. The honest reading is that it was probably the right trade, made explicitly.
Present, and one traded on purpose
The attributes are not commandments. They are capabilities, and a capability can be set aside knowingly for a period.
The distinction that matters is between suspending one deliberately and losing one without noticing. Every failure in this series is the second kind.
The aftermath
What they got wrong.
A success study without this section is marketing. The most instructive error sits inside the turnaround itself.
Manufacturing was outsourced
As part of simplifying the business, LEGO outsourced a large part of its manufacturing to a contract manufacturer. It did not work. Within a few years, the company reversed the decision and brought production back in-house at considerable cost.
The subtraction had gone too far.
Removing the clothing line
Closing the clothing business reduced the number of things the organisation had to be good at. It narrowed the company around what mattered.
Removing manufacturing
Outsourcing manufacturing removed something LEGO had to be good at to exist. Subtraction and outsourcing are not the same thing. The difference is whether the activity is load-bearing.
What happened next
The floor is repaired
The parks were sold, lines were closed and the number of components was almost halved.
A subtraction too far
Manufacturing was outsourced, then brought back in-house at considerable cost.
Sustained growth
By 2015, LEGO had become the largest toy company in the world by revenue.
Complexity returns
LEGO recorded its first revenue decline in more than a decade and cut around 1,400 roles.
That last panel is the one to sit with. Thirteen years after the crisis, with the same discipline available and much of the same leadership in place, the organisation had accumulated enough complexity to need another correction.
Subtraction is a habit, not an event
LEGO did not solve complexity in 2005. It cleared it, and then it built up again, because building up is what functioning organisations do.
Treat any major simplification as buying a decade, not fixing a problem. The organisations that avoid the second crisis are the ones that make removal a routine rather than a rescue.
The conclusion
Measurement first. Then the knife.
You cannot subtract what you cannot see.
Most simplification programmes begin with a target - twenty per cent fewer products, ten per cent fewer suppliers - and then fight about which ones.
Without evidence showing what each activity contributes, the loudest sponsor keeps their initiative and the quietest loses theirs.
LEGO built the instrument first and let it decide. That was unglamorous, it took months, and it is why the cutting that followed was accepted rather than resisted.
Making reality visible is the first act, not the second.
Three questions for your own team.
Can we show, line by line, which parts of what we do actually earn - or only which countries and departments?
What have we added in the last three years, and what did we remove to make room for it?
When someone last told this leadership team something it did not want to hear, what happened to them afterwards?
Find the edge of what you can measure
Identify the point at which your reporting stops resolving. For most organisations, it is well inside the business - you know revenue by client and cost by department, but nothing reliable about the intersection.
Everything beyond that edge is being run on opinion. That is where complexity accumulates, because nothing there can be proved not to be working.
The Shortboard lesson
Visibility before subtraction. Stability before growth.
LEGO did not begin its recovery with a new growth strategy. It began by making the existing business visible. Only then could it determine what earned its place, what created hidden complexity and what had to go. The turnaround worked because the organisation repaired Wave One before asking Wave Two to carry it again.
Ruthless consolidation begins with measurement
You cannot remove complexity by opinion. You must first make its contribution visible.
Build the instrument first
LEGO could see revenue by country and cost by department, but it could not see whether individual products created or destroyed value. Building product-level profitability made the later decisions evidential rather than political.
Apply discipline to the core
Selling theme parks and closing peripheral businesses mattered. Reducing the number of distinct bricks mattered more. The organisation was willing to simplify the product it loved because that was where the complexity actually lived.
Sequence is a capability
Name the position. Make it visible. Subtract what no longer fits. Then build again. The behaviours were not individually unusual. Their value came from doing them in the correct order.
Temporary control needs an expiry
Decision-making was centralised during the crisis and designers lost autonomy. That trade may have been necessary, but only because it was made consciously and later reversed. A capability can be suspended deliberately. It should not disappear unnoticed.
The leadership test
Find the edge of what you can measure, then decide what no longer earns its place.
The evidence
How we know this. And what we don't.
The evidence base is weaker than the failure studies in this series, and the conclusions should be read accordingly.
What makes this strong
- The company announced its crisis publicly and in detail, including its liquidity position, rather than presenting a bad year as an anomaly. That disclosure was contemporaneous.
- The divestments and closures are matters of public record, with dates, counterparties and financial sums attached.
- The component-reduction figures come from independent research conducted with company access, rather than from company marketing.
What to hold lightly
- There is no inquiry, judgment or regulator. Nothing in the available record was compelled or cross-examined.
- Much of the narrative rests on accounts given by the executives involved, after the recovery, through interviews and case studies. They are describing a success they led.
- Financial figures vary between sources and currencies. The figures used in this case study are therefore stated as approximate.
The recovery had other causes too
Licensed franchises, a strong period for construction toys, films and the growth of adult buyers all contributed to what followed. It would be wrong to attribute a decade of growth to a two-year simplification programme.
What the record supports is narrower and more useful: the sequence of behaviours between 2003 and 2005, and the fact that the organisation built its measurement before it made its cuts.
Sources
LEGO Group annual reports and press releases, 2003-2006, including the January 2004 announcement of the year's result; David Robertson with Bill Breen, Brick by Brick, 2013, based on research conducted with company access; published interviews with the chief executive and finance director covering the turnaround period; and contemporaneous reporting of the 2005 theme park divestment and the later reversal of manufacturing outsourcing.
He told the board they might not survive. They made him chief executive.