That sounds contradictory.
It isn’t.
A distribution function gets outsourced because a specialist can operate it more cheaply.
A warehouse is closed because utilisation doesn’t justify the capital.
An internal capability is removed because an external provider has better scale.
A service function is consolidated because the standalone economics improve.
Each decision can be rational.
Each business case can be correct.
And the economics of the overall business can still deteriorate.
Because businesses don’t compete as collections of individual functions.
They compete as systems.
And one of the things we consistently underestimate is the economic value, and cost, of the handoffs between them.
The economics of the handoff
Most make-versus-buy decisions start in roughly the same place:
What does it cost us?
What would someone else charge?
What capital could we release?
What fixed costs disappear?
What happens to EBITDA?
All necessary questions.
But they measure the activity better than they measure the system around it.
Every time an activity crosses an organisational boundary, something else can enter the equation:
Delay.
Coordination.
Duplicated handling.
Information latency.
Different incentives.
Service variability.
Reduced responsiveness.
Ambiguous accountability.
And another commercial margin somewhere in the chain.
Sometimes none of that matters enough to justify ownership.
An external provider may have genuinely superior scale, capability and economics.
But sometimes the apparent saving in one part of the business simply reappears somewhere else.
Not necessarily as an identifiable cost line.
It may appear as lower utilisation.
More inventory.
Longer lead times.
More expediting.
Lower service reliability.
Lost sales.
Management attention.
Reduced pricing power.
Or an inability to respond when the customer needs something outside the standard process.
The supplier invoice tells you the price of the activity.
It doesn’t necessarily tell you the cost of the handoff.
That distinction matters.
Activity economics are not system economics
Imagine an operating capability that produces an ordinary standalone return.
Looked at independently, the conclusion seems obvious:
Why own it?
But now suppose that capability also:
improves utilisation of another asset,
reduces inventory elsewhere,
removes a recurring bottleneck,
protects a high-value customer relationship,
provides information competitors don’t have,
or enables a service proposition competitors struggle to replicate.
What is its return now?
That is a much harder question.
And it exposes a weakness in the way businesses are often analysed.
We allocate revenue and costs to functions, business units and assets because we need accountability.
But economic value doesn’t always respect those organisational boundaries.
One part of a system can create value somewhere else.
Which means an operation can look mediocre on its own P&L while being economically important to the enterprise.
The reverse is also true.
A function can report attractive economics while creating cost and complexity everywhere around it.
That is why local optimisation is dangerous.
A better function does not necessarily create a better business.
Control has economic value. But it has a price.
This isn’t an argument for vertical integration.
Businesses can own far too much.
Assets consume capital.
Fixed infrastructure reduces flexibility.
Additional capabilities require management attention.
Complexity creates risk.
And underutilised assets can destroy returns very quickly.
“Strategic” is also one of the most convenient words management teams have for defending things they don’t want to change.
So ownership deserves exactly the same scrutiny as outsourcing.
The objective isn’t to own more.
The objective is to control the parts of the operating system where control creates more value than it costs.
That is a much higher standard.
And it changes the question from:
Is this core?
to:
What does controlling this allow the rest of the business to do?
Start with the customer promise
One simple question often exposes the issue:
If this activity fails, who does the customer blame?
Usually not the supplier.
They blame you.
That doesn’t mean you should own every activity that touches a customer.
But it does mean that activities materially affecting reliability, quality, responsiveness or speed deserve to be viewed differently from commodity inputs.
Because sometimes what looks like infrastructure is actually part of the product.
The customer may never see it.
They experience what it allows the business to do.
Six questions for deciding where the business should end
When reviewing whether a capability belongs inside or outside the organisation, I would ask six questions.
1. Does it materially affect the customer promise?
Would losing control materially change reliability, responsiveness, quality or the reason customers choose us?
2. Does controlling it improve economics somewhere else?
Does it improve utilisation, throughput, margin, working capital or productivity elsewhere in the system?
If so, a standalone P&L may be the wrong unit of analysis.
3. What is the economic cost of the handoff?
Not just the supplier price.
Waiting time. Rework. Inventory. Coordination. Expediting. Failure demand. Information latency. Management intervention.
What does the interface actually cost?
4. Does control create an information advantage?
Does owning the activity allow us to see demand sooner, understand customers better, identify problems earlier or make decisions faster?
Information has economic value.
5. What options does control create?
Can we enter a market faster?
Add capacity?
Launch another service?
Serve customers differently?
Respond when competitors cannot?
Optionality is not an excuse for poor returns.
But it belongs in the calculation.
6. Does its value compound as the business grows?
Some capabilities are worth only what they produce directly.
Others increase the productivity or strategic value of everything connected to them.
Those are fundamentally different assets.
The second type can become part of the moat.
This changes the way you look at complexity
Management teams are rightly encouraged to simplify.
Most organisations accumulate complexity they no longer need.
Duplicated systems.
Legacy processes.
Excessive variants.
Unnecessary management layers.
Multiple ways of doing the same thing.
Approval structures whose original purpose disappeared years ago.
That complexity should be attacked.
But complexity itself isn’t the enemy.
Unproductive complexity is.
There is another kind.
Complexity that allows a business to provide a service competitors struggle to replicate.
Complexity that removes a constraint.
Complexity that increases customer switching costs.
Complexity that creates proprietary information.
Complexity that improves the economics of another part of the operating system.
Complexity that becomes more valuable as the business scales.
That is productive complexity.
Removing it can make the business simpler.
And less valuable.
So perhaps the better question isn’t:
Can we simplify this?
It is:
Does this complexity earn its keep?
There is a diligence trap here
This matters when buying businesses too.
An investor looking at a vertically integrated company will often find assets and capabilities that appear unusual.
The obvious question is:
Why does this business own that?
Ask it.
But ask another question immediately afterwards:
What else gets worse if we take it away?
That second question matters because the standalone economics can be misleading.
An asset may look subscale.
A capability may appear non-core.
An internal operation may benchmark poorly against an external provider.
Removing it may genuinely reduce cost and release capital.
But what happens next?
Does inventory rise?
Does another asset become less productive?
Does lead time increase?
Does the customer proposition weaken?
Does the business become dependent on capacity it no longer controls?
Does a competitor now have access to the same capability?
Does management lose information it previously received directly?
Does the business lose an option it may need in three years?
If the answer to those questions is “nothing”, the asset probably deserves to go.
But if several parts of the operating model become less productive, the standalone return was never the whole return.
That is the diligence trap.
The boundary of the business is a capital allocation decision
Every capability a business owns consumes something.
Capital.
Management attention.
Organisational capacity.
Risk.
Complexity.
It therefore has to earn its place.
But the return should be measured against the value it creates across the system, not simply the profit allocated to the activity itself.
That could mean:
higher utilisation,
lower working capital,
better margins,
greater customer retention,
pricing power,
lower risk,
faster growth,
greater responsiveness,
or strategic options the business would otherwise not possess.
Only then can we answer the real question:
Is control worth what it costs?
Some businesses own too much.
They confuse control with capability and strategic importance with historical attachment.
Others have gone too far in the opposite direction.
They have outsourced so much of the operating system that they have effectively become coordinators of suppliers while remaining fully accountable for an outcome they no longer fully control.
Neither model is inherently superior.
The right boundary is different for every business.
But it should be designed deliberately.
Not inherited.
Not determined by management fashion.
And not decided by looking at one function at a time.
Because the objective is not to optimise every part of the business.
It is to optimise the business.
And that leads to a much better question than whether something is “core”:
Does owning this make everything around it more valuable?
If the answer is no, challenge why you own it.
If the answer is yes, be very careful what you optimise away.