Why nobody calculates global optima today

- and why this is no coincidence

Executive Summary

In almost all organizations - from corporations to financial institutions to the public sector - strategic decisions are made according to a similar pattern: analyze, discuss, prioritize, decide.

The instruments used differ in detail, not in principle. Scenarios are compared, key figures aggregated, risks simulated, assumptions discussed. Decisions are based on consensus, experience and plausibility.

What is striking is that global, ex-ante calculated optima hardly play a role in practice.

Not because they are impossible. Not because there is no need for them. But because existing organizational, consulting and governance models are not structurally designed to treat decisions as mathematically solvable optimization problems.

This article explains

• why this is the case,
• where the limits of classical decision logic lie
• and why the calculation of global optima is not an evolution of existing methods, but a new category.

1. How decisions are actually made today

Regardless of industry or size, complex decisions usually follow the same process:

1. Projects or measures are described individually
2. Profitability is estimated or modeled
3. Risks are simulated or qualitatively assessed
4. Scenarios are compared
5. Committees discuss and prioritize
6. A selection is made

This logic is familiar, accepted and institutionally anchored. It is compatible with existing roles:

• Management decides
• Controlling measures
• Consulting structures
• Committees legitimize

The crucial point: this logic evaluates decisions - it does not calculate them.

2. Evaluation is not calculation

In practice, decisions are almost always evaluated ex post or locally:

• "Which project has the higher ROI?"
• "Which scenario is more robust?"
• "Which measure seems plausible?"

These questions are legitimate - but they are not equivalent to the question:

Which combination of all possible measures is globally optimal under given restrictions?

The difference is fundamental:

• Evaluation compares known alternatives
• Calculation searches the decision space itself

As soon as more than a few projects, goals or restrictions are involved, this decision space explodes exponentially. At this point, discussion, intuition and isolated key figures are mathematically overwhelmed - regardless of experience or intelligence.

3. Why global optima are rarely calculated

The fact that global optima are rarely calculated is no coincidence, but the result of several structural limitations.

3.1 Organizational logic

Organizations are designed to distribute responsibility, not to bundle it mathematically.

A calculated optimal solution generates

• Unambiguity
• Reproducibility
• Coercion

This contradicts the logic of many decision-making processes, which deliberately leave room for interpretation, consideration and political legitimization.

3.2 Consulting logic

Classic advisory systems provide

• Decision frameworks
• Logics of comparison
• Narratives
• Prioritization aids

But not

• mathematically compelling solutions
• ex-ante verifiable optima

The reason is not a lack of competence, but liability and business model logic. A global optimum is not a proposal - it is an assertion with a burden of proof.

3.3 Governance logic

The following applies in governance structures:

• Decisions must be explainable
• Responsibility must remain assignable

A calculated optimum shifts this logic:

• from opinion to deduction
• from experience to model
• from discussion to proof

Many systems are neither culturally nor legally prepared for this.

Nevertheless, the global optimum

4. The silent assumption behind "Nobody does it that way"

The statement

"Nobody does it that way"

sounds like an argument at first. In reality, it is an observation without normative force.

It implicitly assumes:

If something were relevant, it would already be done.

This assumption is not historically tenable. Almost all structural innovations in decision-making, financial and control systems were not made initially because they did not fit into existing models.

It is not innovation that is missing - it is connectivity.

5. Why decision optimization is not a consulting discipline

Global decision optimization is fundamentally different from consulting:

Consulting	Decision optimization
structures discussion	eliminates discussion
provides plausibility	provides dominance
works with scenarios	works with complete spaces
accepts ambiguity	enforces unambiguity
is liable for process	adheres to model

These differences are not gradual, but categorical.

This is why real optimization does not arise within existing consulting models, but outside them - as a separate infrastructure.

6. Ex-ante instead of ex-post: the change of perspective

The central break lies in the timing:

• Traditional systems evaluate decisions after they have been made
• Optimization calculates decisions before they are made

Ex-ante optimization means

• Making opportunity costs visible before they arise
• Resolving conflicting goals mathematically instead of politically
• Reproducibly justifying decisions

This not only changes results - it changes responsibility, governance and decision-making quality.

7. Why this is inconvenient

Global optima are inconvenient because they:

• Shorten discussions
• Create power shifts
• make implicit assumptions explicit
• Make decisions verifiable

They reduce the room for:

• Gut feeling
• political compromises
• retrospective justifications

Not out of malice - but out of mathematical consistency.

8. A new category: decision-making infrastructure

What emerges here is neither:

• Consulting
• Software in the classic sense
• Controlling
• BI

But decision-making infrastructure.

A level that:

• Formalizes decision spaces
• Mathematically maps conflicting objectives
• Integrates restrictions
• and calculates the global optimum

Not as a recommendation - but as a property of the system.

9. Why the topic is now relevant

Several developments are acting simultaneously:

• increasing complexity
• limited budgets
• increasing liability
• growing governance pressure
• decreasing tolerance for wrong decisions

These factors make it increasingly risky not to calculate decisions.

Not because calculation is perfect - but because non-calculation systematically generates opportunity costs.

10. Conclusion

The fact that hardly anyone calculates global optima today is not proof against the idea. It is proof of the limitations of existing systems.

Decision optimization is:

• not a consulting product
• not a tool
• not a trend

But a necessary response to exponential decision spaces.

Not everyone will need it. But everyone who needs it will not be able to replace it.

FAQ - Frequently asked questions from decision-makers

Isn't global optimization theoretical, but unrealistic in practice?

No. It is unrealistic to manage complex portfolios permanently without formal optimization.

Does this mean the end of management experience?

No. Experience remains relevant - but no longer as a substitute for decision-making, but as model input.

Does optimization replace human responsibility?

On the contrary. It makes responsibility more explicit because decisions can be justified.

Why don't all organizations use this yet?

Because structures, processes and thought patterns need time to adapt to new categories.

Isn't that too rigid for the complexity of the real world?

Rigid are heuristic decisions in complex spaces. Optimization explicitly maps complexity.

Doesn't that lead to acceptance problems?

In the short term, yes. In the long term, acceptance increases through comprehensibility.

Is this scalable?

Yes. This is precisely why algorithmic and hybrid optimization approaches exist.

Is this an attack on consulting?

No. It is a supplement on a different level.

When does decision optimization become unavoidable?

As soon as decision spaces become larger than what humans can plausibly oversee.