Don't try to fit an AI square peg into a round hole

Don’t try to fit an AI square peg into an organizational round hole

AI adoption isn’t enough …

The Operating Model Must Fit for AI ROI

Most organizations are investing in AI.

Fewer are asking the harder question:

Does our operating model still fit the work AI makes possible?

AI can improve individual productivity. It can help people write faster, analyze faster, summarize faster, and produce more work with less effort.

But more AI does not automatically create better enterprise performance.

In many organizations, AI creates more analysis, more recommendations, more pilots, more activity, more cost, and more noise—without solving the deeper issues that block results

The problem is usually not the AI tool.

The problem is the operating model around it.

The AI ROI Gap

The AI ROI gap is usually an operating model gap.

Many organizations are investing in AI before redesigning how work, decisions, authority, accountability, and governance fit together.

That creates an AI ROI gap.

AI cannot automatically fix:

Technology investments without clear business value
Misalignment around mission, values, vision, and priorities
Fragmented workflows and bottlenecks
Unclear ownership and conflicting incentives
Slow or overloaded decision-making
Leaders not equipped to guide human–AI work
Governance gaps that leave authority and accountability unclear
Productivity gains that do not become enterprise-level performance

Adding AI to these problems can make them worse.

A faster broken workflow is still broken.
A faster unclear decision is still unclear.
A faster misaligned organization is still misaligned.

The question is not simply: How can we use more AI?

The better question is: Where has AI materially changed the constraint – the bottleneck to better results – and what should we do about it?

Not Every AI Opportunity Requires Redesign

AI does not require every organization to rebuild itself.

Some AI opportunities are narrow and useful. Some should be handled through individual augmentation, task automation, or ordinary process improvement.

Some should be ignored.

Some should be stopped.

The first step is disciplined diagnosis.

For each opportunity, leaders should ask:

Does AI materially change how this outcome can be produced?
Does it affect a real performance constraint?
Is the workflow strategically important?
Can outputs be verified?
Are errors detectable and reversible?
Who owns the decision?
Who remains accountable?
What new risks are created?
What value would justify the transition cost?
What evidence would cause us to scale, restrict, revise, or stop?

The right answer may be:

Defer
Augment
Automate
Improve
Reconstruct the workflow
Redesign the operating model
Restrict or stop the AI use

A thinking organization does not maximize AI use.

It maximizes sustainable performance under acceptable risk, clear authority, and accountable responsibility.

What We Do

We help CEOs, Boards, investors, and senior leadership teams evaluate and, if needed, reengineer the operating model needed to turn AI investment into ROI.

Our AI work focuses on how AI and an operating model fit together.

The goal is not another AI roadmap, pilot, or training program.

The goal is a stronger operating model that translates into actual results.

That may mean using more AI. It may also mean reengineering the operating model to capture AI value.

The Built to Think Framework

Our Built to Think framework helps leaders redesign the organization as a whole system—not just add AI to isolated tasks.

It incorporates the Optimal Business Success Model™, or OBSM:

Optimal Business Success Model

OBSM recognizes that sustainable performance depends on the alignment of four interdependent factors:

Strategy
Culture
People
Systems and processes

AI can affect every one of these.

But that is exactly why isolated AI adoption is dangerous.

If AI is added to one workflow, one department, one leadership initiative, or one technology stack without redesigning the broader operating model, the organization often suboptimizes. One area may become faster while another becomes overloaded. One team may gain productivity while another inherits more review work, risk, confusion, or rework. One function may automate activity while the enterprise fails to improve results. One AI pilot may look impressive while the operating model remains misaligned.

The problem is not that AI was used.

The problem is that AI was added to a system still designed around yesterday’s assumptions about work, information, authority, accountability, and leadership.

Built to Think helps leaders use AI as part of a coordinated organizational redesign. It helps answer some essential questions as a foundational starting point:

What outcomes should the organization be designed to produce?
Where has AI materially changed the operating constraint?
Which workflows should be improved, eliminated, or rebuilt?
Which decisions should stay human, be AI-assisted, or be delegated within clear limits?
How should authority, accountability, governance, and assurance be designed?
What leadership, cultural, and workforce changes are needed to make the new model work?
What evidence would justify scaling, revising, restricting, or stopping the initiative?

These questions move AI from isolated experimentation to enterprise redesign.

The goal is not to make one part of the organization look more advanced.

The goal is to make the whole organization more coherent, adaptive, and capable of producing sustainable results.

Services to Help You Bridge the AI ROI Gap

We have many ways to help you; below are some of the most common. What services are pertinent will depend on your objectives and situation:

AI Operating Model Fit Assessment

Before scaling AI, determine whether your operating model fits the work AI makes possible.

The AI Operating Model Fit Assessment helps leaders identify:

Where AI may create real enterprise value
Where AI is likely to create noise, cost, or risk
Which workflows are structurally ready for AI
Which decisions require clearer authority
Where governance and assurance are insufficient
Whether the current constraint is technology, workflow, decision speed, trust, capability, data, capital, regulation, or leadership
Whether the right response is augmentation, automation, improvement, workflow reconstruction, operating-model redesign, or refusal

The output is a clearer view of where to act, where to wait, and where not to force AI into the wrong system.

Redesign Contingency Analysis

Not every AI initiative deserves transformation.

We help leadership teams evaluate AI opportunities through disciplined criteria that fit their organization’s uniqueness and situation. Example criteria:

Strategic importance
Effect on the binding constraint
Economic value
Decision consequence
Reversibility
Error detectability
Verification burden
Data readiness
Integration complexity
Workforce effect
Regulatory exposure
Implementation capacity

This prevents two common failures on either end of a spectrum:

Treating every AI opportunity as a transformation mandate.
Treating AI as merely another tool when it has changed the underlying operating logic.

The purpose is to choose the right level of response.

Human–AI Workflow Redesign

Many organizations automate existing processes that should have been rebuilt first.

We help redesign critical workflows around outcomes, not inherited handoffs.

This includes determining:

Which work should remain human
Which work should be AI-assisted
Which work can be automated
Which work should be eliminated
Which decisions can be delegated within limits
Where human review is meaningful
Where evidence and controls are required
How exceptions should be handled
How value will be measured

The purpose is not to automate yesterday’s process.

The purpose is to redesign the outcome-producing system for the world AI makes possible.

Decision Architecture and Delegation Design

AI changes who or what influences decisions.

That requires explicit design.

We can help organizations define:

Which decisions matter most
Who owns each decision
What evidence is required
Where AI may inform, analyze, recommend, draft, or execute
When escalation is required
Who can override
Who remains accountable
How decisions will be reviewed
How the organization will learn from outcomes

Capability is not authority.

A system may be able to make a recommendation. That does not mean it should be allowed to act without clear boundaries.

Leadership and Workforce Alignment

AI transformation fails when leaders treat it as a technology program while people experience it as a change to work, power, value, identity, and trust.

We help leadership teams address:

Human–AI collaboration norms
Role and workforce redesign
Leadership behavior changes
Decision authority
Trust and communication
Capability development
Resistance and adoption barriers
Use of released capacity

The human side is not separate from the operating model. It is part of the operating model.

Fractional Transformation Leadership

Some organizations need more than advice. They need experienced leadership to help coordinate the redesign.

We provide fractional transformation support for leaders who need help moving from AI activity to operating-model fit.

This may include:

Transformation portfolio design
Executive alignment
Board or investor reporting
Workflow and decision redesign
Governance implementation
Internal capability building
Benefits tracking
Leadership coaching and advisory support

The role is not to create consultant dependency.

The role is to help the organization build the capability to think, decide, act, and learn more effectively.

Common Problems We Solve

AI Activity Without Enterprise Value

You have tools, pilots, dashboards, committees, and training—but limited evidence of enterprise-level improvement.

We help determine whether the issue is the use case, the workflow, the constraint, the operating model, or the measurement system.

Fragmented Workflows

AI is being added inside departments, but the end-to-end workflow remains fragmented.

We help reconstruct the complete outcome-producing system.

Decision Bottlenecks

AI increases analysis and recommendation volume, but leaders still make too many decisions too slowly.

We help redesign decision rights, escalation, review, and accountability.

Hidden Delegation

AI recommendations are becoming the practical decision while humans provide only nominal approval. This is, in effect, delegation by default to AI.

We help define meaningful human oversight, evidence requirements, override authority, and accountability.

Governance Without Control

The organization has AI principles or policies but lacks operational controls, ownership, monitoring, and stop authority.

We help design governance that works in the actual operating system.

Productivity Without Sustainable Performance

AI improves local productivity but creates new review burdens, quality issues, risk, distrust, or capability loss.

We help evaluate whether productivity gains are becoming enterprise value.

Workforce and Leadership Misalignment

AI changes work, status, identity, capability needs, and power.

We help leaders address the human and political realities of redesign—not just the communication plan.

How We Approach Assisting You

We Do Not Start With “Use More AI”

More AI is not always better.

The right answer may be more AI, less AI, different AI, better governance, workflow reconstruction, leadership redesign, or no AI at all.

We Start With the Constraint

The central question is:

What currently limits performance—and has AI changed that constraint?

If AI does not affect the real constraint, scaling AI may produce activity without advantage.

We Treat the Organization as a System

AI affects more than tasks.

It affects workflows, decisions, authority, accountability, culture, leadership, governance, incentives, and learning.

Improving one part while ignoring the system can suboptimize results.

We Connect Intelligence to Action

AI can create more “intelligence.”

Organizations still need to convert intelligence into choices, authority, coordinated action, feedback, and learning.

We Include the Right to Stop

A serious transformation process must include the possibility that the answer is:

Not yet
Not here
Not this way
Not at this level of autonomy
Not worth scaling
Stop

The organization that cannot stop cannot think well nor improve results well.

The Competitive Divide

The competitive divide will not be between organizations that use AI and those that do not.

It will be between organizations that force AI into old operating models and organizations that know what to preserve, what to redesign, and what to stop.

Where to Start?

AI often changes operating model economics and effectiveness.

If your organization is investing in AI but not yet seeing clear enterprise value, start with evaluating fit.

Useful questions include:

Where are you investing in AI now?
What enterprise value do you expect?
Where are you seeing results?
Where are you seeing noise, confusion, or stalled progress?
Which workflows or decisions are most affected?
Who owns AI-enabled outcomes?
How are you measuring value?
Where do leaders disagree?
What would cause you to scale, revise, restrict, or stop?

If there appears to be a fit problem and you desire assistance, schedule a conversation with us to discuss your AI ROI gap and how we might help you create better results.

FAQs

Does every organization need AI operating-model redesign?

No.

Some AI opportunities require only individual augmentation, task automation, or ordinary process improvement.

Redesign is justified when AI materially changes how an important outcome can be produced and the inherited operating model prevents the organization from capturing that value safely and coherently.

How is this different from AI strategy consulting?

Most AI strategy work starts with use cases, tools, or roadmaps and assumes AI will provide value.

This work starts with operating-model fit.

We examine the relationship among outcomes, constraints, workflows, decisions, authority, accountability, governance, leadership, and learning.

The question is not only “What AI should we use?”

The question is “How must the organization work so AI will produce sustainable enterprise value?”

How is this different from AI training?

AI training may help individuals use AI tools.

That is useful but insufficient when the real issue is workflow design, decision authority, accountability, governance, incentives, or leadership alignment.

AI training improves user capability.

Operating-model fit improves enterprise performance.

What if we already have AI pilots underway?

That is common and can be a sensible way to begin AI experimentation.

We can help you determine whether:

Current pilots are addressing a real constraint
The workflow is ready
Governance is sufficient
Evidence justifies scale, revision, restriction, or termination

What if our AI problem is actually a leadership or culture problem?

It often is partly a leadership or culture problem. This is not unique to AI but common with any major change or new initiative.

Leadership and culture should not be treated as separate problems from processes and tools. They interact.

The work is to understand the system and redesign the parts that must change together.

Do you implement AI tools?

Not usually. We focus on operating-model fit and Optimal Business Success Model™factors.

When technical implementation is required, we can coordinate with internal teams or external technical partners so the operating model and technical build remain aligned.

Can this apply outside to other than large enterprises?

Yes.

The principles apply wherever leaders are responsible for a consequential system.

That may be an enterprise, portfolio company, business unit, function, product line, regulated process, customer journey, or decision workflow.

What is the best starting point or pilot?

Start with one consequential outcome system.

It should be important enough to matter, bounded enough to govern, measurable enough to evaluate, and representative enough to teach the organization something valuable.

Consider the questions in “Where to Start” above, then contact us if you desire assistance.