Anthony and Simon Framework Explained for Senior Leaders and Decision Making
Most leadership teams don’t fail because they lack data. They fail because they use the wrong kind of decision process for the decision in front of them.
A pricing bug, a roadmap tradeoff, and a market-entry bet all need judgment. But they don’t need the same level of control, analysis, or executive attention. That’s where the Anthony and Simon framework helps.
Robert N. Anthony gave us a practical way to separate strategic planning, management control, and operational control. Herbert A. Simon, a Nobel Prize winner known for his work on decision-making, helped explain how decisions range from programmed and structured to messy, uncertain, and non-programmed.
Put together, the framework shows how senior leaders actually make decisions: by matching the decision type to the right level of management control.
The Anthony and Simon framework in plain English
The Anthony and Simon framework combines two ideas.
Anthony focused on levels of control:
```text
Strategic planning
Big choices about direction, markets, business model, and long-term advantage
Management control
Mid-level choices about resources, priorities, budgets, and performance
Operational control
Day-to-day choices about tasks, workflows, quality, and execution
```
Simon focused on types of decisions:
```text
Structured decisions
Repeated, rule-based, and measurable
Semi-structured decisions
Partly measurable, but still need judgment
Unstructured decisions
Ambiguous, rare, high-stakes, and hard to reduce to rules
```
When you layer them together, you get a simple but powerful map.
Anthony level | Simon decision type | Typical decision | Who usually owns it |
Operational control | Structured | Should this payment be approved? | Team leads, systems, frontline managers |
Management control | Semi-structured | Which product bets get more engineering capacity? | Functional leaders, GMs, directors |
Strategic planning | Unstructured | Should we enter a new market or change the business model? | CEO, founders, board, senior executives |
This sounds academic at first. It isn’t.
A startup deciding whether to retry a failed card payment is dealing with operational control. A SaaS VP deciding whether to shift engineers from onboarding to retention is dealing with management control. A founder deciding whether to move from a product-led motion to enterprise sales is dealing with strategic planning.
Each one needs a different kind of thinking.
The three levels of management control
Anthony’s original model is helpful because it stops leaders from treating every problem like a strategy problem.
That matters because senior teams often get pulled down into operational noise. A metric dips, a customer complains, a release slips, and everyone wants to jump in. Sometimes that’s needed. Most of the time, the better question is: what level of control is this?
Operational control is about reliable execution
Operational control deals with repeated work. The goal is consistency.
Think of:
Fraud checks in a fintech app
Server alerts in a cloud platform
Customer support routing in a SaaS company
Refund approvals in an e-commerce marketplace
QA checks before a mobile release
These decisions should be as structured as possible. If the team has to invent the answer every time, the system is too fragile.
A good operational decision has clear rules:
```text
Signal comes in
Rule or threshold applies
Action happens
Outcome gets logged
Exception gets escalated
```
For example, a payments startup might create a rule that flags unusually large transactions from new accounts. Most cases go through an automated risk score. A small number go to manual review. Only rare edge cases reach senior risk leaders.
That’s healthy. Executives shouldn’t personally review routine fraud flags. They should design the control system, define the risk appetite, and check whether the system works.
Management control is about resource tradeoffs
Management control sits in the middle. This is where leaders turn strategy into operating choices.
It includes questions like:
Which teams get more headcount?
Which customer segment gets priority?
Which product metric matters most this quarter?
Should sales push annual contracts or faster self-serve growth?
Should engineering reduce tech debt or ship a visible feature?
These decisions are usually semi-structured. There’s data, but the data doesn’t make the choice by itself.
A product leader might know that activation is down, enterprise churn is flat, and support tickets are rising. The numbers matter. Still, the leader has to judge whether the real issue is onboarding, product complexity, poor-fit customers, or weak documentation.
That’s management control: turning imperfect information into useful direction.
Strategic planning is about direction under uncertainty
Strategic planning is where decisions become rare, expensive, and hard to reverse.
Examples include:
Moving upmarket into enterprise sales
Expanding from one country to several
Buying a company instead of building the product
Changing pricing from per-seat to usage-based
Shutting down a mature product line to fund a new platform
These are unstructured decisions. You can’t solve them with a dashboard alone.
A leadership team can model market size, analyze competitors, interview customers, and review financial scenarios. That helps. But there’s still uncertainty. Strategic decisions involve belief about the future.
That’s why senior leaders need a different decision process at this level. They need debate, assumptions, options, and clear ownership.
How Simon’s decision types change the leadership process
Simon’s work is useful because it explains why some decisions can be delegated safely while others require senior judgment.
Here’s the practical version.
```text
More structure
|
| Operational decisions
| Rules, thresholds, automation, checklists
|
| Tactical decisions
| Metrics, tradeoffs, judgment, resource allocation
|
| Strategic decisions
| Ambiguity, scenarios, values, long-term bets
|
Less structure
```
The less structured the decision, the more leaders need to slow down and ask better questions.
Structured decisions need rules
Structured decisions are repeated and predictable.
A marketplace might decide when to release seller payouts. A cybersecurity product might decide which events count as low-risk alerts. A subscription app might decide when to send a failed payment reminder.
The leadership job is to make the rule clear.
Good structured decisions have:
Clear inputs
Defined thresholds
Standard responses
Audit trails
Exception paths
If a structured decision keeps escalating to executives, something’s wrong. The process lacks trust, the rule is unclear, or the risk tolerance hasn’t been defined.
Semi-structured decisions need judgment and numbers
Semi-structured decisions are where many senior operators spend most of their time.
A startup might ask whether to keep funding a new AI assistant feature. Early usage is strong, but retention is unclear. Sales likes the demo value, but support sees confusion. Engineering says the model quality needs more work.
That decision can’t be reduced to one metric. It needs a decision frame.
A useful frame might include:
Question | Why it matters |
What outcome are we trying to improve? | Keeps the team from chasing activity |
What evidence do we have? | Separates facts from opinions |
What assumptions are we making? | Shows what could break the plan |
What tradeoff are we accepting? | Makes the cost visible |
When will we revisit the choice? | Stops temporary decisions from becoming permanent |
Semi-structured decisions should end with a clear call, not endless analysis.
Unstructured decisions need options
Unstructured decisions are the ones leaders remember because they carry real weight.
A founder might ask: should we sell to mid-market companies or rebuild the company around enterprise buyers?
There may be no clean answer. Mid-market growth is faster. Enterprise contracts are larger. The product needs roadmap changes either way. The sales team may need a different profile. Customer success may need deeper technical skills.
For this kind of decision, leaders should avoid asking, “What’s the right answer?” too early.
A better process is:
```text
Define the decision
List the real options
Name the assumptions
Test what can be tested
Choose the path
Set review points
```
The output shouldn’t be certainty. It should be a well-reasoned commitment.
A practical decision map for tech and startup leaders
Here’s a simple way to apply the framework.
```text
Decision structure
Structured Semi-structured Unstructured
Operational Automate Escalate rules Redesign process
Tactical Dashboard Leadership review Cross-functional choice
Strategic Policy Scenario model Executive commitment
```
This map helps leaders spot mismatches.
For example, if an unstructured strategic decision is being treated like a dashboard review, the team may overtrust the numbers. If a structured operational decision keeps landing in senior meetings, leaders are wasting attention on work that should be codified.
Example one is the product roadmap
A SaaS company has three competing priorities:
Improve onboarding
Add enterprise permissions
Build AI-generated reporting
At the operational level, teams can track bug counts, cycle time, and release quality.
At the management control level, product and engineering leaders decide how to allocate capacity. They compare customer demand, revenue impact, technical effort, and strategic fit.
At the strategic level, the CEO and executive team ask a bigger question: are we building for self-serve teams or enterprise buyers?
That top-level answer changes the roadmap. Enterprise permissions may matter more if the company is moving upmarket. Onboarding may matter more if self-serve growth is still the biggest bet.
Same roadmap, three decision levels.
Example two is customer churn
A subscription software company sees churn rise.
Operational control asks:
```text
Are cancellation flows working?
Are support tickets being answered on time?
Are payment failures being recovered?
```
Management control asks:
```text
Which customer segments are churning?
Which team owns retention work?
Should success managers focus on onboarding or renewals?
```
Strategic planning asks:
Are we selling to the wrong customers?
Is the product too broad?
Does our pricing attract poor-fit accounts?
```
A weak leadership team argues about the churn number. A strong one separates the control levels and assigns the right work to each.
Example three is AI adoption
A company wants to add AI features to its product.
Operational decisions include model monitoring, response quality checks, privacy reviews, and fallback rules.
Management control decisions include staffing, vendor selection, feature priority, and customer rollout timing.
Strategic decisions include whether AI is a feature, a workflow layer, or the foundation of a new product category.
That last choice is not a Jira ticket. It’s a strategic decision with business model implications.
How senior leaders should use the framework in real meetings
The framework becomes useful when it changes the way decisions get made.
Try this simple meeting filter.
Ask what kind of decision this is
Before debating the answer, name the level.
```text
Is this operational?
Is this tactical?
Is this strategic?
```
That one move saves time. It also reduces the chance that a senior team spends 40 minutes on a workflow issue and 4 minutes on the real business risk.
Match the forum to the decision
Not every decision belongs in the executive meeting.
Decision level | Best forum | Output |
Operational control | Team review, incident review, workflow owner | Rule, fix, checklist, escalation path |
Management control | Functional leadership review | Priority, resource allocation, metric target |
Strategic planning | Executive session, board discussion | Direction, commitment, investment thesis |
Senior leaders should protect their agenda. If everything reaches the top, the organization hasn’t built control. It has built dependency.
Define the decision owner
A decision without an owner becomes a discussion.
For structured decisions, the owner maintains the rule or process. For semi-structured decisions, the owner weighs inputs and makes the call. For unstructured decisions, the owner usually needs senior authority because the tradeoffs cut across functions.
Ownership doesn’t mean ignoring input. It means someone is accountable for closing the loop.
Write down the assumptions
This is especially important for strategic decisions.
A decision record can be short:
Decision
We will shift new sales efforts toward enterprise accounts.
Reason
Enterprise accounts show stronger expansion potential and better retention.
Key assumptions
The product can meet enterprise security needs within two quarters.
Sales hiring can shift without slowing current pipeline too much.
Customer success can support fewer, larger accounts.
Review date
End of Q2.
```
This is simple, but it prevents revisionist history. Later, the team can ask whether the decision was bad or whether an assumption changed.
Download-ready summary notes
Use this as a one-page reference for leadership planning, operating reviews, or decision audits.
Anthony and Simon Framework
Anthony’s management control levels
Operational control
Purpose
Run repeatable work reliably.
Decision type
Mostly structured.
Best tools
Rules, checklists, thresholds, automation, exception handling.
Example
Auto-approve standard refunds under a defined limit.
Management control
Purpose
Allocate resources and improve performance.
Decision type
Mostly semi-structured.
Best tools
Metrics, tradeoff discussions, prioritization, leadership reviews.
Example
Shift engineering capacity from new features to retention work.
Strategic planning
Purpose
Set direction under uncertainty.
Decision type
Mostly unstructured.
Best tools
Scenarios, assumptions, executive debate, board-level review.
Example
Move from self-serve SaaS to enterprise sales.
Simon’s decision types
Structured
Repeated, measurable, rule-based.
Semi-structured
Partly measurable, requires judgment.
Unstructured
Ambiguous, rare, high impact.
Leadership rule of thumb
Automate structured operational decisions.
Review semi-structured tactical decisions with data and judgment.
Treat unstructured strategic decisions as commitments under uncertainty.
```
The real value is knowing what not to escalate
The Anthony and Simon framework isn’t just a school model. It’s a filter for executive attention.
Operational decisions should become reliable systems. Tactical decisions should become clear trade-offs. Strategic decisions should become explicit commitments with named assumptions.
That’s the real leadership move: don’t treat every issue as equal.
When a decision shows up, ask three quick questions.
What level of control is this?
How structured is the decision?
Who is the right owner?
If the answer is clear, the organization moves faster. If the answer is unclear, the framework tells you where the confusion lives. That alone can save a leadership team from chasing noise while the real strategic calls wait in the background.
Comments