From messy problems to decisions you can stand behind
Complex decisions rarely start clean.
Uncertainty, interactions and competing objectives make it difficult to see what really matters.
I help teams cut through that complexity - so they can make clear, defensible recommendations leadership can stand behind.
Where decision models break down
Most decision models fail in predictable ways:
- Key uncertainties are left as implicit assumptions or ignored
- Variables are treated as independent when they really move together
- Models are built as one-off exercises, while decisions happen in stages
- Outputs focus on numbers, not what actually drives outcomes
- And the process itself becomes slow, manual, and difficult to repeat
The result is analysis that looks rigorous, but requires significant effort to maintain, relies on disconnected steps, may be slow to generate answers, and is hard to trust under pressure.
I work with teams to fix this by bringing structure, simulation, process design, and clear thinking to complex decisions:
- Identify and focus on the variables that actually drive outcomes
- Capture how those variables interact (not just in isolation)
- Quantify uncertainty in a way that reflects how the real system behaves
- Build models and processes that can evolve as decisions unfold
A structured, iterative approach to complex decisions
Complex decisions aren't solved in a single pass.
As you build a model, you uncover new variables, challenge assumptions, and refine what the decision actually is. The process is inherently iterative.
I use a structured, simulation-based approach that treats decision-making as a cycle - not a one-off exercise.
Clarify the Decision
Prevent solving the wrong problem
- •What decisions are actually being made?
- •What does success look like? (a choice, an objective to optimise, constraints to be satisfied?)
- •Who is making the decisions, and when do they occur?
Most problems start loosely defined. Tightening the decision space early ensures the model is solving the right problem — not just a convenient one.
Frame the System
Define the world the decision sits within
- •What are the key performance measures and the boundaries of the system?
- •What uncertainties exist – and how do they behave (noise, trends, shock, rare events)?
- •How do variables interact, scale, and constrain each other?
- •What ranges and edge cases actually matter for the decision?
This step moves beyond metrics into understanding the structure and behaviour of the system.
A model is only as good as the way the problem is framed.
Build the Simulation Model
Turns the system into something that you can explore
- •Represent the key variables, decisions, and their relationships
- •Model how the system evolves over time, including staged decisions
- •Incorporate new information as it becomes available
- •Design for ranges, distributions, and interactions – not single-point estimates
The goal is not just to calculate outcomes, but to simulate how the system behaves under different conditions.
Analyze, Challenge & Evolve
Reveals what matters – and what's missing
- •What actually drives outcomes?
- •How do different strategies perform across scenarios?
- •Where does risk and variability come from?
- •What assumptions break under pressure?
- •What new variables, interactions, or decisions emerge?
This is where the model is tested, challenged and refined. And where insights are generated.
Iterations here often lead to: reframing the decision, introducing new variables, and changing how the system is represented.
The process is inherently iterative – each pass improves both the model and the decision itself.
Communicate the Decision
Turns analysis into action
- •What is the recommended path forward?
- •What are the key drivers behind that recommendation?
- •What risks and trade-offs need to be understood?
- •What range of outcomes should the decision-makers expect?
Analysis only matters if it can be understood and acted on.
The goal is a decision that is: clear, defensible, and robust under scrutiny.
Data, Processes & Systems
Enables the model to work in practice
- •How is data sourced, validated, and kept up to date?
- •What level of trust and transparency exists in the inputs?
- •Who owns the data, the model, and the outputs?
- •How are updates, reviews, and decisions operationalised?
- •How can workflows be streamlined to reduce manual effort and disconnected steps?
Strong decision models rely on more than structure — they require reliable inputs, clear ownership, and repeatable processes.
The objective is not just a better model, but a system that can be trusted, maintained, and used consistently over time.
Clarify the Decision
Prevent solving the wrong problem
- •What decisions are actually being made?
- •What does success look like? (a choice, an objective to optimise, constraints to be satisfied?)
- •Who is making the decisions, and when do they occur?
Most problems start loosely defined. Tightening the decision space early ensures the model is solving the right problem — not just a convenient one.
Analyze, Challenge & Evolve
Reveals what matters – and what's missing
- •What actually drives outcomes?
- •How do different strategies perform across scenarios?
- •Where does risk and variability come from?
- •What assumptions break under pressure?
- •What new variables, interactions, or decisions emerge?
This is where the model is tested, challenged and refined. And where insights are generated.
Iterations here often lead to: reframing the decision, introducing new variables, and changing how the system is represented.
The process is inherently iterative – each pass improves both the model and the decision itself.
Build the Simulation Model
Turns the system into something that you can explore
- •Represent the key variables, decisions, and their relationships
- •Model how the system evolves over time, including staged decisions
- •Incorporate new information as it becomes available
- •Design for ranges, distributions, and interactions – not single-point estimates
The goal is not just to calculate outcomes, but to simulate how the system behaves under different conditions.
Frame the System
Define the world the decision sits within
- •What are the key performance measures and the boundaries of the system?
- •What uncertainties exist – and how do they behave (noise, trends, shock, rare events)?
- •How do variables interact, scale, and constrain each other?
- •What ranges and edge cases actually matter for the decision?
This step moves beyond metrics into understanding the structure and behaviour of the system.
A model is only as good as the way the problem is framed.
Data, Processes & Systems
Enables the model to work in practice
- •How is data sourced, validated, and kept up to date?
- •What level of trust and transparency exists in the inputs?
- •Who owns the data, the model, and the outputs?
- •How are updates, reviews, and decisions operationalised?
- •How can workflows be streamlined to reduce manual effort and disconnected steps?
Strong decision models rely on more than structure — they require reliable inputs, clear ownership, and repeatable processes.
The objective is not just a better model, but a system that can be trusted, maintained, and used consistently over time.
Communicate the Decision
Turns analysis into action
- •What is the recommended path forward?
- •What are the key drivers behind that recommendation?
- •What risks and trade-offs need to be understood?
- •What range of outcomes should the decision-makers expect?
Analysis only matters if it can be understood and acted on.
The goal is a decision that is: clear, defensible, and robust under scrutiny.
Where this makes a difference
This work is most valuable when decisions are important — but the current approach isn't holding up.
For example:
- •Budgets and forecasts that require significant manual effort to update
- •Long-term plans built on assumptions that are hard to test or explain
- •Strategic decisions with multiple pathways, but no clear way to compare them
- •Ad-hoc analysis that takes too long to build and can't be reused
- •Models that produce numbers, but not insight into what's driving them
- •Outputs that are difficult to explain or defend to stakeholders
In these situations, the challenge isn't just uncertainty — it's a lack of structure, clarity, and repeatability.
This approach is commonly applied to:
This approach helps turn one-off analysis into a structured, repeatable way of making decisions.
Supporting tools and workflows
This work can be delivered using your existing tools and models, or supported by Bear Decisions— a platform designed to bring structured scenario and simulation analysis directly into Excel.
For many teams, improving the structure, thinking, and processes are enough.
Where appropriate, Bear Decisions can be introduce to:
- Accelerate model development and iteration
- Explore scenarios adn uncertainties more effectively
- Reduce manual effort and disconnected workflows
My focus is always on the decision - not the tool.
Engagement Options
Advisory
Support on an existing model or decision
Model Design
Build or restructure a decision model
End-to-End Engagement
From problem framing through to recommendation
Training
Help teams adopt better decision modelling practices