Analyze Workflow & Interpreting Results

Work through the Analysis process in Bear Decisions: filter scenarios, drill into instances, and read outputs without common misinterpretations.

This aligns with Analyze, challenge & evolve in how we approach client work—testing what drives outcomes, how strategies perform across scenarios, where variability comes from, and which assumptions break under pressure. When you are ready to package a recommendation for others, use Communication.

Chart mechanics live in Dashboard & Visualizations.

When Scenario Analysis Matters Most

Scenario analysis with Bear Decisions is most valuable when you're making important decisions with long-term consequences, facing significant uncertainty in key variables, and need to communicate risks to stakeholders while testing the robustness of your strategy.

Interactive Analysis Features

A key part of the learning process is to interactively explore the results of the analysis to understand the challenges and opportunities presented by these critical uncertainties, key decisions, their relationships - and then be able to make a data-informed decision or recommendation that is robust and defensible.

This can be done both within the app and by pushing alternative values back to your Excel model.

Filtering and Focus

Use the left-hand analysis tree to exclude specific cases from your calculations so you can test their impact on the overall outcome:

• Filter in and out problematic scenarios to test robustness or identify where additional mitigations are required
• Focus on specific uncertainty ranges to understand sensitivity
• Test how conclusions change when you include or exclude edge cases (when you present a simplified view externally, pair it with what was left out—see Communication)

Update Probabilities and Signals

You have already inputted the first set of probabilities and signals for your analysis, but you can update them to reflect new information or to test the impact of different assumptions. This helps you understand the value of either spending a bit more or waiting a bit longer to get greater confidence in a signal you are receiving. It also helps you bring other stakeholders along on your journey by testing out their hypotheses around what the likelihood of each case is - and rapidly seeing if this actually changes the overall conclusion.

Selected Instance Drawer

Dive deeper into a specific scenario by rapidly updating your Excel model with the specific instance/outcome directly from Bear Decisions.

To achieve this, click on any interesting point in your visualizations to open the Selected Instance Drawer with the context of the selected point:

• Case Selections: Understand which specific decisions and uncertainty cases contributed to the result
• Update Excel: Update your Excel model with a specific scenario's inputs
• Restore Original: Return to original initial value inputs or a previously saved configuration
• Clear: Remove all instances from the drawer and close it

Note that you may also collapse the drawer by clicking on the selected instance at the top of the drawer.

When dealing with strategies, you are choosing a specific sub-section of the strategy path to view the results of. Note that the Selected Instance Drawer will show both the correct and incorrectly identified signal contributions, so please be careful when choosing which instance to update.

Copy to Clipboard

Use the Copy to Clipboard button to copy the results of your analysis to the clipboard. This can be useful for sharing with others or for further analysis in other tools.

Export as PNG

Use the Export as PNG button to copy an image of the chart to the tab "BD_Images". This can be useful for sharing with others, for use in presentations or reports.

Staged Decision Process

When leveraging strategies, you need to consider both how the application will show the results and how your model will generate the results to ensure that the calculation and results are consistent.

In App Impacts

Specifically, the results shown in the application will typically be based upon blended outcomes of interpreting the signals correctly throughout the process when you are grouping by Strategy.

For example: if there is a Yes/No uncertainty, and you have an 80% signal, and the base likelihood of Yes is 60%, then the application will show a blend of 85.7% Yes contribution and 14.3% No contribution, for a signal perceived as Yes outcome. This blending is why the value will be different to what you would see if you navigate to the Yes outcome when grouped by Decision.

This is not the case for the Cumulative Distribution Plot, which will show the actual distribution of the base outcomes, or if you select P50 within the Waterfall Chart as the uncertainty input, which will select the 50th percentile outcome of the base outcomes.

Model Calculation Impacts

When deferring a decision from the 'Now' period, you need to ensure that the model is structuring that change appropriately to allow for divergence over time rather than a step change immediately. This could be that you are planning an expansion decision where the first capital is not spent until next year, but the decision is being made later within the current year. Alternatively it could be that you are spending some smaller amounts of capital now, but accept that if the decision to expand is not made, then this would have been wasted as sunk cost.

Similarly, with uncertainties that you are tracking and want to interpret as a signal for later decisions, you need to ensure that sufficient factors are enabled to be picked up early (or later).

Common Interpretation Mistakes

Confusing Expected Value with Certainty

Expected Value represents your best estimate, but actual results will likely differ. Don't treat it as a prediction—use it as a baseline for comparison and planning.

Ignoring the Full Distribution

Looking only at Expected Value or Mean misses critical insights. Use Range Charts, Tornado Charts, and CDP Charts to understand the full spectrum of possibilities and if there is a specific bias in the results. A decision with a slightly lower Expected Value might be preferable if it has much less downside risk.

Over-Relying on Single Metrics

Different transformations (NPV, IRR, Sum) and different variables (EBITDA, Cashflow, Net Profit, etc.) can lead to different conclusions. Use multiple perspectives to build confidence in your interpretation.

Forgetting Your Assumptions

Your analysis is only as good as your input assumptions and probability assignments. Regularly question whether your scenarios and likelihoods still make sense.

Process, not just a Result

Going through the analysis will identify where additional work may be needed to improve the plan or decision - mitigating key risks with new decision cases or even additional incremental decisions. Iterating on the results is important and expected.

Interrogating results: drivers, risk, and sensitivity

Use the app to stress-test the model the same way you would in a working session: what actually moves the outcome, where variability comes from, and whether conclusions hold when assumptions move. For how to translate findings into memos, decks, and stakeholder-ready messages, see Communication.

Identifying key drivers

Use different grouping options (see Analysis Configuration Options) to understand what matters most:

• Group by Decision: Which choices have the biggest impact on outcomes? Do you need to make all of these decisions now or can you defer some?
• Group by Strategy: If you are able to defer some decisions, how confident do you need to be in the signals you are receiving to make the decision? When do you need to make the later decision? How much option value are you creating by deferring some decisions?
• Group by Uncertainty: Which external factors create the most risk? What is the relative importance of each uncertainty area/case for mitigation?
• Filter scenarios: How robust are your conclusions when you add or remove extreme cases?

Risk assessment and mitigation

Look for patterns that suggest additional work is needed:

• High-Impact Uncertainties: Which external factors, if they go wrong, could derail your strategy?
• Adjust Probabilities: How likely does a specific uncertainty area/case need to be for it to be critical to the decision and directionally change the choice?
• Decision Sensitivity: Are some decisions only good under specific conditions?
• Edge Cases: What scenarios would require completely different strategies?

Strategic insights

Bear Decisions helps you identify strategic insights that pure intuition might miss:

• Illogical Combinations: Some decision combinations may not make strategic sense together
• Conditional Strategies: Certain decisions may only be optimal under specific uncertainty conditions
• Risk-Return Tradeoffs: Understand when higher expected returns come with disproportionate risk

Making confident decisions

Perfect information is impossible. Use the prompts below for your own judgment before you communicate outward; for the recommendation package itself, structure it around the questions in Communication. You often have enough to decide when:

• One decision clearly dominates across most reasonable scenarios
• You understand the key risks and have mitigation strategies
• Additional analysis wouldn't change your fundamental conclusion
• The cost of delay exceeds the value of additional information

Next Steps

Continue to Communication →