1. Integrating insights in the decision process
Welcome to this final lesson!
2. Decision-making under uncertainty
Throughout this course, we've explored various ways to handle uncertainty: probability models, simulations, decision trees, and more. But in business, decisions don't come with labels.
So, how do you bring everything together and apply what you've learned in real-world decision-making? That's what we'll focus on in this lesson, a practical approach to structuring decisions when uncertainty is involved.
3. Integrating insights for decision-making
When faced with uncertainty, it helps to break down the decision process into clear steps. If you're familiar with the data science process, you'll recognize these steps.
Before diving into analysis, be clear about the question you're trying to answer. Are you evaluating risk? Comparing options? Estimating an outcome?
4. Integrating insights for decision-making
To identify key uncertainty factors, ask: What do we not know? Are we unsure about customer demand? Costs? Market trends? Recognizing uncertainty helps guide your approach.
5. Integrating insights for decision-making
Based on the type of uncertainty, select a method that fits. If you need to estimate probabilities, you might use Bayes' Theorem. If you're comparing multiple possibilities, scenario analysis can help. If risk is a concern, Monte Carlo simulation might be the best choice.
6. Integrating insights for decision-making
Once you've applied the technique, interpret what they mean. Are the risks acceptable? What trade-offs exist?
7. Integrating insights for decision-making
Finally, communicate your insights to stakeholders using visualizations and key metrics.
8. Example: expanding into a new market
Let's apply this framework to a real-world scenario.
A company is considering expanding into a new market. They know the opportunity is promising, but they are unsure about customer demand, costs, and potential risks.
The first step is to define the problem. In this case, the key question is: Should we expand into this market?
The second step is to identify uncertainties. The biggest unknowns that could impact the decision are customer demand, cost fluctuations, and competitor reactions.
9. Example: expanding into a new market
The third step is to select the right technique to analyze these uncertainties. Since both customer demand and costs are highly variable, a Monte Carlo simulation is chosen. This method allows the company to model a wide range of possible outcomes by simulating thousands of potential conditions. Instead of relying on a single revenue projection, the company can assess the likelihood of different levels of profitability, helping them make a more informed decision.
Note that you're not necessarily restricted to one method. You could include different methods to look at the problem from another angle. For example, scenario analysis is useful for exploring different cost scenarios.
To structure different choices, you might map options using a decision tree.
10. Example: expanding into a new market
The fourth step is to run the analysis and interpret the results. After running the Monte Carlo simulation with different demand and cost assumptions, the company finds that in 60 percent of simulations, the expansion is profitable, while in 20 percent, they break even, and in 20 percent, they experience a financial loss.
The fifth step is to communicate findings effectively. The results of the simulation can be visualized using probability distributions, showing executives how likely different revenue levels are.
Since the analysis shows a strong probability of profitability but some risk of loss, the company decides to move forward cautiously. Instead of launching at full scale immediately, they choose to start with a pilot program in select locations. They also develop a contingency plan, such as adjusting pricing or marketing strategies, if early demand does not meet expectations.
11. Let's practice!
Time for some final practice!