# How to Perform Decision Tree Analysis

What is decision tree analysis?

Your company has to make a decision that carries some weight. It could be as simple as “should we expand our sales team?” or as complicated as “should we acquire one of our competitors?”

So, how do you make a decision this daunting?

There are quite a few factors that go into high-stakes decision-making: multiple and at times competing options, the probabilities of success or failure, and the potential upside (or downside) for every possible choice—the list goes on.

Trying to do this with multiple stakeholders can quickly get overwhelming. You need a proper method for objectively determining what your “best” choice actually is. A great way to do this is with decision tree analysis.

An example of a decision tree (Source: Wikipedia)

Why is decision tree analysis important?

Decision tree analysis is the process of graphically charting out business decisions. In a nutshell, you list out every decision and every possible consequence while assigning probabilities and utility values (usually expressed in dollars) to each outcome.

When done right, decision tree analysis compartmentalizes (and, ultimately, simplifies) complex decision-making into neatly organized, comprehensible choices.

Here’s how a decision tree works:

1. Business decisions are mapped out as branching paths (e.g., Choice A vs. Choice B).
2. Each choice has specific, objective consequences that are displayed as branches from each decision. Each of these consequences is assigned a value. Sometimes these values are straightforward dollar amounts (e.g., how much a new hire costs). Other times, the utility value may not be as obvious (e.g., replacing one tech solution with another, comparably priced one) and must be determined using other means.
3. Every consequence has a probability expressed as a percentage value. After all, a single choice may lead to multiple outcomes (e.g., hiring Jane could lead to more sales, no change, or fewer sales) that may be outside your control.

Decision tree analysis examples

Of course, you don’t need to conduct a full decision tree quantitative analysis to make the right decision. On a day-to-day basis, the decisions you have to make are relatively simple, and the right choice is usually obvious.

But when it comes to complex, costly decisions that may have significant consequences for your business, decision trees help you visualize every possible outcome of your choices. Proper visualization and data model methods give you much better visibility into not just what the “best” decision might be, but also what the second, third, and fourth best work outcomes are.

### A simple, everyday decision tree analysis example

Let’s take a simple (non-business) example. It’s Friday night, and you and your partner are hungry. Do you order delivery or have a sit-down dinner at a local restaurant?

It’s a binary choice, but your outcomes are dependent on a number of associated factors. For instance, you know that both you and your partner would be happier having a sit-down meal. But it’s also Friday night, meaning it’s too late to get a reservation and you might have to wait in line. And waiting in line will definitely make you and your partner less happy.

Each choice also comes with risks: dining at the restaurant has high utility value, but there’s a small chance you could end up waiting for an hour. Ordering delivery is the less volatile option, but you run the risk of passing on a great night out because you’re playing it safe. There’s also a very small chance that your delivery simply won’t arrive!

Sure, it’s a silly example. But you can see how the decision tree analysis approach can be useful for even mundane day-to-day decisions.

Now that we’ve seen a simple example, let’s take a look at a more applicable business example.

### A complex, business decision tree analysis example

Let’s say you’re considering building out a new sales team. The question you have to ask and answer is whether it’s worth the cost.

Here’s a non-exhaustive list of what you have to consider:

• Building the team: Do you hire a sales team of two, three, or four? Do you outsource hiring to a staffing agency that charges you a finder’s fee? Or should you save that money and just do it yourself for free? Do you split an existing sales team or build one from scratch? Do you hire mostly junior salespeople with a strong VP of Sales or an all-veteran team of closers?

• Cost-benefit analysis: For each outcome, what potential increase in revenue might you see? What about the direct monetary cost for each outcome? What about indirect costs that are harder to measure (e.g., if you do the hiring yourself, it will take longer and take up time that could be better spent elsewhere).

• Probability of success: For each situation, what is the likelihood of success or failure? With so many variables, determining the right probability values may not be straightforward at all. And how might that impact the cost-benefit breakdown for each outcome?

As you can see, even a relatively straightforward business decision isn’t simple when you actually break it down into its component parts.

Better decisions = better business

If you look at the bigger picture, the math is simple: better decisions are more likely to achieve better outcomes for your business. Decision tree analysis is a fantastic tool for risk management and decision making because it can carefully identify benefits and drawbacks, as well as the probabilities of success and failure, for every possible choice you could make.

Using the hypothetical example above, for example, you can account for different levels of risk for each decision:

• If you hire a support team of four people rather than two, there is a higher likelihood that your revenue will increase. However, there is a 100% chance that your expenses will double.

• Alternatively, outsourcing sales to a remote team with a commission-only or mostly commission-based structure may significantly lower your risk. But you may be sacrificing more long-term profits to higher commissions in exchange for lower short-term salary expenses.

• If you decide not to hire an additional sales team, can you quantify the opportunity cost that you’re passing up on? How does that stack up to the cost savings of not going ahead with the plan?

Whatever the case may be, decision tree analysis is a sensible way to quantify your choices and figure out which decision produces the best potential set of outcomes with an acceptable degree of risk.