A Guide to Decision Trees

Ever feel like decision-making is one part science, two parts guessing, and a dash of “let’s hope this works”? Well, welcome to leadership, folks.

You can download a copy of  A Guide to Decision Trees here.

Whether you’re running a multinational or trying to figure out where to get lunch, you’re faced with a never-ending stream of decisions. Daily, weekly, annually. Some are easy; some make you want to crawl under your desk and pretend email doesn’t exist.

The problem is that a lot of decisions are made on intuition, based on gut feelings, half-baked data, or by the loudest person in the room. As a leader, part of your job is ensuring your decision-making is consistent, aligned with those around you and clear to the teams you lead.

So, wouldn’t it be great if there was a single system to make smart decisions? A system that helped examine your decisions to determine potential outcomes, assess various risks and ultimately predict your chances for success? Well, good news, there is.

It’s called a Decision Tree.

Decision trees take all the craziness swirling around in your head — the what-ifs, the risks, the potential rewards — and lays it all out in a simple map. They’re a tool to effectively visualise your decision making process and predict your expected values. So, you can avoid unwanted outcomes and get to the best possible result. Systematically and strategically. With a clear, consistent decision-making approach on hand, you might even foster sustained performance across your organisation. That’s a win! Whether you’re deciding to invest in a new market, launch a risky product or try a new creative approach, decision trees provide a method to the madness.

So, buckle up and let’s learn how to build one.

Step 1: Start with the Big, Hairy Decision: The Root Node

Ok, the first thing you need is a root. Your root node is the main decision you’re trying to make — the one keeping you up at night. Maybe the reason you’re reading this. It might be something like, “Do I quit my job and become a TikTok influencer?” or “Do I yield my crops now or later?”

For now, let’s be serious. As a marketer, you often face critical decisions regarding market expansion. So, let’s imagine you’re in the pursuit of year-on-year business growth.

Your Root Node is: “How do I grow our market share?”

This decision will have ripple effects across budgets, teams, and markets, so you need to think it through.

Whatever your hairy, scary root is, get it down and the rest will branch out from there.

Step 2: Branch Out: Draw Your Options

Now, from your root node, you need to draw branches representing the different actions or paths you can take. You probably already had your options in mind. What do you need to decide between?

Let’s draw our first two branches:

Branch A: Launch Paid Campaign in New Market (US)
Branch B: Increase Spend on Current Campaign in Existing Market (UK)

 

Step 3: Pinpoint The Possibilities: Identify Your Leaf Nodes

Once you have established your options, you’ll want to work out the possible outcomes of your glorious (or disastrous) choices.

But before you get ahead of yourself, you need to think about the factors that could affect your decision. Is there anything that might get in the way? Is there a risk? Will a chance of rain ruin your crops?

We call this a Chance Node.

A chance node is something outside of your direct control. It’s where fate – or market demand, consumer behaviour and competitor actions – comes into play. You can estimate the odds, but you can’t guarantee the result. It reminds you to plan for the good and the bad.

We can identify a Chance Node for Branch A.

Chance Node: Competitor Launch
There’s a chance that a competing company enters the market at the same time. This could hurt your sales.

Do you know how to decode your nodes?

Whether you draw your decision tree by hand or use a fancy software template, you need to know that there are two types of leaf nodes:

Square Leaf Nodes:
A square means that there’s another decision to make. It means you’re not done yet.

Look, your starting root is a square.

Circle Leaf Nodes:
A circle represents a chance event. It’s not a new decision, it’s just a possibility. One that you unfortunately can’t influence.

Look, your Competitor Launch is a circle.

 

Step 4: Crack Out the Crystal Ball: Estimate Your Results

Once you’ve accounted for your chance nodes, it’s time to predict the outcomes. This is where the real work begins.

Now, nobody’s asking you to become Nostradamus here. Unless you have access to a crystal ball, you’re going to use a mix of sales figures, market research, past performance, and some good old-fashioned educated guessing.

Let’s go back to our example…

Branch A: Launch in New Market (US)
This is your bold move, but it comes with the risk of a competitor launch. You’ll need to predict your sales and market share for both potential outcomes.

Branch A1: If Competitor Launches
• Predicted Market Share: 2-3%
• Predicted Sales: $2M annually

You’ll be facing a strong competitor here, so grabbing even 2-3% of the market will be tough. You estimate that this equates to $2M in annual sales.

Branch A2: If Competitor Does Not Launch
• Predicted Market Share: 4%
• Predicted Sales: $4M annually

If you’re the only major player in the market, the world’s your oyster. A 4% share in a big market like the US could land you $4M in annual sales.

Branch B: Increase Spend in Existing Market (UK)
There’s less uncertainty here but will you see explosive growth?
• Predicted Market Share Growth: 1-2% increase
• Predicted Sales: $2M annually

You already own a significant piece of this market, so any additional spend would likely push your market share up by 1-2% and boost sales by $2M.

 

Step 5: Bet on the Future: Assign Probabilities

Now that we have our predicted sales for each outcome, it’s time to estimate how likely each scenario is. This is where you consider market conditions, competitive behaviour, and your own analysis.

Sometimes probabilities come from talking to experts. Or, if you have the luxury of testing a decision before going all-in, try it and derive probabilities from there.

Branch A1: If Competitor Launches
Probability: 60%
Based on what you know about your competitor’s strategy, there’s a good chance they’ll launch alongside you.

Branch A2: If Competitor Does Not Launch
Probability: 40%
While less likely, there’s still a solid chance that you’ll be the only one making waves in this new market.

Branch B: Increase Spend in Existing Market (UK)
Probability: 100%
Luckily, there’s no uncertainty here. You control your spending, so this option is a guarantee if you go for it.

So, what’s the point of probability? Let’s find out…

Step 6: Crunch the Numbers: Calculate the Expected Values

It’s time to bring in some maths. It might not be the sexiest part of marketing but hang in there. You’re about to figure out which decision gives you the best shot at success.

For each branch you’ve outlined, you’ll calculate the Expected Value (EV) of the financial outcomes based on the likelihood of each scenario.

The formula is simple:
EV = Predicted Outcome x Likelihood

Let’s apply this to our tree…

Don’t fear, the Expected Value is just a fancy way of saying the weighted average or mean. If there’s more than one outcome, it’ll be the sum of all possibilities.

Look at our tree. We’ve calculated the EV for every option like this:

Branch A1: If Competitor Launches
EV = $2M x 60% = $1.2M

Okay, you’re not quite done yet. Branch A has two possible outcomes. So, you need to calculate the Expected Value of proceeding with Branch A regardless of what happens. You’ll work this out by adding both Expected Values together.

In our tree, Branch A would be:
EV = A1 + A2 = $1.2M + $1.6M = $2.8M

Step 7: The Moment of Truth: Make Your Decision

Right, it’s crunch time. You’ve mapped every option, calculated risks and rewards, and stared at more numbers than a tax accountant in April. Now’s the moment of truth. Do you go bold or play it safe?

Let’s look at what we’re dealing with…

On the face of it, Branch A is the winning bet – $2.8M. But you’d be rolling the dice on that pesky competitor showing up. Branch B is the safe, solid option — it’s a guaranteed $2M.

But wait, sometimes your big Expected Value can shrink after factoring in costs. So, you might want to go the extra mathematical mile and calculate your net gain. To figure out if your costs eat up your profit, simply subtract your expected costs from your EV and you’ll be good to go.

Now you’re ready to ask yourself: is it worth the gamble?

If you love the thrill of high risk/high reward and trust your gut (even when it’s in knots), Branch A is calling your name. But if you sleep better knowing you’ve locked in sure-fire results (even if they’re a bit modest), then Branch B might just be your ticket.

Decision Trees: Your New Best Friend

Decision trees tell you how it is. They don’t sugarcoat the risks; they clearly show what’s at stake, and what you stand to gain.

Sure, at the end of the day, it’s still you that makes the call. And ultimately, this comes down to your business strategy and risk tolerance. Thankfully, with decision trees, every decision is a step forward. By embracing the learning that comes with each choice, you’ll continue to make better decisions in the future.

Now, whether it’s launching a campaign or deciding between two mediocre lunch options, you’re better equipped than 90% of decision-makers out there. Those are good odds.

Good leaders make good decisions. Master Decision Analysis and more with the Mini MBA in Management.