If you own a business, I assume you are not unfamiliar with solving problems. Troubleshooting supply chain issues. Catering to customers who seem to want nothing to do with anyone else but the CEO. Papering through bills with revenue calling for attention. Every business owner has been in any of those moments. After all, the job of the ‘big’ man or woman at the top is mainly solving the big existential problems.
While most business problems arrive as a knot and everyone has a theory, the instinct is to start fixing whatever feels most urgent. That is how big executives and teams burn months on the wrong problem.
Trying to introduce you to a tool called the issue tree to help you think through a problem and arrive at useful insights and solutions. The issue tree is the simplest tool for cutting through the knot. Used by consultants at McKinsey, BCG and Bain, it breaks a big, vague problem into smaller, distinct pieces you can actually investigate. You don’t need a consulting background to use it. You need a pen, a blank page, and some discipline.
But wait. What is an Issue Tree? An issue tree is a diagram that starts with one key question on the left and branches into sub-questions on the right, each level more specific than the last.
Think of it as an outline for your thinking. At the far right of the tree are questions concrete enough to answer with data, an interview, or a quick test. Two rules make it work. These rules are known as MECE: Mutually Exclusive, Collectively Exhaustive.
To be clear, mutually exclusive means branches on the tree don’t overlap. Nothing sits in two places at once. Collectively exhaustive means together, the branches cover the whole problem. Nothing important is missing.
If your branches overlap, you’ll duplicate work. If they have gaps, you’ll miss the real cause.
Before we dive into the steps, I would like to cite an example of a business deploy the issue tree briefly. A regional food distributor sees margins slip. The root question he asked was, why has gross margin dropped from 24% to 19% in six months?
Moving ahead to create branches on the issue tree, he invited his team for a brainstorm session. The team splits the issue tree into price and cost branches. Price data shows average selling prices are steady, so that branch is eliminated in a day. Under costs, they split it into purchase cost and delivery cost. Delivery costs are flat. Purchase cost, though, has jumped, and the next level (by supplier) reveals that one supplier raised prices 30% while the team kept buying from them out of habit.
In a week, the team has a precise problem and a clear action: renegotiate or switch suppliers. Without the tree, they might have spent a quarter running a pricing review that would have found nothing.
See how it works? Let’s take the steps for using an issue tree to tackle business problems.
Step 1: Define the Problem as a Sharp Question
A tree is only as good as its root. “Sales are bad” is not a root. “Why did revenue fall 15% in the last two quarters, and what will restore it to last year’s level by December?” is.
Note that a good root question must be Specific, measurable and actionable. You will need to spend real time here. Ten minutes sharpening the question saves ten days of analysis.
Step 2: Choose the Right Type of Tree
There are two main types. Diagnostic trees (why trees) find the cause of a problem. For example, ‘Why is profit declining?’ Solution trees (how trees) generate options for reaching a goal – ‘How can we grow revenue by 20% next year?’
Use a diagnostic tree when something has gone wrong, and you don’t know why. Use a solution tree when you know the destination and need routes. Many projects use both. They take diagnoses first, then build a solution tree on the cause they find.
Step 3: Split the Problem into MECE Branches
This is the heart of the method. Find a logical way to divide the problem, and use one dividing principle per level. For instance, in determining process steps, break a customer journey into awareness, consideration, purchase, and retention. Ask where the drop-off happens. Or when determining segments, split by customer type, region, product line, or channel.
Here is a simple diagnostic tree for falling profit:
Why has profit fallen?
├── Revenue is down
│ ├── Fewer units sold
│ │ ├── Fewer customers
│ │ └── Customers buying less often
│ └── Lower average price
│ ├── Discounting more
│ └── Shift to cheaper products
└── Costs are up
├── Fixed costs
│ ├── Rent and utilities
│ └── Salaries and overhead
└── Variable costs
├── Raw materials
└── Logistics and distribution
Notice that every split is clean. A cost is either fixed or variable, never both.
Step 4: Go Deep Enough, Then Stop
Keep splitting until each end branch is something you can test. “Costs are up” is too broad. “Logistics costs per delivery rose 22% after the fuel price increase” is testable. Most trees need three to four levels. If you’re on level seven, you’ve probably drifted into detail that won’t change the decision.
Step 5: Prioritise the Branches
Now the tree earns its keep. Don’t investigate every branch. Use early data, quick conversations, and your own judgment to find where the problem most likely lives.
Ask two questions of each branch:
1. Impact: If this is the cause, how much of the gap does it explain?
2. Ease: How quickly and cheaply can we test it?
Start with high-impact, easy-to-test branches. Cross off the ones the data rules out. This is how you avoid boiling the ocean.
Step 6: Test Hypotheses and Build the Answer
Turn each priority branch into a hypothesis: “We believe revenue fell because we lost three major accounts.” Then test it against data. Each result either confirms the branch or eliminates it. As you prune, the tree converges on the real driver, and your recommendation writes itself, already structured and defensible.
However, there is a key caveat. The issue tree is not magic. It is structure. It forces you to define the problem, break it logically, focus on what matters, and let evidence lead. Whether you’re diagnosing a decline, planning a market entry, or pitching a client on a strategy, the habit is the same: before you solve, map. Draw the tree first. The answers get much easier to find.
