The McKinsey Approach to Problem Solving

Today’s article is based on McKinsey Staff Paper 66: The McKinsey Approach to Problem Solving. I’m going to give my own spin on it, but I would definitely recommend reading the paper and my previous article on it.

The goal of this article is to give you a better idea of how McKinsey consultants approach problems, why they approach problems this way, and what they look for in the case interviews.

The McKinsey Approach to Problem Solving can be broken up into six parts:

1. Creating a Rigorous Problem Definition

2. Structuring the Problem

3. Prioritizing Issues

4. Plan & Execute Analyses

5. Synthesize Findings

6. Developing Recommendations

Creating a Rigorous Problem Definition

Knowing exactly what problem you’re trying to solve is the first (and most important step) of the problem solving process.

Unless you know exactly what the issues on the table are and what your criteria for success is, you won’t be able to plan efficiently and recognize your solutions when you see them.

The way McKinsey does it is they have a “problem statement worksheet” that they work to fill out before they begin any engagement. This worksheet has the following sections:

1. A SMART (Specific, Measurable, Action Oriented, Relevant, & Time Bound) Initial Question. This would be something like: How can we improve our client’s profits from its retail fuel operations by $250 million in the next 5 years?

This is why it’s very, very important to understand what the initial question is when you’re doing a case interview. Always repeat the question and the key information, and make sure you understand the client’s objective. Even better, see if you can quantify the objective – size, time frame, etc.

The reason this is important is because if you don’t know what you’re shooting for, you’ll never know when you’re done and what’s in scope and out of scope.

Having a very specific initial question up front prevents you from doing unnecessary analysis and keeps you safe from losing focus / going on tangents (this is why in my article on how to take good notes I recommend that you write the objective at the top of your notes and keep referring back to it).

2. A Discussion of Context. In establishing the problem definition, it’s important to take stock of the initial situation. This would be things like the internal and external situations / complications the client is facing such as capability gaps, money available for investment, industry trends, and the like.

What does this mean for case interviews? I’d say the discussion of context should occur after you repeat the problem statement. Usually you have the chance here to ask clarifying questions. Don’t waste them! Ask about the industry trends, the client’s financial situation, how the client’s business model works, etc.

It’ll make the problem much easier to understand, and ultimately solve.

3. Success Criteria. As we discussed earlier, you need to have an understanding of the client’s objective. Usually there will be a set of quantitative and qualitative criteria that your solution needs to meet. It’s extremely important to understand these criteria so you can understand when you have a solution.

In the context of a case interview, it’s very important you get an understanding of your client’s objective and how it will be measured. For example, the client’s objective might be to lower the ROIC by 10%. Make sure you understand what ROIC means, how to calculate it, and why they want to lower it. This way, you can check each of your potential solutions to see if it would meet the client’s goal.

4. Scope and Constraints. Scope refers to the market or market segments you’re operating in, and constraints are things you can’t do.

For a case interview, this would be establishing what market you’re in (i.e. US, Europe, etc) and also taking the time to eliminate potential solutions before you dive into the case.

For example, in profitability turnaround cases one set of solutions would involve entering new markets or introducing new products. You can save yourself a lot of time up front by asking if the client finds the above solutions acceptable or if they’re out of scope.

Structuring the Problem

Issue Tree

In previous articles I’ve discussed how to create an initial structure. What I wanted to mention here is that apparently McKinsey creates either “hypothesis trees” or the more common issue trees/initial structures that we’ve discussed before.

The difference appears to be that the hypothesis trees are much more focused (see above). Instead of covering all possible areas, the trees are produced by listing out all relevant hypotheses and each are explored systematically.

The normal way of structuring the problem into different categories is better if you know less about the situation. The hypothesis tree approach seems to be better suited for more structured situations like improving profit, sales, and lowering costs. I’ll write more on this later.

Prioritize the Issues

The article recommends prioritizing each hypothesis for exploration. The way they talk about it is creating a 2×2 matrix with impact and ease of impact on each side (though you can use other criteria) and categorizing each hypothesis into different quadrant.

The hypotheses that are high impact / high ease of impact are the ones that you should consider first.

An example of this might be any profitability enhancement case: you can always increase prices, sell more product, or lower costs. Increasing prices is the option with the highest ease of impact and can also be high impact as well, so you’d often start there. This is how you prioritize issues, and it makes a lot of sense to talk about these in cases as well.

Plan & Execute Analyses

What is unusual about the McKinsey method of problem solving is that they already have the solution in mind when they began an engagement. From the previous steps we discussed they already have a set of prioritized hypotheses they want to explore.

All information gathering / interviews with experts / analysis is planned in advance to explore each hypothesis in detail and uncover the evidence to prove or disprove the existing hypotheses.

It’s interesting — I can see how this might come across as either arrogant or lazy to the people paying for the studies. But really, it’s just an example of focused problem solving. You need to go in with a solution first and investigate only the minimum number of areas to prove its validity, because otherwise you’ll waste time and money conducting unnecessary analysis.

Synthesize Findings


McKinsey seems to favor the dot-dash storyline method of constructing a synthesis, and arranges the storyline into a pyramid structure with supporting evidence (dashes) under main points / arguments (dots).

The purpose of a synthesis is to create a clear, easy to follow, inductive, and logically unassailable conclusion for the listener.

Develop Recommendations

Coming up with strong recommendations isn’t enough. You need to develop a pragmatic action plan with the following three elements:

1. Relevant initiatives with sequence, timing, and mapping to make the plan happen. It should also offer a balance of quick wins, sustained impact, and consider available resources and competing priorities.

2. Clear owners for each initiative.

3. Key success factors and challenges for delivering on the initiatives.

When doing a case interview, it would be very useful to mention these three things at the end of the case discussion and in the next steps / risks section of your conclusion.


The McKinsey Approach to Problem Solving can be broken up into 5 parts:

1. Creating a Rigorous Problem Definition

2. Structuring the Problem

3. Prioritizing Issues

4. Plan & Execute Analyses

5. Synthesize Findings

6. Developing Recommendations

Hope it helps!


2 thoughts on “The McKinsey Approach to Problem Solving

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s