First of all, I think there are three distinctions in principles between solving a McK case and a BCG case.

1. Initial structure: compared to McKinsey cases, the initial structure should be very similar on the top 1-3 levels, but BCG style cases don’t require you to have a comprehensive list of key factors and drivers. Having a few key factors and drivers given as examples are always better, but for BCG style cases, there is no need to strive for a complete list. McKinsey cases are interviewer-driven, and the interviewer will tell you which sub-brunch to explore after the initial structuring. Therefore, the initial structuring phase is the only phase that you can show your structure for all sub-brunches. Unlike McKinsey cases, BCG cases allow you to explore all sub-brunches during the case; therefore, there is less requirement in showing your structures completely up front.

However, from my point of view, the above doesn't mean that your structure can be non-MECE on the top 1-2 or even 1-3 levels. I have met with case partners who just pick one sub-brunch in a complete issue tree to start with, assuming that he will expand the issue tree later in the case. There are two main drawbacks to this: (1) If you don't have a complete workplan to start with, it's impossible to expand your issue tree logically and rigourously latter (2) You miss the opportunity to showcase your ability to structure

This is the sample issue tree for the Airline case. For a BCG case, your initial struture should ideally cover up to the orange level, and minimally, cover up to the yellow level

2. Hypothesis driven and synthesis: because BCG style case needs the candidate to drive by herself, it’s a lot more important to stay very hypothesis driven and give synthesis every step of the way so that your audience can follow you with minimal effort. Hypothesis is why you want to explore one particular sub-brunch and synthesis is what your conclusion is after you are done exploring one sub-brunch.

3. Rigor: there are two parts to rigor, one is logical thinking and the other is logical execution. Logical thinking is heavily tested in any case interviews but BCG style cases also test logical execution because BCG cases are candidate-driven. Logical execution is, what Victor Cheng refers to as process of elimination. Basically, you go down to each sub-brunch of your issue tree, and determine if there is any potential for solutions. Generally, you want to drill along one sub-brunch until you are sure there is nothing there before switching to another.

I will focus on explaining the above three principles when analyzing this BCG sample case.

A very simple but to-the-point case recap would look like this:

A low-cost airline company’s profitability decreases due to higher fuel cost. The goal is to return the company to profitability in the short run, and ensure profitability in the long run.

Some key initial insights that ideally should have crossed your mind after the case prompt:

The airline’s business model is currently based on a low-cost strategy, which implies competitive pricing.

This case format online is interactive, but the first two questions together is basically the initial structuring question. Like McKinsey cases, it want you to outline a MECE appraoch with some key factors and drivers. This is what I explained in Principle 1. You need to have at least a structure that covers up to the yellow part of the sample issue tree.

The next two revenue related questions are where you have started drilling down your issue tree. You have chosen to explore revenue first. In a real case interview, you need to offer your hypothesis that you think revenue may have more potential so that you'd liek to look at revenue first.

I listed two options given after Question 5. I listed both of them because it doesn't matter which option you want to explore first. There is no right or wrong. Both options belong to the same brunch, and eventually you want to drill down to each of them.

This way Question 6 is answered is clearly a demonstration what a candidate should do in a BCG case interview. You should expand the sub-brunch of your issue tree logically (MECE) and then explore each sub-brunch. Here, ticket sales is expanded into number of tickets sold and ticket price. Then to increase revenue, one should explore both to see if there is any opportunities.

I listed two options for Question 7. I listed both of them because it is when I choose whether I want to drill further into the option of increasing price. I would choose to drill further to make sure that there is no hope in changing price so I would go for the second option. It gives me more confidence to eleminate the price sub-brunch.

Question 8 is a quantative question. I chose to ask for historical data but the interviewer said there is none. And then some current data is given to me. The calcualtion is easy, the way one should approach this quantative question should be the same with McK style cases.

Question 9 is another quantitaive question with interpretation of a figure. It's easy. The same principles of sovling a quant question apply.

Ideally, Question 10 shouldn't be asked in an actual interview, because the candidate is expected to interpret the answer from Question 9 proactively. But it's extremely important to take a close look at the answer to Question 10 because it is when you conclude that increasing price is not an option, make the synthesis, and eliminate this sub-brunch, and switch to another sub-brunch.

Question 11 shouldn't appear in an actual interview either because the candidate should have offered to switch to looking at ways to increase the number of tickets sold without being prompted by the interviewer.

Like I said last time, Question 11 shouldn't be a question in the sense that once you reach the conclusion that increasing price is not an option, you should proactively suggest what to do next. By process of elimination, you should have proposed to do sth. about the number of tickets sold.

The logical next step may be a bit hard in the sense that you need to break the "number of tickets sold" into its driving factors: No. of Tickets Sold = No. of Customers * No. of Tickets/Customer = No. of Flights * Customers/Flight * No. of Tickets/Customer

Therefore, to increase the No. of Tickets sold, one needs to increase one or two or all of the three factors: No. of Flights, Customers/Flight, No. of Tickets/Customer

At this point, what I would suggest is to start with one of the factors that's most helpful in terms of increasing short term profitbility, and then proceed to factors that are less important with regards to this goal. Let's say you choose to attach the "Customers/Flight" factor first.

This interviewer is really just saying that there's nothing we can do about this factor. At this point in time, you are given two options to choose from during the online interactive process. One is “The client could opt for bigger aircraft to accommodate more passengers on each flight. Thus, the total number of passengers would increase.”, and the second is “Well, are there any flights at specific times of the day where the occupancy is low? The client could look at offering special discounts to customers for these flights.”

I recommend choosing the first one. Whenever you get to a point where you feel you may have enough proof to eliminate a branch, you should feel free to do so. The second option is when the candidate decides to drill further. It's not wrong, but it will waste you time. It's also against the 80/20 rule. If in general there is no opportunity in improving the load factor, spending time to see if certain flights' load factor can be improved just won't yield much.

What you can continue to drill is other ways to increase Customers/Flights, which is to increase the size of the plane.

The interviewer gives you proof that increasing the size of the plane doesn't work.

Now it's time to investigate the other factors. What the case did was to move on to increase the number of flights. “Okay. In that case, the client could increase the frequency of flights or fly to new destinations.”

The interviewer rejected this possibility too. So what's left is tickets purchased per customer. The case didn't explore this but I think it should be something to be considered. But let's just assume you moved on to the last factor but the interviwer said no again. Since there are three factors you think that constitutes the number of ticets sold and you have eliminated all of them, the logical conclusion is that there is nothing you can do about the volume side either.

Once you realize the above, it's time to synthesize! I feel if there is one thing you don't need to do in McK cases but absolutely need to do in BCG cases is synthesizing.

If you look at the initial structure, you have elimiated the yellow branch "Ticket Sales". So say it out loud that it seems there is no opportunity in increasing ticket sales. Then dirve the case to the next branch, which is non-ticket-sales.

The value added branch is quickly dismissed. And then you move on to the cost branch.

The structure for the cost part is fuel cost vs. non-fuel cost. The interviwer suggested that there isn't much to do with fuel cost so that branch is elimiated. What's left is the non-fuel cost.

I will skip the steps in the cost section, basically the candidate drives through all sub-branches and reach the conclusion that there is nothing to do with respect to the cost either.

Don't forget to synthesize at this step.

This online sample case is hard in the sense that it ends up with an analysis of whether the firm can survive the storm given that there is no short term strategy to revert profitability.

It requires some business intuition. Don't fret if you find it hard. Business intuition grows as you practice.

I will not walk you through the rest of the case. I think I have made my points clear, at least I hope I did. 1. There is no substantial difference between McK and BCG style cases. 2. BCG styles case require you to drive more in that you need to logically go through every branch of the issue tree and use the process of elimination, and you need to synthesize every time you elimiated a major sub-brunch. 3. Start practicing McK cases first, give yourself time to learn and grow. Don't fret if you can't handle a BCG case well now.