I'd like to get your thoughts of what is the best final recommendation structure is for McKinsey. Thanks!
What final recommendation structure would you recommend for McKinsey interviews?
Overview of answers
1. Recommendation first, in clear and exact terms. E.g. "I recommend the client shut down shut down their china factories to return capacity to US."
2. Follow up with reasoning next: E.g. "There are three reasons (edit: MBB loves the "threes") for this: 1) we immediately save xxM in costs, which is yy% of overall revenue, thus improving margins by zz% which was our original goal 2) we reduce negative PR around our chinese mfg 3) long-term benefit from new US-friendly mfg policies" -- these reasons are just hypotheticals btw.
3. Follow up with 2-3 risks (risks == things that are not fully understood/known): E.g. "1) US capacity is more difficult to grow in face of unknown demand projections 2) difficult to serve APAC demand from US due to unknown political climate 3) heavy up-front divestiture costs"
4. Finally, if applicable - you may want to "upsell" a bit by saying: as next steps, we'd love to work with the client to offer the best divestiture plan, and look into the ideal capacity expansion scenarios in the US.. etc etc.
Hope this helps
the previous answers are all fine; however they do not cover the biggest mistake I see in final recommendation: not repeating the objective.
Many candidates forget part of the objective when providing the answer, in particular when there is a time constraint on the final sum up. They thus do not provide the correct answer to the interviewer due to that.
Also, you should not feel constrained to provide three reasons. If there are two reasons supporting your conclusion, don’t try to put a third one – it will be very awkward and the interviewer will notice that you are making up the third reason. I saw that several times from candidates: “There are three reasons for that: (i) We would increase sales by XYZ; (ii) we could keep a competitor out and …ehm.. [silence – making up incorrect answer]”.
Thus my suggestion for your final sum up would be the following:
- Repeat the objective: “Dear Mr CEO, you asked us to find why profits are going down and a way to increase them by $5M in 2 years” – This allows you to be 100% sure you are fully answering the case. More than 50% of the people don’t do this, so you will be already above average with this step
- Provide an answer-first conclusion: “After a quick analysis we found the main reason for the profit decline is increasing labour cost abroad. We also found a way to increase profits by $3.5M. This is coming from two sources: (i) relocating part of the production in Mexico and (ii) starting our own website.” Again, if there are only two areas you found to increase profits, you have to mention two. Of course, if you found a third one, would be good to add it.
- Provide next steps – including risks: “As additional steps, we would propose the following: (i) There may be some issues with the relocation in Mexico due to potential layoffs in the US which would be worth to be investigated. (ii) We would be also happy to continue the analysis to identify additional $1.5M in profits. To do so, as we saw it could make sense to investigate the launch of a new product in the market. We would like to provide a full analysis on this in a couple of days”.
The best structure is to start with your answer first, support it with data (3 main insights of your case), precise the next step if needed and repeat your answer.
The key to your recommendation is to be data-driven and client-friendly. This a general consulting thing. I am really not sure there are different recommendations structure depending on the firm you are interviewing with.
The only specific to McK is that they really apply 'the Pyramid Principle' (book written by a McK consultant, B. Minto). The core idea is to 'start with the answer first'. However, most of the firms apply it. Why ? It is client-friendly.
I have seen many people preparing for cases, and explaining me what they have been doing chronologically for the last 30 min.
If I am the CEO of Fortiune 500 company, i meet you in the elevator,I don't wanna lose a minute with you. I don't pay you to know what you have been doing. I pay you to get answers, know they are solid (data-driven), and that i can build my strategy on it. That is what is tested when you're asked to conclude. Will you be able to perform well if you meet a CEO in the elevator or will you fuck up ?
Here is an example :
Regarding the profitability issue we have been working on, the best solution seems so far to target the new segment XXX with new product YYY to boost revenues with a marketing campaign because :
1. We have no levers to reduce costs of current products, revenue is the issue
2. Other segments are not growing, this segments grows 20% a year
3. We have all the capabilities and channels to address it.
I am on my way to check potential logistics issues with the Operations department, but with the available data, our recommendation is to target new segment XXX.
Whatever you're doing, always wonder what is the most-client friendly. Regarding conclusions, it consists in (maybe repharsing the problem/question in 1 sentence and) starting with your answer first, supporting it with data, and repeating your main conclusion.