Hey Victor, No worries: a few years of coaching taught that learning is a non-linear journey :) I think your problem with structuring cases is that you are overthinking the main buckets. Intro and example My approach is to see buckets as "areas of analysis" you want to see during your case. They do not need to be directly linked to a specific sub-question of the case, but rather should give you a wide pool of analyzed data to then converge on the solution. I always advice my coachees to use this approach because it's just simple and easy: if you start overthinking at 01:00, you are in for a looong interview :) Let's take a standard market entry case: "A global PE fund contacted McKinsey to understand whether they should enter the market of Padel (i.e., a racquet sport, not dissimilar from pickleball) in the US. The sport is pretty popular in Europe, but almost unknown in America, so they are wondering what the size of the market is, what competition could look like, and what services they could offer." The approach I would suggest for a beginner is this one: Market - Size, growth, and profitability (i.e., understanding how big the pie is) - Customer segments (e.g., more appealing to 20-40yo or 60+) - Competitive dynamics (e.g., fragmented vs. consolidated) Target business model - Product/Service offering (i.e., what are we selling here?) - Operating model for the PE fund (e.g., buying a few smaller companies and merging, launching a new player, entering with an adjacent Port.Co.) - Capabilities (e.g., operations, labor, assets, etc.) Financial analysis - Investment criteria (e.g., payback, ROI, ...) - Revenue (P*Q) - Costs (FC + VC, incl. investment) Risks - Market-specific (e.g., competition with pickleball, ...) - Firm-specific (e.g., lack of sports management experience, ...) Why this approach is helpful Let's go a bit more in depth to see why I think this type of structure helps you. 1. It's extremely comprehensive: you are starting with a framework that considers a lot of aspects that one should check out in a real case. Can you really consider a market entry without thinking about customer, competitive dynamics, investment criteria? These would be lacking in a Process approach, or in an approach that uses sub-questions -- the approach above does not leave any stone unturned 2. It's easy to understand: the interviewer can follow you nicely. You avoid taking a bolder/newer approach (e.g. Process, which is more tailored to implementation projects, typically less strategic) and the interviewer instantly recognizes. I am pretty confident that out of 100 McKinsey Senior Partner, 95 of them got in by using this standard approach :) 3. It still follows a process. In a real engagement (I have done more than a few market entries for PE), you would do exactly the same: first, you start the workstream on the Market, then you start thinking how you play in that market (i.e., Target Business Model); once you have gathered those two pieces of info, you can create your business plan (i.e., financial analysis); lastly, you want to caution your client about risks (or maybe 'Implementation and risks') Why the Process approach is more complicated With Process, let's say that you have 4 buckets: (i) Strategy Definition, (ii) Pilot Project, (iii) Implementation - 0 to 1, (iv) Extension - 1 to 10. A few issues: 1. The 'Strategy Definition' bucket is so broad, that it's hard to be comprehensive/exhaustive within this bucket. You would almost need to replicate my buckets above: a strategy needs info on customers, competitors, needs a financial analysis, potentially risks, etc. 2. The content of the next buckets is heavily dependent on the first: what can you say about Pilot Project if you don't know what the strategy is (e.g., roll-up of padel courts vs. entering with a PE-owned company doing a spin-off) 3. You are implying something you don't know about the client engagement: what if the client doesn't care about implementation (e.g., most PE funds have an implementation or operating improvement team)? Then, you threw away one bucket --> Is the Process approach always wrong? No! But it's harder because it's more specific and has a structural dependency from Bucket1 to Bucket2 ... to BucketN. If the question is about "Launching a product", then it might make more sense, but a lot of cases are not a good fit for this approach Why the sub-question approach is more complicated Let's take the three sub-questions as buckets: (i) what the size of the market is, (ii) what competition could look like, and (iii) what services they could offer. 1. You are missing the main point (i.e., would this be a profit-generating business with a good ROI)? That's because the real answer to the case might be the logical consequence of the three questions you are asking (on top of some more elements), but it's a different analysis that requires different instruments (i.e., revenue and cost estimates, investment criteria, etc.) 2. You are not being super-MECE. Very often, two questions are intertwined: here, you are splitting market between the size and the competition -- but those are two sides of the same coin. When you do a market analysis for a company, you tell the not just the $$$ but also the competitive landscape, I don't love dividing these two things in different buckets. 3. You sound very mechanical. A big part of every real MBB project is taking the questions the client has and transforming them into a coherent plan of attack (with workstreams, deliverables, etc.). My experience showed me that the client might be missing some angles or might be asking only 1 of the 3 real questions. If you just repeat what you hear, you don't sound like someone that can rethink problems/questions from a different perspective -- while I am sure you are a very nuanced thinker! --> Is the sub-question process always wrong? No! It's great when you have three self-contained, fairly independent tasks. I don't think this is the right case for that: the main task we have to deliver is giving an answer to a broader question (absent from the sub-question), i.e., is this a market we should enter? Conclusion My key takeaways: The standard "areas of analysis" approach is by far the easiest to start with (and, again, probably used by 95% of McKinsey Senior Partners in their interviews 15-20 years ago) All the frameworks that you see are just potential examples. Every structure must be fully tailored (both in its main buckets and underlying bullet points) to the specific nuances of the case at hand A very helpful way to view these frameworks is as "potential puzzle pieces." You should select and assemble the right pieces based on the specific needs of the prompt you are given. My answer to your question "why do I keep getting lost?" is that you are probably overthinking with very complex approaches :)))) Do I have to always use the standard approach? No! Start from the simplest one, but then feel free to add more layers of complexity once you feel more confident For advanced candidates: if you are able to formulate a solid hypothesis based on the initial clarifying questions, it is highly encouraged to include that at the beginning/end of your structure. For instance: "My initial hypothesis is that this is a highly fragmented market, and therefore, leveraging M&A could drive growth and capture economies of scale, provided it is financially feasible." I hope this helps, but please let me know if you have other questions. While coaching at Berkeley, I built a pretty comprehensive guide on MBB Structure Frameworks. Feel free to DM me if you are interested in it (the invite is open to everyone!). It helped a ton of coachees learn enough "puzzle pieces" to then compose their best structure in a few seconds. Best, Tom