Top-down or bottom-up chart interpretation

Hi there, 

when getting a complex chart, I would suggest to do the following. Take a moment to process the data & ask questions to interviewer, summarize intersting observations you make, bring observations together to get to insights & recommendation. If the chart is super complex it is best to not look at numbers first but at proportions, etc. Running numbers might lead you into a rabbit whole if you have not understood the big picture first.

Andreas

Hi David,

From a quick glance I don't see a problem with how you are communicating it!

Ultimately, if you communicate clearly, concisely, and in the context of the objective/question then you are doing well! You have done that in your above example.

Hi there, 

Sharing with you a resource that explains in more depth top-down vs bottom-up:

• Free Guide: Mastering Structuring & Brainstorming

Best,
Cristian

In my opinion, you are 1000% correct. 

Always go top down. Always!!!

Observations and insights are irrelevant if they are not leading to a recommendation or solving the specific problem. If they are not tied up to the case objective, they are just “random (and irrelevant) interesting facts”. Bain calls this being “answer-first”.

Moreover, top down communication, particularly when analyzing a chart is much more effective from a time management point of view. You only focus on what is important and you convey immediately WHY it is important.

You are communicating as CEOs (and consultants) want you to communicate. You start with the answer upfront, and then provide your arguments. As such, the consultant/CEO can analyze your arguments already knowing where it leads to. 

If you go bottom-up, the interviewers have no idea where you are heading to (and most likely you don't know as well). Worse, if you start with observations you are basically telling them what they already know. They don't need your help to read slides. They need your help to draw insights and provide recommendations based on them.

This is actually the MOST IMPORTANT rule to follow when looking at exhibits. You first think about the case goal, and then look at the exhibit to understand how it impacts the recommendation. If you are going bottom up, it is VERY UNLIKELY that you will follow this thought process. Once again: always be “ANSWER FIRST”. Particularly for Bain interviews, this is a MUST.

Of course, other may have a different opinion, but I am quite adamant on this one. Top-Down!

(editiert)

Hi there,

Top down only works if you're actually clear about the ‘bottom’. However, when doing chart analysis, this is often not the case yet because you're exploring as you're talking. Here's my advice:

• Step 1: Talk briefly about the obvious insights - this will impress nobody but is important for your own understanding and buys you a little time
• Step 2: Identify opportunities to show your quant skills - almost all exhibits have ‘hidden’ insights that can only be uncovered by crunching numbers
• Step 3: Execute math where simple and/or align with interviewer for more complex computations that take a moment - explain why you're doing this
• Step 4: Contextualize your insights with the data points from the case so far i.e., backward looking, and also suggest the relevance going forward

Hope this helps a bit. Best of luck!

Moritz

Great question! 

So generally McKinsey and other interviewer-led companies love it if you go top-down. They expect you to take some time to look at the chart before giving an answer and if you can do it this way up it's a large bonus. 

Interviewee-led interviews expect you to go level one and up. This is because they expect you to talk through the chart as you're analysing it and so it's more natural that you'll get the smaller insights first and then work your way up. 

Good luck!

Hi there,

Chart interpretation should always be done in a top-down manner, where you address the actual question / reason why you're looking at the chart first.

Contrary to some comments and beliefs here, please don't provide random insights just for the sake of it, esp if unrelated to the actual question, just to “show off” your quant skills. That's not going anywhere. 

Instead, focus on answering the question, add supporting evidence, use hyphoseses where helpful and be clear on the next steps to push the case further. Interviewers appreciate control over the narrative and action-bias.

Hope this helps - if you'd like to learn more on how to effectively interpret charts in any type of scenario, feel free to reach out via PM.

Regards, Andi