Market Sizing

estimation market sizing
New answer on Aug 17, 2021
3 Answers
1.1 k Views
Anonymous A asked on Aug 14, 2021

How would you approach “How many supermarket trolleys are in the city of London?” 

Overview of answers

Upvotes
  • Upvotes
  • Date ascending
  • Date descending
Best answer
Anonymous B updated the answer on Aug 15, 2021

This is a neat question. I have an approach, but I'm not sure if it's optimal.  We need:

 

(a) # of markets in London

This will be London pop. X (Markets/people)

Finding the number of people served by a single market is tough and the thing I am most unsure of.  It's roughly driven by how much food a person buys per week vs. the volume of the market and how fast a market can restock. I'm not sure how to work out the latter, but my hometown of 20k has 3 supermarkets, so let's just say Markets/people = 3 markets/20k people = 1 Market/6k people (-ish)

Then we have London pop. X (Markets/people) = 8 M X 1 Market/6k people = 1300 markets  in London

 

(b) # of carts per market

At one cart per person, this is just the maximum number of people in the market at once. 

Suppose 1 shopping trip is 1 hour, and a person shops once per week.

 

6 k people/week X (1 week/7days) X (1 day/8 shopping hours) = 110 people/hour in the store, on average.  If 1 hour is 1 trip, this is 110 carts per store.

------------------

Now multiply these two numbers together:

Markets in London X carts per market = 144 k shopping carts

 

Some extra considerations: 1 hour per shopping trip might be a little long. It could be more like 30 min.  Also I assumed that the shoppers are spread out evenly throughout an 8 hour window.  There are probably peak hours, and a market likely has enough carts to meet that peak demand.  So maybe instead of 110 carts/store we could call it 150.

 

I think the weakest point here is definitely in understanding how many people a single market serves.  Interested in hearing if anyone else has additional thoughts.

 

 

(edited)

Was this answer helpful?
7
Anonymous A on Aug 15, 2021

Thanks for replying, I agree it is a very complicated exercise. I wonder how to account for large supermarkets which normally have more trolleys vs city mini markets

Anonymous B on Aug 15, 2021

Yeah I'm not sure. In what context did this come up? I would be surprised if anyone asked you to go deeper than this in an interview

Anonymous B on Aug 21, 2021

On additional approach is to bypass the total number of markets altogether. In this approach, the end goal is to just compute the number of people in London shopping simultaneously. It doesn't matter whether they are shopping at big or large markets, they will all need carts. I'll assume a single trip is around 30 minutes, so we want to know how many people are shopping simultaneously during a 30 minute window. First, we should break the population of London (9 million people) down into people who do shopping trips. There are 2 main groups: those in families and those not in families. Take 50% of people to be in families, where just one person does the shopping for an average household of 4. This is about 1.1 million shoppers. For the single people, all of them do shopping so that's 4.5 million shoppers for 5.6 million shoppers in total. I'll assume that everyone shops once a week. You could argue that single people need to shop less frequently, but I see shopping as more of a scheduled weekly ritual. This comes out to 5.6 million trips per week = 800 k trips per day. If people are only shopping during an 8 hour window, that's 100k trips per hour, which is 200k trips per half hour. This means that during every 30 minute window (during normal shopping hours), there are 200k people simultaneously shopping. To get a utilization rate, we can say 1/5 of the shoppers are shopping for families and definitely use one. The remaining 4/5 of shoppers are shopping only for themselves, and use a cart say 50% of the time. This is 60% utilization overall. Multiplying this by the number of shoppers gives 120k carts. This is basically the same answer as before. But this is the magic of market sizing, all of your ignorance cancels out in the end (though I think some would strongly disagree with this perspective). Btw, if you're interested in seeing how insane market sizing can be, I'd check out the Crafting Case Market Sizing module in their free course. The BMW call center cost estimation is astounding.

(edited)

Clara
Expert
Content Creator
replied on Aug 17, 2021
McKinsey | Awarded professor at Master in Management @ IE | MBA at MIT |+180 students coached | Integrated FIT Guide aut
Was this answer helpful?
Ian
Expert
Content Creator
replied on Aug 16, 2021
#1 BCG coach | MBB | Tier 2 | Digital, Tech, Platinion | 100% personal success rate (8/8) | 95% candidate success rate

Hi there,

You'll really get the best out of this exercise if you try it yourself. Why don't you post what you're thinking and we can help?

Hint: Do you think this makes more sense top-down or bottom up?

General Tips

Remember that there's rarely a "best" answer with market sizing. What's important is that you break down the problem the way it makes sense to you. Importantly, break it down so that the assumptions you make are the ones you're most comfortable in.

For example, do you know all the major brands? Great go with that. Do you understand all the segments of that country's population (either age or wealth or job breakdown)? Go with that. Do you know the total market size of a similar industry? Then break it down that way.

Some tips:

  1. Just like in a case, make sure you understand the question - what are you really being asked to calculate
  2. Decide whether a top-down or bottom-up approach is best
  3. Figure out what you know you know, and what you know you don't know, but could estimate
    1. This helps you determine how to split out buckets
  4. Stay flexible - you can start with a "high-level" market sizing, but gauge your interviewers reaction....if it looks like they want you to do more...then go along level deeper in terms of your splits
Was this answer helpful?
How likely are you to recommend us to a friend or fellow student?
0
1
2
3
4
5
6
7
8
9
10
0 = Not likely
10 = Very likely