Segmenting by age groups for market sizing?

Question

Hey guys,

could some of you give me their methods of segmenting people per age groups? Is it reasonable to assume that groups of equal age length (say 0-20, 20-40 ... etc.) have a somewhat equal proportion of people in it?

I know that in western countries the distribution might be slightly skewed as there is a greater percentage of elderly people. However, in my opinion this would make the sizing too sophisticated. What I would do is that I would acknowledge that the distribution might not be evenely and that there might be a greater proportion of elderly but I would think that for coming up with a quick market sizie it is sufficient to build on that assumption.

Is that good?

Answer 1

It is usually fine to assume that population is distributed evenly by age. However, some countries are well-known for having a younger or older demographics. If you get a sizing case for such a market, it is usually a good idea to tweak the assumptions slightly or atleast show the interviewer you're aware of these nuances.

For example, sizing the smartphone market in India, and doing a segmentation by age, you can assume that age <30 is 50% of population (median age is 28), and assume an even split for >30 population. It might look something like <15 (25%), 15-30 (25%), 30-50 (20%), 50-70 (20%), >70 (10%).
Similarly, if you get a question for Japan (median age is 50 i think), you can assume >50 population to be 50% and <50 to be remaining 50% and divide evenly within that.

The intent is to differentiate yourself in an interview by showing that you're aware of these nuances. It is perfectly ok for you to just mention these factors, and align with the interviewer that you are going to make simplifying assumption of dividing population evenly by age.

Was this helpful?

Answer 2

Hey there,

Ideally this will depend on the country/geography and population sizes and which industry you are looking at e.g. social media users or shampoo users. I advise following options depending the case/industry that you get given:

1. Age ranges by life cycle stages: babies, children, adolescents, adults, middle-age, and seniors

2. Generation-based: Baby boomers, Gen X, Millennials, Gen Z

Use common sense and keep it simple. You are not expected to come up with very precise ages ranges.

Answer 3

Hi there

You cannot assume the age groups are the same

Part of market sizing is being able to determine which segments (whichever they may be) are different based on what you know of the world.

If you're market sizing for India, you need to recognize that half the population is young.

If you're market sizing for Japan, you need to realize that half the population is old.

If you don't do this, you will 100% be dinged.

One other people of advice: Try not to do 0-20, 20-40, 40-60, 60-80, 80+. This overcomplicates! just do young, middle-aged, old

Answer 4

I think it comes down to what you are solving for where you want to make sure your increments are relevant to the problem (e.g., child, adult, vs. elderly population, etc.). From a pure math perspective, equal age increments makes things more simple so that would be my default. The geographic nuances and other assumptions are important to share upfront with your interviewer who should also give guidance if they want you to segment differently.

Answer 5

Hi there,

Yes, your approach is totally fine.

If the country is skewed on young/old, just mention to the interviewer you know the population is skewed and ask confirmation if it is fine to split equally.

Most of the time they will say it is fine and you can continue without complications.

Best,
Francesco

Answer 6

Hi, in addition to the solutions proposed by the other coaches in the discussion, I would like to suggest similar cases in the platform to practice with:

https://www.preplounge.com/en/consulting-forum/how-much-would-you-charge-to-clean-all-the-windows-in-seattle-4965
https://www.preplounge.com/en/consulting-forum/market-sizing-milk-consumption-5087
https://www.preplounge.com/en/consulting-forum/how-would-you-calculate-the-value-of-a-cow-4982
https://www.preplounge.com/en/consulting-forum/estimate-number-of-traffic-lights-in-a-london-5692

Hope it helps,
Antonello

Answer 7

Hello!

Indeed, the good thing about segmenting is the fact that those groups:

Don´t act/react them same: so you can tailor the approach
They are not evenly distributed: if you look at the population pyramid, it has totally nothing to do in developped vs. delopping countries. You can easily see a couple in Google, that will suffice to get the idea.

Hope it helps!

Cheers,

Clara

Answer 8

Hi there!
Segmentation by age groups can be strikingly different under factors such as marital status (a mother of 25 and a woman without a child have a lot of differences in behavior), work, financial wealth, and geography.
but, more often than not, the following groups have the most common features of behavior:
16-18, 18-21, 21-25, 25-30, 30-35, 40-45, 45 - 60, 60 and above
There may be other variations, very often used:
12-17, 18-24, 25 - 34, 35-45, 45 - 55, 55 and above

Each country has very different demographic indicators, so different groups may have completely different percentages of the population.

Was it helpful?
GB

Answer 9

The entire point of estimation techniques is to quickly get to a reasonable answer with reasonable assumptions. It is totally acceptable to assume equal distribution as long as you state that you are doing so and why (e.g., for ease of calculation). If the interviewer does not agree they will ask you to change your approach. Yes it is not perfect but with clear communication you will get the direction you need from the interviewer.

Udayan