0
$\begingroup$

I'm trying to create weighted averages for 2 different variables, but am struggling with how to do the right calculation. Let's say I have the following, by Market Size / Penetration % / Market Share %, and then for each region you multiple the variables ($ = Market Size * Penetration * Market Share):

Region A: $\$900,000 / 75\% / 5\%$

Region B: $\$200,000 / 75\% / 0\%$

Region C: $\$5,500,000 / 0\% / 0\%$

Region D: $\$28,400,000 / 0\% / 0\%$

This leads to a sum of $\$33,750$ as only Region A will be a non-$0$ number.

Obviously total market size is $\$35,000,000$, but how do I calculate the overall penetration and market share %?

Originally I was going to calculate average penetration by multiplying each region's market size by its penetration, summing those values, and then dividing by the total market size (and then doing the same for market share).

This gave me a penetration of $2.4\%$ and a market share of $0.1\%$, but then when I multiply the total market size of $2.4\%$ and $0.1\%$, I don't get $\$33,750$.

I get a total of $\$1,061$ instead.

Any help would be much appreciated!

$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

The penetration% and market share% are not independent. The regions with higher penetration% (A and B) have relatively higher market share% than the other regions.

Therefore, you can't simplify multiply the overall penetration% with the overall market share% to get the total dollar amount.

$\endgroup$
4
  • $\begingroup$ Browngreen, what if they were independent though? What if you input random numbers for penetration and market share, would you be able to get some type of overall % for each? $\endgroup$
    – NuclearPenguins
    Commented Jul 12, 2017 at 18:16
  • $\begingroup$ Even if in truth the two variables are independent, if you input random numbers for a small sample like this (only four regions), the two variables would likely be dependent within the sample (unless for example, all of the penetration numbers were the same). For a much larger sample, the two variables would be roughly independent, and the final value could be approximated using the overall penetration and market share %'s. $\endgroup$
    – browngreen
    Commented Jul 12, 2017 at 18:43
  • $\begingroup$ Ok, how would the formula work for that then? $\endgroup$
    – NuclearPenguins
    Commented Jul 12, 2017 at 18:51
  • $\begingroup$ If penetration and market share are independent and random and you have a large sample, you could approximate with the formula you used at the end: total market size $\times$ overall penetration% $\times$ overall market share%. $\endgroup$
    – browngreen
    Commented Jul 13, 2017 at 1:23

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .