How would you estimate the number of new users Amazon gets in a day?

1. CLARIFY:
   1. Are we focused on US or global? You choose. (I will focus on global relationships.)
   2. Is it OK to focus on Amazon Prime / eCommerce memberships (not Video, Music, Whole Foods, etc.)? Yes.
   3. Would you confirm we are channel agnostic? Yes. 
2. BACKGROUND: Amazon is an eCommerce giant that offers Amazon Prime membership for a subscription fee. Prime offers deals on items like free 2 day shipipng. There is also Amazon Prime Day during which Prime members access to tons of items on Amazon for big discounts. Prime Membership is offered in ~ 18 countries. 
3. EQUATION: # of forecasted Amazon Prime users / Time forecast to convert users to Amazon Prime
4. BREAKDOWN UNKNOWNS: 
   1. # of Forecasted Amazon Prime Users:
      1. Total Population of Markets with Amazon Prime: Amazon Prime in ~18 countries, including US, India and Brazil. US (330M), India (1.4B) and Brazil (213M) are the most populuous regions - representing a total of a little over 2B. Let's assume the total population of markets including the other smaller countries is ~2.5B.
      2. Internet Penetration: Assume 60% of those countries has access to internet. 2.5B * .6 = 1.5B people with internet access.
      3. Amazon Market Capture: Assume Amazon will eventually capture 40% of the eCommerce market it arrives in. 40% includes people using Amazon but have NOT bought a membership. 1.5B * .4 = 600M Amazon users. 
      4. Amazon Prime Conversion: 
         1. Amazon Prime is most established in the US. There are ~ 112M US Amazon Prime users, which is roughly 1/3 of the total US population (112M Amazon Prime members / 330 M US population). Let's assume that the US penetration is high, so other country penetration will be less than the US. 
         2. We can divide  likelihood of conversion of an Amazon user to Amazon Prime member into 3 groups. Let's assume there is an even distribution of the 600M Amazon users across these buckets (i.e. 200M of each type of user). 

         3. Likelihood of Conversion	Total Amazon User Population	Likelihood of Conversion to Amazon Prime	Total Converted Amazon Prime Members
            Low	200M	10%	200M * .1 = 20M
            Medium	200M	15%	200M * .15 = 30M
            High	200M	20%	200M * .2 = 40 M

         4. Total Amazon Prime Members: 20M + 30M + 40M = 90M
   2. Time Forecast to Convert Users: 
      1. Amazon Prime Users / Year: Let's assume it takes 5 years to convert these users to Prime. Let's also assume there is an even distribution of users converting to Prime each year. 
         1. 90M Prime Users / 5 years = 18M new Prime users / year
      2. Amazon Prime Users / Day: 18M new prime users per year / 365 days / year = ~49K new Prime users / day.

• Assumption: 
  • US 
  • Amazon marketplace users (not prime or AWS etc)

# or users with no amazon account - users in age group 18 - 75(~60%  of 350mn) = 210mn

50% of users who already may have an account : 105mn users 

105 mn  users do not have an account - let’s say 70% of this user base will create an account. Some families create 1 account per family rather than individual account;  some may never convert. ~73.5 mn users who do not have an account but may create one, 

Assuming the conversion to happen over the next 5 years. 

So 73.5 / 5 = 14.7 mn users a year.

14,700,000/365=40,273.9726

40K new users/ day fo US. 

Taking international market into account: 

Amazon provides service in around 18 countries. Being an American company with 20+ years of trust, user adoption is high in US as compared to other countries. Some countries like India where its operations started late, adoption rate maybe significantly lower compare to countries like US,UK or Germany.

So, instead of doing linear calculation of multiplying 40K by 18 , I would take an average of 9 countries . 

40,000 * 9 = 360,000 user per day all over the world where Amazon operates.