The Google Estimation question is really 80-95% logic and 5 to 20% math. However, the average candidate gets so worried about the math, they don’t pay attention to the logic. My framework for approaching the estimation question puts logic front and center. 

I put the focus on the basic equation and then the logic for how we fill in the variables to reach a basic number. 

Tips & Tricks

• Variables: No more that 2 to 3 variables for your equation

  • occasionally a 4th variable is okay if you need to divide for the final answer (think # of servers needed)

• Logic For Assumptions: Use logic to fill in values for each variable

• Round up or down: You want numbers ending in 0, 2 or 5 because the math is easier.

• Think of two solutions before diving in: The follow on question is often, can you think of another approach. If you spend 5 to 15 mins on one approach, it an be difficult to come up with another, so think of two ways before you go too far. (Think: Top Down vs Bottom Up)

• Build a cheat sheet of common facts: Think major world populations, size of emails, images, videos, server capacities, etc

A Few Key Steps

1. Word Problem Approach: Develop an equation with 2 to 3 variables (4 at most)

   • Sense check: If you can think of another approach before you dive too far in, it will help. They

2. Use logic to fill in the variables

   • i.e. US population is ~300M and there are ~100M households so I will assume 25% of 100M households will use my product

3. Volunteer to question your own assumptions

   • Break down 1 or 2 of your variables looking at High, Medium and Low Users (as an example) or High, Medium and Low frequency of use, etc.

4. Review your Logic

   • Compare your first result to your newly questioned results

Cheat Sheet Example:

• US population: 330M

• California population: 40M

• NYC population: 8M

• World population: 7.8B

• Size of continental US: 3M square miles

• US households: 130M

• Average people per household in the US: 3

• Life expectancy: 72

• Median household income: $65,000

• Weight of an average car: 4,000 lbs

• Percent of population under 20 or over 65: ~25% (give or take)

Technical

• Amazon S3 Standard cost: $0.023 / GB / month

• Average file size for a 90-min 720p movie: roughly 3.5GB

• Average file size for a smartphone camera picture: roughly 3-5 MB

• Average CTR for a search ad: 1.91%

• Average landing page conversion rate: 2.35%

• Average WiFi bandwidth: ~10Mbps

• Cost of iPhone 12 Pro: $999

• Cost of Google Pixel 5: $699

• Cost of Amazon Echo (4th Gen): $99

General sense of revenue (2020)

• Dropbox revenue: $2B

• Airbnb revenue: $3B

• Google revenue: $181B

• Facebook revenue: $86B

• Apple revenue: $274B

• Amazon revenue: $386B

• Netflix revenue: $25B

• Google (Alphabet) net income: $40.27 billion

• Apple R&D expenditure: $18.75 billion

General sense of user populations

• Netflix (Q2 2021): 209 million subscribers

• Google G Suite (March 2020): 2 billion+ monthly active users

• Uber (Jun 2021): ~1 million drivers in the US

• Twitter (Q1 2018): 192 million daily active users

• Number of Americans that own a smart speaker (Jan 2020): 60 million

• Number of products Amazon sells: over 12 million (not including Amazon marketplace sellers, which brings the total to 350 million)