Estimate the number of search queries answered by Google per second?

estimation Market sizing
New answer on Feb 29, 2020
3 Answers
5.3 k Views
PK asked on Feb 18, 2020
Director of Product at PayPal. Based in SF. Worked at Twitter, Yelp, Salesforce, Houzz in the past.

In order to understand QPS for Google we can look at the # of queries handles by Google per day and then infer the QPS from the daily data assuming a uniform distribution.

I would like to take a top-down approach and understand how many queries are down across the world per day. Also, assuming that by queries specifically, we are referring to searches on and not via other Google services.

The world population is 7B and 50% of those have access to the internet.

Thus we have about 3.5B people and it is also safe to assume that anyone who has touched internet also uses Google.

Of the above 3.5B we can assume that:

1. 20% of them are in the developed world. (700M)

2. 60% are from developing countries (2.1B)

3. 20% of them are from underdeveloped countries. (700M)

We can also assume that users in developed countries have access to high-speed internet thus we can assume that the number of queries they do is quite high. While I work in tech and run several dozen google queries per day, it is not true for most people to query so often. Based on my experience I think the range is very wide and thus I will choose 5 as the avg number of queries per day per user.

For developed countries, this number will be a lot lesser as not many people have access to smart phones. Thus I am assuming the number of queries in the developed world is a lot lesser. Assuming its 40% of developed world. 2 queries per user per day

For underdeveloped countries, this will be even lesser and maybe more like 0.25 queries per day per user.

Based on the above we can evaluate queries for each part of the world:

Developed --> 700M * 5 queries per day --> 3.5B queries per day

Developing --> 2.1B * 2 --> 4.2B queries per day

Underdeveloped --> 700M * 0.25 --> 350M queries per day

Thus total query per day --> 3.5B + 4.2B + 350M ~ 7.8B queries per day --(A)

Thus, total queries per second --> (A)/(24*60*60)

Note: I am not feeling very confident about my choice of how I arrived at 5 queries per user per day in a developed country. Any suggestions on how to best estimate the daily queries in developed, developing, and underdeveloped countries?

Overview of answers

  • Upvotes
  • Date ascending
  • Date descending
Best answer
Content Creator
replied on Feb 20, 2020
Top rated Case & PEI coach/Multiple real offers/McKinsey EM in New York /6 years McKinsey recruiting experience

I like your approach it is consistent and well thought through- also the internet says there are 3.5B Google queries a day so you aren't very far off.

A few things to consider to change your assumptions

1. Search is no longer capitalized by google (e.g., most searches for products start on Amazon)

2. Verticalized search is hurting google (e.g., LinkedIn for jobs)

3. Number of search queries is therefore lower in Developed markets

4. Access to internet in developing markets is not necessarily the same as in the US for example. A lot of 'smart' phones in India are very basic and have limited access in the traditional sense

All the best,


Was this answer helpful?
Content Creator
replied on Feb 28, 2020
BCG |NASA | SDA Bocconi & Cattolica partner | GMAT expert 780/800 score | 200+ students coached


I would base my segmentation on population age, since young people are used to search on google for every doubt and old people don't use internet.


Was this answer helpful?
Content Creator
replied on Feb 29, 2020
McKinsey | NASA | top 10 FT MBA professor for consulting interviews | 6+ years of coaching

I find it a good approach


Was this answer helpful?
Udayan gave the best answer


Content Creator
Top rated Case & PEI coach/Multiple real offers/McKinsey EM in New York /6 years McKinsey recruiting experience
Q&A Upvotes
92 Reviews
How likely are you to recommend us to a friend or fellow student?
0 = Not likely
10 = Very likely