“Hey Google, go find this for me using everything you know about me; use the revenue from the advertising auction as payment; and be quick about it, I only have a second to spare!”
Google has been effective in doing just that. In the most recently reported fiscal year, Google's revenue amounted to 136.22 billion US dollars. Google's revenue is largely made up by advertising revenue, which amounted to 116 billion US dollars in 2018.
Now read on
When you perform a query on Google, it will typically return search results on to the screen in significantly less than one second having utilised somewhere between 700 and 1000 computers scattered across various Google data centres.
But before you ever enter your query, Google has been diligently harvesting data from hundreds of billions of web pages using web crawler software, also referred to as spiders. The spiders follow links to discover pages and then stores that data in a massive, constantly updating index.
Sites that are larger or more popular, will be visited by the spiders more frequently to capture new information. And it is important to remember that a Google search is not an entire search of the entire web, rather it is a search of the index it maintains.
Getting back to the query
When you submit the query using the google search bar, a number of tools are used to identify and analyse the words to determine what the search means.
These tools include natural language processors to correct spelling errors, establish the correct meaning for any synonyms used in the query, and hunt for trigger words such as “today”, “flight”, “shop”, “weather”.
In the event of any spelling errors, Google adjusts the spelling to what it believes to be correct and uses that version of the query to submit to the Google search algorithm. In the situation where the query is too general, or ambiguous then Google will use its association database (where it stores past queries, web pages and the associations between them) to refine the search query.
Google also uses context, which can be words and phrases associated with a single domain, or your search and browser history, or your location, or anything else known about you deemed by Google to be relevant.
The query is sent to the index servers, and Google lists all the websites that rank highly for the query. Then the query is sent to the document servers to retrieve web pages or documents. At this time, snippets are created to provide descriptions for the results of the search.
Finally, using the PageRank and other factors, Google decides which pages to place at the top of the list shown on your screen.
And now to the personal data
And not to forget, while all this activity is going on, the Google search engine uses the keywords (e.g. Nike, travel, restaurant, locksmith, investment, coupon, sale etc etc) you used in your query to hold an auction amongst advertisers who offer products related to those keywords.
In fact, Google makes information available to advertisers that enables them to be selective about which groups of people they are interested in as potential customers based upon their demographics or past history of visiting that website. Google itself provides a rich seam of information on this topic at https://support.google.com/google-ads/ .
Of course, Google has been collecting and retaining information since you took that first initial step of accessing Google products and services in return for handing over your personal data. How many years ago was that; all remembered?
So what information is available about you to pass to the advertisers: Gender, age, location, family status, income, browsing history, devices, educational level, home ownership, mobile network provider, what products you have bought, what apps you have installed, what times of day you use Google, your email history, your website history, and what you are doing right now.
Does Google know what you are doing right now!
YES, YES, YES! for example; is your signal moving from one mobile phone tower to another (you are travelling), or are you close to Nike store and are searching for “Athleisure wear”, or you are at an airport, or are you at the location Google recognises from your map queries as “home” or “work” or “boyfriends flat”.
And if all this data from your online life was not enough, look at the current direction of travel with Google. Acquiring credit-card data (check out article on BBC) so that in-store spending can be further analysed to supplement everything they know about online spending..
Next time you check your search results, notice the Ads
If one or more advertisers are bidding on the keywords deemed relevant by Google, then there is an auction which determines whether or not the ad is to be shown, and in which ad position it will show on the page.
Typically, unless your keyword is so obscure as to not warrant the interest of an advertiser, the first handful of listings in the search results are paid advertisements.
Those advertisers have bid in the online auction (quicker than the time it takes to fulfil your search query) to gain the coveted first or second spot on those listed results.
As most queries are now made using mobile telephones, its really only the first few spots in the list that are worthwhile as the majority of mobile users do not scroll further down the search listings.
Should you “click” on an advertisement, that information will enrich the “AdWords” file of the advertiser and they can use that information to understand the trigger responsible for selling the product to you and how that information can be leveraged in future auctions.
In reality, when you submit a search query, you are saying “Hey Google, go find this for me using everything you know about me; use the revenue from the advertising auction as payment; and be quick about it, I only have a second to spare!”
Length: 1059 Words.
Time to read 5 Minutes
2018, December 28
Photo by Franki Chamaki on Unsplash