The story so far: Algorithms play a crucial role for search engines as they process millions of web searches every day. With the quantity of information available on the internet growing steadily, search algorithms are becoming increasingly complex, raising privacy and other concerns and drawing the attention of regulators. Last month, U.K.’s digital watchdog said they will take a closer look at algorithms, seeking views on the benefits and risks of how sites and apps use algorithms, as well as inputs on auditing algorithms, the current landscape and the role of regulators.
An algorithm, essentially, is a series of instructions. It can be used to perform a calculation, find answers to a question or solve a problem. Search engines use a number of algorithms to perform different functions prior to displaying relevant results to an individual’s search request.
Tech giant Alphabet Inc’s Google, whose flagship product is the Google search engine, is the dominant player in the search market. Its search engine provides results to consumers with the help of its ranking systems, which are composed of a broad set of algorithms, that sort through web pages in its search index to find the most appropriate results in quick time. Its search algorithms consider several factors, including the words and expressions of a user’s query, relevance and usability of pages, expertise of sources, and the user’s location and settings, according to the firm.
While Google captures a significant chunk of the general search market, there are alternative search engines such as Microsoft’s Bing and DuckDuckGo available for users to explore. The latter, a privacy-focused search engine, claims it does not collect or share users’ personal information.
In January, market leader Google generated 61.4% of all core search queries in the U.S., according to database company Statista. During the same period of time, Microsoft sites handled a quarter of all search queries in the U.S.
As the algorithms used to deliver results would vary from one search engine to another, when a user inputs a query, the results would also differ. Moreover, results from different users would be rarely similar, even when searching for the same things, since the algorithms take into account multiple factors, like their location.
Algorithms are often built using historical data and for specific functions. Once developed, they go through frequent updates from the companies to enhance the quality of search engine results presented to users. Most large search engine providers also bank on machine learning to automatically improve their users’ search experience, essentially by identifying patterns in previous decisions to make future ones.
Over the years, Google has developed search algorithms and updated them constantly, with some major updates like Panda, Penguin, Hummingbird, RankBrain, Medic, Pigeon, and Payday, meant to enhance some function or address some issue. In March, it introduced another update to improve the search engine’s ability to identify high-quality product reviews.
Search engines exert huge control over which sites consumers can find. Any changes or updates in their algorithms could also mean that traffic is steered away from certain sites and businesses, which could have a negative effect on their revenue.
The search giant’s trackers have allegedly been found on majority of the top million websites, as per a DuckDuckGo blog post. “This means they are not only tracking what you search for, [but] they’re also tracking which websites you visit, and using all your data for ads that follow you around the internet,” it added.
According to a Council of Europe study, the use of data from profiles, including those established based on data collected by search algorithms and search engines, directly affects the right to a person’s informational self-determination. Most of Google’s revenues stem from advertisements, such as those it shows consumers in response to a search query.
DuckDuckGo, in addition to providing an alternative to Google’s search engine, offers mobile apps and desktop browser extensions to protect users’ privacy while browsing the web. The privacy-focused firm, in a blog post, said that editorialised results, informed by the personal information Google has on people (like their search, browsing, and purchase history), puts them in a “Filter Bubble” based on what Google’s algorithms think they are most likely to click on.
These search algorithms can be used to personalise services in ways that are difficult to detect, leading to search results that can be manipulated to reduce choice or artificially change consumers’ perceptions.
Additionally, firms can also use these algorithms to change the way they rank products on websites, prioritising their own products and excluding competitors. Some of these concerns have caught the eye of regulators and as a result these search algorithms have come under their scrutiny.
The European Commission has fined Google €2.42 billion for abusing its market dominance as a search engine by giving an illegal advantage to another Google product, its comparison-shopping service.
Moreover, under the Commission’s proposal on the Digital Services Act, transparency measures for online platforms on a variety of issues, including the algorithms used for recommending content or products to users are expected to come into force.
“Majority of algorithms used by private firms online are currently subject to little or no regulatory oversight,” U.K.’s Competition and Markets Authority has said earlier in a statement, adding that “more monitoring and action is required by regulators.”