Whenever you have a pressing question, your most natural instinct would be to whip out your smartphone and type that question into Google. On the search results page, you’ll be greeted with 10 links that Google’s algorithm has determined to be the websites that best answer your question.
However, this is set to change as Google wants to redefine the search experience by adding more AI-powered capabilities. While this technological advancement would introduce a new level of interactivity to the search process, we need to consider the impact it may have on both users and digital publishers.
Why is search so valuable?
Ranked among the world’s top five companies in market capitalization, Google, along with its parent company Alphabet, owes much of its success to its search engine. In Q1 2023, Google’s advertising revenue from Google Search amounted to some USD 40.3 billion, constituting a significant portion of the company’s overall revenue of USD 69.7 billion.
Google’s search algorithm has demonstrated its exceptional accuracy compared to other search engines, where it has consistently shown its capability to effectively connect users with high-quality websites that align with their specific search intent for any given query. As a result, Google has dominated the search market, accounting for more than 93% of searches worldwide, while Bing in second place only accounts for less than 3%.
However, Microsoft has forged a close partnership with OpenAI, enabling the integration of ChatGPT capabilities into its Bing search engine. It also helps that Microsoft has invested billions into OpenAI. This development poses a potential challenge to Google’s search dominance, prompting the need for the Silicon Valley giant to innovate and introduce its own iteration of AI-powered search in order to stay competitive.
What are the changes to Google’s Search?
Since 2019, Google has been leveraging AI, including implementing the natural language processing (NLP) model BERT, to enhance its ranking of search results. In May, Google announced the Search Generative Experience as the latest innovation in Search. This development introduces their generative AI model into the user’s search experience, resulting in a more seamless and intuitive interaction with search functionality.
At the top of each page, Google will include an AI-powered summary, providing users with a preview of webpage content without needing to click the link. This helps to save time and presents relevant information upfront, which is great for users who just want quick answers.
Moreover, Google includes the sources it used to generate the summary, which is helpful in validating the accuracy of the information. This is in contrast to ChatGPT which may at times, “fabricate source names, direct quotations, citations, and other details” according to OpenAI. While ChatGPT could be useful in providing questions about general knowledge, it’s also not up to date — its database only includes information up to 2021, so Google may still prove to be more useful in providing the latest information.
Another unique feature is that the user to ask follow-up questions in a conversational manner, saving them the hassle of having to generate a new search journey. For example, a user could be searching for ‘cheap shoes near me for the summer’, and the next query they could enter is ‘how about shirts?’ and Google will use the same context to provide their search results.
With the incorporation of Google’s generative AI model to search, the algorithm gains improved comprehension of conversational queries, making it more accurate when users are employing voice searches. This functionality proves particularly beneficial for individuals with visual impairments or for those who require urgent information while driving.
This personalized approach helps Google generate answers tailored to a specific query. In the past, the algorithm can only serve content that has already been created, which may not be personalized to the searcher’s specific needs. However, the new AI-based search is able to generate new, personalized content based on existing information, for example, developing a training plan to run 10 km within a specific timeframe. This increases the likelihood of the searcher being able to achieve their desired outcome.
With these changes, Google stands to benefit as this interactive experience should result in users spending more time on the search page instead of immediately clicking on another website, similar to Facebook’s Instant Articles and TikTok Shop, all platforms designed to retain the user instead of redirecting them. The increased user engagement could allow Google to potentially earn greater advertising revenue by suggesting relevant products to the user’s search.
The problems with using SGE
However, with the advent of this new technology, there are some important considerations as well.
Addressing content ownership concerns will be a critical challenge for Google’s relationship with content publishers. Websites heavily dependent on search traffic may experience a decline in traffic volume directed from Google, which could affect their revenue if it is primarily based on advertisements.
This issue is crucial for Google to solve, as it needs to strike a balance between keeping users engaged on its platform and ensuring that content publishers are incentivized to continue creating content. As a search engine, Google does not generate its own content, and is dependent on websites to supply the information that it can index and deliver to internet users. This reliance is especially important for uncovering new information that is not widely known, as generative AI models can only create content based on existing knowledge, and lack the ability to generate entirely new content independently.
One possible solution that could be proposed is a revenue-sharing model with both the publisher and Google. The Australian government passed the News Media and Digital Platforms Mandatory Bargaining Code in 2021, which governs commercial relationships between news businesses and digital platforms like Google. This code ensures fair compensation for businesses whose content is showcased on these digital platforms’ websites.
A similar revenue-sharing model has already been implemented on YouTube, Google’s video-sharing platform, where a percentage of ad revenue will be shared with the content creator. While this will be harder to implement on websites, Google could consider exploring a similar model to ensure fair compensation for digital publishers.
These changes will inevitably impact the field of Search Engine Optimization (SEO). Content publishers, who traditionally employ various on-page and off-page strategies to rank higher on search engines, will now need to reassess their approach.
In order to optimize their content to be included in these new algorithm-generated summaries, publishers may have to adopt new strategies. Nevertheless, this could be a push for publishers to focus on generating high-quality content, as the emergence of ChatGPT and other language models has made it easier to create generic content that does not serve the user’s intent.
With these evolving changes, publishers need to adapt to them in order to maintain their relevance in search results.
Another problem with AI-powered summaries is the possibility of misinformation. Even though the model has gone through extensive testing, some of the provided answers may lack the necessary contextual nuances, failing to deliver information that is tailored to the user’s circumstances. This is most damaging for sites that cover topics that have the potential to affect a person’s quality of life, including finance and health.
Google already implements strict guidelines that allow results only from sites that have displayed EEEAT — experience, expertise, authoritativeness, and trustworthiness — for these topics. Only websites that have proven credibility will be able to rank on the search results for these topics.
Furthermore, Google’s Vice President of Google Search, Elizabeth Reid, spoke about this issue in a video interview with Bloomberg Technology. She mentioned that AI-powered search information won’t be displayed for certain topics, for example, the dosage of paracetamol for a three-year-old. As a result, AI-powered summaries may not be suitable for all types of queries, especially those that could have a significant impact on an individual’s life.
This leads to the last point on how overreliance on AI can potentially undermine our capacity for critical thinking. While AI can provide valuable insights, we need to balance them with our intuition to critically evaluate the information being displayed.
Google has acknowledged some limitations of its AI model, where it may misinterpret the information it uses to support its snapshot, which could alter the meaning of its summarized answer. By maintaining our critical thinking skills, we can navigate the digital landscape more effectively, make informed decisions, and approach search results with a balanced perspective.
While Google’s SGE seems to be a high-powered version of the once famous “I’m feeling lucky” button, its new AI functionality can enhance the search experience by making it more interactive and accessible. However, there are certain topics that go beyond the scope of its summarized content and require deeper critical analysis.
Despite the convenience of AI-generated summaries, we still need to rely on ourselves to assess the reliability and effectiveness of the information presented to us. This highlights the ongoing necessity for human engagement and critical thinking in the search process, ensuring that we approach search results with a balanced perspective and remain mindful of the limitations of AI technology.