AI has changed the world of search as we knew it. As the technology paves the way for further major updates to how we search, what are the ramifications for the ad tech industry?
A shapeshifting search landscape
AI-powered searching has opened up a whole new world over the past couple of years. With the launch of AI chatbots such as OpenAI’s ChatGPT, the capabilities of search have taken on a new face. We’ve seen an AI race commence among big tech, with Google and OpenAI leading the competition to drive innovation in the sector.
Google’s dominance in the ad tech ecosystem – particularly in search over the past couple of decades – is pretty undisputed. Having thwarted competitors left and right, Google became the target of antitrust cases which accused it of abusing its dominance as a search engine. Allegedly, it pays agreements of more than USD $10bn (£7.86bn) per year to maintain dominance by ensuring it’s the default search engine on mobile phones and desktops.
Now, Google has introduced a revamped, AI-enhanced version of Google Search with the launch of AI Overviews. Although not yet available globally, the feature provides users with AI-generated answers which appear at the top of the search results. Unlike its previous version of Search, the update is able to process and answer increasingly complex and particular questions with nuances. Google said it expects the feature to be available to over a billion users by the end of the year. It hasn’t all been smooth sailing though – since, Google has had to reduce the scope of the feature. It had trouble with some of its responses which ended up being widely mocked on social media. Users were advised to eat one rock per day, and use non-toxic glue to better stick the cheese to their pizza, among other amusing suggestions.
The release of Google’s revamped search engine was almost definitely spurred on by increasing competition from rival OpenAI, whose ChatGPT has gained monumental traction. The AI chatbot arrived on the scene in November 2022, immediately gaining huge popularity. By January 2023, ChatGPT had set a record for its fast-growing user base: within only two months of its launch, it had amassed an estimated 100 million monthly active users. In effect, this made it the fastest growing consumer application in history, even overtaking the initial pickup of social media platforms such as TikTok. Fastforwarding to this year – OpenAI’s chatbot received 1.8 billion monthly visits in April 2024. Now, ChatGPT also powers many Bing products including its Search and Copilot AI. Several other competitors have appeared on the horizon too, such as Perplexity AI which has also gained major traction.
What does the introduction of AI Overviews mean for publishers?
Publishers have been in an increasingly precarious position lately. A recent study by the Reuters Institute for the Study of Journalism found that 63% of survey respondents claimed to be worried about a sharp decline in referral traffic from social media sites. Just a couple of months ago, Facebook shut down its news tab for users in Australia and the US, with Meta aiming to reduce news content available on its services. When we look at the rapidly changing nature of search, it makes sense that publishers are tentative about the changes.
Validly, many publishers are concerned about whether their referral traffic will be negatively impacted by Google’s new search, which would consequently affect their ad revenue. The Guardian, for example, got around 30% of its visitors from Google Search in April – plus 5.2% if we include Google News. With just under half of traffic coming from Direct (via a URL taking the user straight to the site), Google is clearly responsible for an enormous portion of traffic. When it comes to smaller publishers, many are even more reliant on Google Search for their traffic. So far, Google has failed to reveal data addressing these concerns, raising further worries about transparency.
Google itself has suggested that its AI Overviews could actually help increase traffic for publishers – in its blog, Google stated: “the links included in AI Overviews get more clicks than if the page had appeared as a traditional web listing for that query…And when people click to links from AI Overviews, these clicks are higher quality, where users are more likely to spend more time on the site.” So far the sentiment has been mixed, but many have expressed predictions about how this could negatively impact publishers. It’s currently unclear to what extent they could be affected.
What will happen to ad spend?
A week after launching AI Overviews, Google announced it would start testing Search and Shopping ads in the feature. It said that advertisers would have the opportunity to appear within the AI Overview in a section clearly labelled as sponsored, if and when their content is relevant to the query searched and the information produced by the overview. Google’s blog explains: “There’s no action needed from advertisers: ads from existing Search, Performance Max, and Standard Shopping campaigns have the opportunity to appear within the AI Overview.”
At its Marketing Live event, Google explained that AI Overviews would show interactive shopping ads that would be prioritised over organic search results. This can be seen as a new opportunity for advertisers – however, this also raises other questions and potential concerns industry-wide.
Sustainable searching: Are we regressing?
As urgency surrounding climate change becomes more pronounced, the ad tech industry has been increasingly adopting greener practices and tech. But with the use of AI skyrocketing, will the industry’s green efforts go to waste? Although AI has been hailed by some as a potential solution to the climate crisis, there are certainly problems with this idea. Increased efficiency from AI is a benefit, but that doesn’t appear to be offset by the use of energy and resources needed to power it. With the invention of more powerful AI models, comes a drastically higher energy consumption.
Firstly, a significant amount of energy is required to create and train AI models. Researchers have estimated that to create GPT-3, 1,287 megawatt hours of electricity were consumed and 552 tonnes of carbon dioxide equivalent were generated – this may not be so bad, but this only takes into account its creation. And once up and running: adding generative AI to Google Search alone increases its energy uses by over tenfold.
Other ongoing environmental costs related to AI usage also intensify – with water consumption, for example. Google’s 2023 environmental report showed a 20% jump in its water consumption in its data centres, amounting its yearly total for 2022 to 5.6 billion gallons. This can be attributed to Google’s increased computing capacity which has largely been driven by AI in recent years. (A high volume of water is used at data centres to cool the systems.) Looking ahead to the next couple of years, global AI could account for up to 6.6 billion cubic metres of water withdrawal in 2027 – which is roughly the total annual water withdrawal of half of the UK. With Google’s latest Search updates, its consumption of water and energy is sure to rise further.
Social searching
Considering the possibilities for the future of search wouldn't be complete without touching on the increasing use of social media platforms as search engines. TikTok particularly has become a search engine for many. Research commissioned by Forbes Advisor found that when searching, Gen Z turn to Google 25% less than Gen X. The study also demonstrated a 30% decline in search engine use for brand discovery among Gen Z compared to Baby Boomers. Around a quarter of respondents even said they primarily use social media to search online. This makes some sense when taking into account the younger generations’ increasing use of mobile devices versus desktop. Interestingly, TikTok appears to be implementing material generated by ChatGPT into its search feature.
Looking to the future: What’s next for search?
From searching on TikTok to messaging chatbots, people’s searching habits have been evolving significantly. As the industry advances, social will be an interesting space to watch. Although AI chatbots have gained enormous popularity, search engines are used for many reasons, some of which a chatbot like ChatGPT won’t provide the best results – for the moment, at least. Currently, one of the most important functions of a search engine is simply to direct users to a particular website – a process which has no need to involve AI.
In our weekly poll a couple of weeks ago, we asked ExchangeWire readers whether they expect ChatGPT to someday overtake Google Search’s popularity globally, or whether Google’s AI-enhanced updates could keep it in the lead. The results revealed that most respondents (72%) banked on ChatGPT overtaking Google Search.
With developments in AI occurring at such a rapid pace, it’s impossible to know exactly where the industry will be in even just a few years. ChatGPT has certainly changed the outlook of ad tech, but as it stands, Google may well maintain its dominance in the search engine landscape for the foreseeable future. As AI-enhanced searching becomes increasingly popular among the general public, the industry must hold itself to higher environmental standards, and look where possible to greener energy sources.