How Conversational AI Could Disrupt Search Advertising
Microsoft is reportedly planning to integrate OpenAI’s ChaptGPT with Bing, bringing an advanced conversational AI chatbot to a major search platform. Some believe it will push Google to integrate its AI Chatbot LaMDA with its search platform. The move could drastically disrupt the digital media landscape for publishers and advertisers if search platforms use chatbots to serve more information natively in results rather than providing a list of outbound links.
As search platforms integrate native chatbots, consumers will be conditioned to search using more conversational multi-word and question-oriented queries. Currently, the average keyword searched on Google is just 1.9 words or 8.5 characters, according to a Backlinko study. Further, only 14.1% of search queries today are in the form of a question and voice-based chatbot applications such as Alexa, which are more conducive to conversational search queries, only make up 1.2% of daily search volume according to our estimates. But over time, we anticipate that AI applications will condition users to search more frequently using conversational, multi-word search queries, doubling the total addressable market for AI chatbots between 2022 and 2025.
ChatGPT is a chatbot developed by the for-profit research lab OpenAI. It went viral on the internet last month, crossing 1 million users in five days, when people discovered that ChatGPT could summarize vast amounts of information in a natural and contextualized format. For example, users can ask ChatGPT to “summarize how combustion engines work,” and the application will return a thorough but easy-to-understand explanation. The same prompt asked to a traditional search engine could require clicking-through multiple search results to develop the same level of understanding about the topic.
Notably, ChatGPT doesn’t attribute links or sources in responses to generic questions. This has major implications for publishers who rely on traditional search engines for web traffic. Currently, 30% of CNN’s 600 million monthly website visitors come from search engines, according to SimilarWeb. Even social media platforms depend on search – 11% of Facebook’s web traffic comes from search. If Google and Microsoft use conversational AI chatbots natively in search, it could drastically reduce traffic to publishers - but doing so would also undermine the search advertising business.
Search is the largest digital marketing channel and will account for 41% of global digital ad spending this year, according to Insider Intelligence. Approximately 80 percent of Google’s revenue comes from advertising, and so it's in the company’s best interest to maintain good relations with publishers. Nonetheless, Google may be forced to disrupt itself before it becomes the one being disrupted by Microsoft, whose search engine Bing has doubled its share of the global search market from 4.5% in 2019 to 9% in 2022, according to Statcounter.
We’ve seen from other digital media shifts (e.g. print to digital) that in the struggle between maintaining traditional business models and consumers seeking convenience, consumers will get their way. Thus, Google and Microsoft will likely disrupt their own search advertising business models to try and strike a balance between consumer and advertiser preferences.
“I think what we’ll see is a shift to quality over quantity in search advertising,” according to one marketing executive at a large consumer goods company. “As an advertiser, we’re already aware that the vast majority of our search traffic comes from the first-page of search results. In the same way consumers prefer the path of least resistance to finding a solution to their need, so too are advertisers who prefer the path of least resistance to finding a customer for their solution. It’s a win-win if Google and Microsoft can figure out how to use this technology to help both consumers and advertisers.”
In some ways, the shift in advertising dollars from traditional search to retail media platforms like Amazon and Walmart has been a result of advertisers seeking that "path of least resistance" in using search products with the greatest consumer intent. Amazon also has its own conversational AI chatbot, Alexa, which powers its voice-assistant Echo devices. Similar to how ChatGPT uses a natural language processing model to serve users the single most relevant response to a query, Alexa uses a model to recommend the single most relevant product in a query. The general trend is that these tech giants will use AI to serve consumers fewer, more relevant information, and publishers and advertisers must adapt to stay relevant.
When prompting ChatGPT with the question, “which tv brand should I buy,” the application responds with a recommendation for Samsung, LG, Sony, and Vizio. The same prompt on Google returns a bevy of product pages and reviews that span a much broader range of brands – including Philips, TCL, and Panasonic – who presumably are going to great lengths to appear in those search results. So what can these brands do to optimize for rank and relevance in conversational AI applications?
Newer, more advanced AI applications like ChatGPT rely on multi-layer transformer networks that help the application self-train without manually labeled data – this means they are somewhat black boxes and there isn’t someone in a control room somewhere deciding how the application should respond to a query. Thus, the only way for brands to organically optimize for a future of AI-powered search is to make sure they have a strong digital presence that spans social, eCommerce, video, and other channels as these datasets will likely be used to train AI models. A brand will probably be more likely to be mentioned by an AI application if it can trace that brand to many instances online with a positive sentiment.
If you're an executive with an opinion about ChatGPT and the future of search, we want you to join our community. Apply now to become an Executive Partner at Hibben Group to join other leading corporate executives.