New Era of SEO: Brands Adapt to AI Search Engines

Advertising agencies and technology startups are racing to help brands increase visibility in the search results of artificial intelligence chatbots, opening a new chapter in the development of search engine optimization.

Companies like Profound and Brandtech have developed software solutions that track how often specific brands appear in the results of OpenAI’s ChatGPT, Claude from Anthropic, and Google’s new feature Overviews.

Brands such as fintech company Ramp, job search portal Indeed, and Scottish whisky producer Chivas Brothers (owned by Pernod Ricard) are already using such tools, hoping to reach millions of users who increasingly use generative AI tools as a new method of searching for information online. This is a new trend that poses a serious threat to Google’s core revenues.

New Tools for a New Era

– This is not just about indexing your website in search results. This is the recognition of large language models as the ultimate influencers – said Jack Smyth, a partner at the marketing technology group Brandtech, which has developed its own interface for brands.

New tools can predict the sentiment of AI models towards certain companies by sending a series of text queries to chatbots and analyzing the results. Based on the collected data, brand rankings are created, and agencies then advise clients on how to increase the likelihood that models will mention them in responses.

These moves come at a time when advertisers are feeling increasing pressure due to the expanding application of artificial intelligence in creating and targeting marketing campaigns. Meta and Google are developing ‘self-service’ tools that allow brands to run campaigns independently, which poses a threat to the traditional business of agencies and media buying intermediaries.

According to a study by consulting firm Bain, as many as 80 percent of consumers rely on AI-generated responses in at least 40 percent of their searches, resulting in a decline in organic web traffic of up to 25 percent. Additionally, 60 percent of searches now end without clicking on any additional links.

Google’s Response to Challenges

Despite this, Alphabet, Google’s parent company, announced last week that its revenues from core search and advertising activities rose by nearly 10 percent in the first quarter, reaching $50.7 billion.

These results calmed investors concerned about the rising popularity of competing AI chatbots like Musk’s Grok, as well as the possibility that Google’s own Gemini chatbot and AI response summaries could threaten the business model by reducing the number of clicks on ads.

Brandtech has developed a product called ‘Share of Model’, which provides clients with insights into analyses and advice on adapting text and visual content on websites to better meet the demands of AI search.

The startup Profound, which raised $3.5 million in seed funding led by Khosla Ventures in August, offers a data analytics platform that allows brands to track the most common queries in their industry and measure their own success in AI searches.

– Traditional search has been one of the largest monopolies in the history of the internet. And now, for the first time, it seems that the walls of that castle are crumbling. This is a moment of transition from CDs to streaming, said James Cadwallader, co-founder of Profound.

New software requires a deep understanding of how specific models represent brands. ChatGPT, for example, uses traditional web search and then evaluates information sources based on relevance to the user, including an assessment of the credibility and authority of the sources.

Adam Fry, head of search for ChatGPT at OpenAI, points out that users today are asking increasingly specific and precise questions, such as: ‘Can you find a quiet restaurant for a family of five in New York?’, instead of simply ‘restaurants in New York’.

The Future of Optimization

At the same time, Perplexity, an AI search engine, is testing sponsored ‘questions’ that are offered to users as suggestions for continuing their search after the initial query.

– Large language models understand more content and can more thoroughly identify contradictions or misinformation. The process is significantly more thorough than merely browsing links. Therefore, it is much harder to manipulate results, and the only real strategy for brands becomes creating the most relevant and high-quality content – said Denis Yarats, co-founder of Perplexity.