Can AI Search Engine Overtake Google in Search Marketing?

If we were to select the crown jewel of the internet industry, search engines would undoubtedly be one of the strongest contenders. In both eastern and western worlds, Baidu and Google have almost occupied the iron throne respectively for over two decades due to their search engines.

However, with the rise of generative artificial intelligence (GenAI), the search engine arena has been reignited with competition. More and more companies believe that with the power of GenAI, overturning Baidu and Google’s iron throne is a very possible course.

With major players like Microsoft, OpenAI, Alibaba, and ByteDance entering the field, Perplexity, the Silicon Valley AI startup that pioneered in AI search, is facing tougher times ahead. To avoid falling behind in this AI search competition, Perplexity took the lead in commercialisation of its AI search engine. However, compared to traditional search engines, marketers seem to still in the observation stage about AI search.

AI Search Engines vs. Traditional Search Engines?

To understand the differences and advantages of AI-powered search engine compared to traditional search engine, here is some high level summary:

The core of AI search engine is understanding user queries

AI search engines, such as Perplexity and Microsoft Bing, utilize natural language processing (NLP) and machine learning (ML) to better understand user queries. Unlike traditional search engines that rely on keyword matching, AI search engines can interpret the context and intent behind the user’s search, providing more accurate and relevant results.

Interpretation of users intent: direct answers vs. link lists

Traditional search engines typically provide a list of links that users must sift through to find the relevant information. In contrast, AI search engines like Perplexity and Microsoft Bing can directly provide answers to the user’s query, summarizing the information from various sources and presenting it in a concise manner. This saves users time and effort.

AI search engines provide users the interactive search experience

AI search engines offer an interactive experience, allowing users to engage in a conversation-like interaction. Users can ask follow-up questions, and the AI will continue to provide detailed answers based on the previous context. This is in stark contrast to traditional search engines, where each new query is treated independently and without memory of previous interactions.

Integration of multiple technologies

AI search engines combine traditional search indexing with the reasoning and text transformation capabilities of large language models. For example, Perplexity uses OpenAI’s GPT-3 model to read and summarize content from multiple links, ensuring that the answers provided are supported by relevant references and links.

Privacy and Ad-free experience

Many AI search engines, such as Komo, You.com, and Perplexity, emphasize user privacy and offer ad-free experiences. This is a significant departure from traditional search engines, which often rely heavily on advertising revenue and may compromise user privacy to generate profits.

Content Creator Compensation

Some AI search engines, like Yep, introduce new business models that reward content creators directly. Yep uses a 90/10 revenue split, where 90% of the ad revenue goes to the content creators, providing a more equitable distribution of income.¹

Enhanced Search Efficiency

AI search engines are designed to optimize search efficiency. For instance, Waldo allows users to customize their search interface, reducing the need to open multiple tabs and making the search process more streamlined. Similarly, Andi’s interface is designed to save users time by providing a clean, ad-free reading experience.

In summary, AI search engines offer a more personalized, efficient, and interactive search experience by leveraging advanced NLP and ML technologies, providing direct answers, and often prioritizing user privacy and content creator compensation.

An emerging AI Powered Search marketing model

To move forward with its commercialisation process, Perplexity recently announced a new search marketing model that completely reforms Google’s original ad auction system, aiming to provide advertisers with a way to bid for question-and-answers targeting specific queries.

Perplexity’s new AI Search Marketing model is tailor made for AI search engine. Traditional search engines like Google, Baidu, Sogou, DuckDuckGo, and Yandex use crawlers to fetch web pages from the internet, then save them in their own databases and create indexes. When a user initiates a search, the system retrieves records matching the user’s query conditions and returns the results to the user in a certain order.

But the results were recommended to users without truly understanding their needs. Therefore, the results provided are essentially reference materials, and whether the problem can ultimately be solved depends on the user’s ability to extract effective information from these references.

However, AI searches represented by Perplexity are question-and-answer based. They combine the content indexing technology of traditional search engines with the reasoning ability and text conversion capability of large language models.

The charm of AI search is not only reflected in finding results that are more relevant to user queries but goes a step further by using AI to think on behalf of users, presenting answers directly. This question-and-answer mode of AI search means that the paid ranking of traditional search engines is no longer applicable, as inserting advertisements directly into answers would greatly affect the user experience. This is why Perplexity’s previous attempts to sell ad spaces to advertisers have been largely ineffective.

Google and Baidu use paid rankings, resulting in advertisements appearing as top links, but users can choose to skip them, and browse the search results they need while filtering out ads. If AI search were to incorporate ad placements, the generated answers to user’s question would become sponsored content. So while Google’s paid ranking system is indeed highly efficient in terms of monetization, bringing in hundreds of billions of dollars in revenue annually for Google, it is incompatible with AI search.

The inability to directly sell ad placements in the answers generated by AI search engines is the root cause of Perplexity’s current proposal for advertisers to bid on sponsored questions. While this model can serve the needs of advertisers, this strategy also meets the demands of content creators such as news media and publishing institutions. After all, Perplexity’s working model involves submitting user questions to large language models, then searching for existing resources on the internet, allowing AI to compose answers based on the information found.

So what the new AI-powered search marketing model looks like?

Perplexity’s recently introduced advertising model integrates ads into its search results in several key ways:

This approach blends traditional ad revenue models with the advanced capabilities of AI, offering a new and interactive way for brands to engage with users.

What else should we know about AI search marketing?

According to Caitlin Halpert, Vice President of marketing agency Journey Further, “I see the advertising seamlessly integrating with how the platform operates today.” Perplexity’s move can integrate advertising seamlessly without sacrificing user experience because, with the help of AI, Perplexity can replace users’ questions imperceptibly.

The specific operational mode of Perplexity’s bidding for sponsored questions should be similar to the keyword association technology of traditional search engines. It automatically recommends related keywords by analyzing the keywords in the questions asked by users. For example, if a user’s question includes keywords like “greeting card” and “price,” Perplexity can then show related questions and answers sponsored by greeting card manufacturers, enabling users to see more relevant answers promoted by businesses.

Compared to the straightforward approach of paid rankings, sponsored questions might be subtle and imperceptible, greatly reducing people’s wariness towards search results. Needless to say that when answers are smart enough to interpret user’s true intent, the user is more likely to read around and dig out more useful and relevant information, i.e. potentially more chances for businesses to influence their potential customers.

Perplexity’s claim to reshape Google’s ad auction system may not just be an exaggeration. This new AI powered search marketing model of matching users’ questions to advertisers sponsored questions and answers has the potential to reform how search marketing worked for decades.