Generative AI tools powered by Large Language Models (LLMs) like ChatGPT (which powers Microsoft Bing), Google Bard, and many emerging platforms are rapidly changing the way we search for information online. These tools are designed to generate human-like responses to user queries, providing answers that are often more nuanced and contextually relevant than those served through traditional search engines. This begs the question, how will generative AI tools impact search? As these tools continue to improve, we can expect them to become an integral part of the search process, helping users find the information they need more efficiently and effectively.
Advantages of LLM for Search
One of the key advantages of LLM chat tools is their ability to understand the nuances of language and context. This allows them to generate responses that are not only accurate but also provide a more complete understanding of the topic at hand. Additionally, these tools are often able to learn and adapt to new information, improving their accuracy and relevance over time.
Here’s an example:
A coworker and I were hoping to find the perfect, post-work happy hour within a certain geographic zone. I tried searching “Best St. Louis Happy Hours,” but many of the top hits were blog posts, often outdated and not in the specific area of town we needed. Enter: Bing with ChatGPT. First, I told the tool that we wanted to be somewhere between the towns of Town and Country (where the Perficient HQ is located) and Clayton and that we wanted happy hour menus.
Bing was able to crawl the web and provide 5 top choices. One choice looked particularly interesting, so I asked Bing what was on the happy hour menu. Upon hearing $6 wine, we were sold! In under 2 minutes, Bing found the best option quickly and easily. However, I never once visited the restaurant’s website. This leads us to the big question: will AI impact web traffic and SEO? And is this the end of traditional search as we know it?
Initial Adoption of AI Chat in Search
LLMs like ChatGPT have the power to be transformative, as we’ve already begun to see in Microsoft Bing. Before the launch of Bing with ChatGPT on February 7th, the site averaged between 40-50 million daily visits. Now, the site is exceeding 100 million visits with over 1 million sign ups in the first 48 hours of the technology’s announcement. And there is still huge upside as adoption continues, since many tools are still in beta phases.
Though this groundbreaking technology is unlike anything we’ve seen before, it is still driven by requests and questions, also known as search. While Bing ultimately helped me pick a happy hour spot, I was not able to see what the restaurant looked like or other details available via traditional search. Until chat tools can integrate images, news, or videos and be reliably monetized, search engines will have a strong built-in incentive to keep “traditional” search front and center.
And while many users are excited by AI possibilities, we have been trained how to search for over two decades and those habits won’t change overnight. However, many marketers find themselves wondering how this will impact web traffic and SEO. How will we be able to optimize websites now that consumers can find the answers they need without ever visiting the source site?
Optimizing for AI
Though technology and search behaviors are evolving, SEO fundamentals are still valid.
Google cares less about if content is written by people, as long as that content is written for people (and not LLMs). Providing comprehensive answers to user queries still works, especially focusing on long-tail keywords. Additionally, subject matter experts (SMEs) are still more valuable than content farms or AI-generated content. EEAT (experience, expertise, authority, and trustworthiness) will continue to be prioritized by Google, and the expertise that a SME brings to the table may help your content rank higher than AI-generated content.
One way to optimize your pages for LLM generative AI tools is to structure your pages in a way that is easy for search engines to understand. Ensure your pages have clear headings and subheadings, schema markup, and clear and concise points that answer the questions users are searching for. If this feels like déjà vu, this is the same strategy for optimizing for not only search, but also voice search, snippets, as well as new LLM chat engines.
Consider implementing AI-driven chatbots and virtual assistants into your own websites. You can now leverage the power of ChatGPT in a conversational interfaces to answer questions about your website and cut down on call center volume. Providing users with a more engaging experience on your site will give them more reason to visit your site vs. finding answers via Bing with ChatGPT, for instance.
LLM chat tools have the potential to revolutionize the way we search for information, enabling us to find answers to even the most complex questions with ease. By understanding natural language queries and providing more relevant results, these models are helping users find information more quickly. As LLMs continue to evolve, we can expect to see even more changes in the way that search engines and users interact with the web.