LLM search overview and FAQ

How we search and find things is changing fast.

LLMs are initiating a shift from constrained, search box queries that return lists of links to natural language questions that return conversational answers. With this shift, we see AI agents further relieving human toil across tasks, leaving further space for our human-most thoughts and innovation.

These times are exciting and liberating, but we realize they’re also a bit scary and ambiguous, too.

At ércule, we’re rapidly researching and exploring all things web and LLM search so our clients can stay ahead of these shifts. This FAQ is meant to be a resource for any marketer who has questions around what’s happening across LLM search today and what to prepare for next. We hope it’s helpful!

Have a question that’s not answered here? Reach out and we’ll answer it here.

Am I losing traffic to AI overviews? What can I do?

If you were building search-optimized content prior to AI Overviews being rolled out, almost definitely you will see a decline in clicks for any top-of-funnel, informational content.

Google's AI Overviews provide comprehensive answers directly ahead of standard search results, reducing the need for users to click through to websites for informational queries.

Many reports indicate that clickthrough rates have decreased up to 30% due to the introduction of AI Overviews.

At ércule, advise clients to avoid very top of funnel, informational content that will likely be served by AI overviews—or AI mode—in the future.

Instead, we encourage teams to spend their strategic writing resources on content that brings forward their unique perspective and point-of-view and write only content that they’re uniquely qualified to write. This content can then be repurposed into various content formats through LLM-assisted workflows to maximize its value.

Last but not least, we also recommend closely tracking conversion rates and user journeys across your site to improve site experience and engagement for visitors. Often we see observe that a loss in traffic which we can roughly attribute to AI overviews improves conversion rate (with little-to-no change in net conversions), despite the loss in traffic.

What is Google’s AI mode, and is it the new way to search Google?

AI mode is Google’s latest innovation on their search experience. It’s an LLM-first search experience that feels more like what users are used to from ChatGPT’s web search functionality than the traditional Google search experience:

AI mode was introduced in March 2025 and is currently rolled out to all US search users as a Google Search Labs experiment. To access, it, you can go to google.com/aimode and follow the appropriate instructions to enable it for your personal Google account. (Note that AI mode is not currently available for Google Workspace users.) It’s not clear if this will become the dominant search experience, but we think some version of this is likely. In the meantime, there are some important elements to work out on the Google side (like how traffic will be attributed from AI mode, and where Google ads fit in.)

We’re closely following the rollout and response to AI mode and will keep clients and this FAQ updated we as learn and observe more.

Which LLMs currently use real-time web search vs only training data?

As of May 2025, every major public LLM provider has web search integrated in one way or another. There are important differences to know in how web search is triggered, depending on the LLM provider. For most LLM chat surfaces, if the LLM has sufficient training data for a query, it may not trigger a web search. Currently how this works depends on the provider, and the level of control and visibility is changing often.

As a way to “force” web search more reliably, you can prompt the LLM to “Search the web for latest info as of today’s date.”

• ChatGPT (GPT-4o): Utilizes live web search by default when training data is incomplete. You can also specifically trigger a web search using the Tools dropdown within ChatGPT by selecting search the web:
• Claude 4 (Opus and Sonnet): Offers web browsing capabilities but only searches web under specific conditions. Best to always request in prompt if recency is important.
• Google Gemini: Able to search the web, with no specific option to do so. Best to always request in the prompt if recency is important.
• Perplexity AI: Always searches the web with cited sources

Do LLMs credit sources when using our content?

Often - yes! ChatGPT, Claude, Gemini, and Perplexity often provide citations or links to original sources. However, the visibility and prominence of these citations vary. For instance, Google's AI Overviews may include links, but they are less prominent compared to traditional search results. AI mode uses a query fan out approach.

Importantly, it might work through 100+ pages to construct its answer, but only cite 5 pages. You can asssume those 5 pages might have the most direct answers to the most important questions related to the query, or at least the closest semantic connection.

The real kicker? These citations may not even be ranking first or second page for the terms they’re being found for. It’s a truly different world than traditional search\

How do we make sure our content is “indexed” by LLMs?

There is no “rank” for LLMs. Sources might be weighted differently by the model to determine things like authority, but whether or not you show up in a query is incredibly dynamic—even moreso than search—and also non-deterministic in nature.

That said, here are some things we recommend for optimal visibility:

• Identify blog post author and their role to establish authority
• Use authoritative sources in your copy and cite your sources
• Lead with active verbs and use short sentences so content is easy to parse
• Structure content like components that can function indepdenently (eg: a paragraph should be able to live alone without missing key details)
• Add FAQs at the end of post to capture potential queries a user might ask, that could land them on your content

Technical site concerns should also be reviewed, but are rarely an issue. We’re happy to help if you’re trying to sort through a confusing audit or need some guidance here≠

Can we influence what LLMs say about our brand?

Indirectly, yes. By ensuring that authoritative and up-to-date information about your brand is available across reputable sources—such as your website, Wikipedia, and industry publications—you can influence the data LLMs access and, consequently, the narratives they generate.

What should we communicate to stakeholders about LLMs' impact on our digital strategy?

The rise of AI-generated answers is reshaping how users find things. A strong content strategy to maintain and increase visibility and authority is more important than ever. Investments should be made in content optimization (for current, ranking articles), net new content, website messaging and updates, and comprehensive LLM and search visibility tracking.

Is traditional SEO still relevant for LLM visibility?

Yes, abolutely. The premise is the same, but the rules have changed a bit (for the better.)

SEO historically has helped brands go from unknown to known, across a variety of topics that they otherwise might not have been surfaced for—through content. Historically, domain authority has played a large role in whether or not a site has hope to rank for “competitive” terms.

LLMs care less about if you’re known or not, but whether or not you deserve to be in the conversation.

But get this - whether or not you deserve to be in the conversation is still largely deternined by SEO! To be fair, ”search” is still a very emerging word that one day will represent the common principles around how we find things. For now, we have different definitions across LLMs and search engines, but we see them getting closer by the day.

So what does this mean for your SEO strategy? Continue everything you’re doing, but also have a complimentary strategy for addressing the shortcomings of the #1 rule of SEO fight club: “Don’t talk about your brand.”

Turns out, you need to talk about your brand everywhere now. Your content people are probably the right people to help you figure out the placement and timing for these mentions, but for many teams this will look like a collective effort while more data on what’s most impactful by industry comes to light. Have a question that’s not answered here? Reach out and we’ll answer it here.