What Marketers Need To Know

People prefer the easiest route. When they search for information, that now means they turn to generative AI tools, like Gemini and ChatGPT, or even AI Overview in Google Search.

These LLMs deliver neatly packaged answers. They don’t require the information seekers to scroll through a results page, click on a promising link, consume the site’s content to find the answer, and click the back arrow if they don’t find what they need.

That’s why marketers must adjust their search strategies to keep attracting online audiences to their brand and its content. To optimize your content for the search revolution, read on for recent research and advice from a few of CMI’s go-to experts in the field of search, analytics, and AI.

AI isn’t an alternative to Google; it will be the new Google

Even the search behemoth itself — Google — is immersed in the AI-answer world in ways that create a “serious situation for SEO,” writes Search Engine Journal author Roger Montti.

“In 2025, AI is not just AI Overviews. AI is Gemini, its new features coming to Gemini and possibly the release of features developed from Project Astra, a multimodal universal agent … All of these are a departure from the traditional search box as the point of contact between users, Google, and websites,” Roger explains.

Will people really move from the familiar search box to Gemini? Google’s recent statements and products indicate that the company is betting on it.

Google’s AI lab co-founder says the company will “turbo charge” the Gemini app in the next two years. Google’s universal assistant — Project Astra — will converse with users audibly and visually through their phones and prototype glasses and give answers, recommendations, and more information based on those chats.

And that’s just what’s happening at Google.

Microsoft has integrated its Bing search engine with its AI tool, CoPilot. ChatGPT, which is already a slang term for generative AI, has released SearchGPT, which draws from Bing.

And given the growing volume of players in the diverse generative AI game, expect ongoing disruption in this revolution.

Don’t panic; learn the game board

Revolutions aren’t unique. If you’re a seasoned professional, you know that before Google was THE search engine, there was Alta Vista, Ask Jeeves, MSN Search, Yahoo!, and more. Each had its own nuances, from algorithm preferences to how it delivered the results.

Answer-focused AI may be in a similar stage. So, now’s the time to understand what each tool cares about the most when generating responses and decide where your SEO team should invest more of its time.

Will Seer and Jason Stinnett at Seer Interactive recently broke down the three types of AI search:

• Training-first systems such as Claude and Llama “generate responses based solely on their training data — snapshots frozen in time until the next model update.”
• Search-first systems such as Google AI Overviews and Perplexity “rely heavily on real-time indexing of web pages … It’s like SEO on steroids, with a dash of content curation.”
• Hybrid systems such as Gemini and ChatGPT “blur the lines by deciding, often in real time, whether to answer based on their training data or fetch updated information from the web.”

Discover your audience’s preferences

Now that you know the potential systems, how do you know what your audience prefers? You should do what every good marketer does: Look at your web analytics.

To do a quick check, use a simple GA4 filter as explained by Orbit Media’s Andy Crestodina:

• Go to the Engagement > Page path and query string report.
• Set a long date range.
• Click the “Add filter +” button at the top of the report.
• In the filter settings on the right, select “Page Referrer” as the dimension. Select “exactly matches” as the match type. Enter “https://chatgpt.com” as the value.
• Click “Apply”.

Then, you can see the traffic coming from ChatGPT. Apply the same process but change the values for the “Page Referrer” to other domains, such as https://gemini.google.com and https://perplexity.ai.

Given the newness of AI as an alternative search method, don’t expect a lot of traffic to come from any of the sites, Andy says. As that changes, you’ll already be monitoring the impact to better understand what answer AI tools your audience prefers.

For example, Orbit Media’s site saw AI referrers made up 0.2% of its total traffic in the previous 30 days, but that’s twice as much as the previous month, Andy explains.

By checking the numbers regularly, you also can see how (if) your traditional search traffic declines month over month.

Want to spend more time setting up this monitoring permanently or creating a custom channel group? Orbit Media has step-by-step instructions on how to do both.

Ramp up your AI referral engines

Of course, just as you don’t sit and hope for search engine traffic, you shouldn’t wait for AI sites to deliver traffic. Investigate the emerging trends and know how your brand appears (or doesn’t appear) in the current systems.

Chris Long of Go Fish Digital did the latter. The agency conducted a ChatGPT Search test to see how it performed for “digital PR agency,” a query that does well for Go Fish Digital on Google Search.

It found ChatGPT Search listed Go Fish Digital, but it didn’t include the agency’s notable clients as it did for some other firms mentioned in the results.

To remedy that, the Go Fish team revisited the source page most frequently cited for its ChatGPT Search mentions. It added a “notable clients” section to the content in a bulleted list, a format the search engine LLM would be more likely to appreciate.

A week later, ChatGPT Search added Go Fish’s notable clients to its results mention.

Recognize the big picture for brand mentions

Seer Interactive took a wider view with its proprietary tool to investigate factors affecting brand mentions in AI-generated responses.

It took 300,000 keywords related to the finance and SaaS industries and got their traditional search engine rankings and other related data, such as backlinks, format, etc. It also selected 10,000 questions that would mostly trigger answers that mentioned a brand from the People Also Ask results. It ran those questions through OpenAI’s GPT4o API and counted the brand mentions in the answers.

The results led to these conclusions:

• Brands ranking on page one of Google strongly correlated with their mentions in the AI answers. Bing rankings mattered but slightly less.
• The quantity of backlinks had a weak or even neutral impact.
• Content format variety (videos, images, etc.) didn’t have much of an impact.

In a different test, Seer Interactive dug into the importance of keyword variations to generative AI tools — something that traditional search rewards.

Seer asked 1,000 versions of this question: “Who are the best brands for composite decking?”

The 1,000 answers mentioned a total of 74 unique brands, with an average of six brands included in each answer.

Seer Interactive’s Christina Blake says the results indicate that AI tools prioritize semantic intent over specific words. Thus, marketers should focus on the themes and concepts behind the questions, not minor variations in phrasing as they might do with traditional search.

Do the hard work for the easy search route

In the coming year, more and more people in your target audience will change their search behavior and opt for the easier route.

To ensure that your content and brand show up in those nicely wrapped packages, update your marketing playbook. Don’t abandon traditional search rankings, given their potential role in AI-generated answers. But do spend less time perfecting the keywords and other minutia that doesn’t really matter to your audience or the generative AI tools.

All tools mentioned in this article were suggested by a contributor to the article. If you’d like to suggest a tool, share the article on social media with a comment.

HANDPICKED RELATED CONTENT:

Cover image by Joseph Kalinowski/Content Marketing Institute