Rediscovering SEO in the Age of AI

Every planning season, someone declares SEO dead. The reality is simpler. People still look for answers. They now get more of them inside AI assistants, social feeds, and AI Overviews that appear at the top of Google. In March 2025, those Overviews showed for about 13 percent of U.S. desktop queries, up from 6.5 percent in January (Search Engine Land). This shift is real, and it is accelerating.

So, what is “AI SEO” in practice? At a basic level, it is the work of becoming the source that assistants prefer to summarize and cite. It is also the discipline of being present in the broader web conversations that shape an AI system’s memory.

Britney Muller puts a sharp point on it. Mentions without URLs relate to training data and popularity. URL citations in AI answers usually reflect what already ranks, because assistants often pull sources through search APIs. You are not replacing SEO. You are rediscovering it, then extending it to new discovery surfaces.

AI SEO vs. Traditional SEO

• Primary goal. Classic SEO aims to rank and drive clicks. AI SEO aims to be cited or mentioned to build authority, trust, and considered visits.

• User behavior. Classic search encourages scanning links. AI encourages reading a synthesized answer, then clicking with clearer intent.

• Query format. Classic search skews short and keyword based. AI queries are longer and conversational.

• Success metrics. Classic dashboards favor rank, traffic, and CTR. AI discovery rewards citation frequency, brand mentions, and share of voice in assistants.

A useful reality check: according to an Ahrefs study most pages cited in AI Overviews already rank in the top 10. Fundamentals are still the price of entry.

What’s Old Is New Again!

I have, for years, used a combination of SEO and Public Relations as a way to have maximum impact with my marketing campaigns. With LLM interactions, it’s time we lean more heavily into the traditional PR metrics for our content and online experiences.

Instead of clicks and sessions, you are optimizing your content for topic authority, presence, and recall. Third-party proof matters even more now, so reviews, analyst notes, standards bodies, and credible media citations will help assistants trust you. Queries are longer and more conversational. People ask full questions, then hop to Reddit, YouTube, and forums for validation.

Under the hood, a lot still rhymes with classic SEO. Helpful, expert content still wins. Technical clarity still matters. Distribution, PR, and ads still move the needle. The twist in what is familiar is where the value shows up. Being cited in an AI answer can deliver fewer clicks, yet more qualified visitors. In several studies, AI search traffic converts far better than traditional organic traffic. The graphs below come from a SEMRush study into AI and Search Traffic, and one of the takeaways from the study was that the average LLM visitor is worth 4.4x the average traditional organic search visitor.

There is an underlying assumption here that more people will adopt these features, but also as platforms make AI answers, or AI assistants, more of the default it’s easy to go along with this idea. We’ll have to watch user adoption of these features as more people become more familiar with AI-features.

Rethinking the KPIs

I am not fully convinced we should demote traffic across the board, but the pattern is hard to ignore. Some metrics are losing strength: organic sessions, pure rank, CTR, and bounce rate tell a smaller part of the story.

Other metrics look to be the things we should pay more attention to moving forward. Metrics like:

• Tracking topic share and brand strength.

• Tracking citations across reputable sources (trade publications, news media, governing bodies for your industry, etc.)

• Weighing reviews and creator coverage by quality, and not only by the stars of your industry.

  • SparkToro is a fantastic tool to find out which channels, creators, or shows your target audience engages with online.

• Monitor post-exposure actions, including brand searches, direct visits, demo requests, sign-ups, and sales chats. Consider display as scaffolding that firms up reach and helps you measure view-through lift on brand demand.

5-Steps for AI-friendly Content

1. Publish genuinely helpful and credible work. Add original research, real experience, and clear recommendations.

2. Structure pages for easy parsing. Use logical headings, concise paragraphs, lists, and tables. These help humans and help assistants extract answers.

3. Answer questions directly. Use question-based subheads. Provide the short answer first. Expand with context below.

4. Build brand presence beyond your site. Show up in reputable publications, creator channels, expert forums, and well-moderated communities. Assistants learn from the whole web, not only your domain.

5. Keep content fresh and accessible. Update statistics and examples. Confirm that important crawlers can access your site. If you want assistants to see your content, allow GPTBot in robots.txt.

   1. Learn more about OpenAI crawler bots

Bottom line

If your company has not been keeping up with SEO fundamentals with the content it creates, then start there. It’s still very early days with this “AI Transformation,” but what we’re seeing is that having a marketing mix that focuses on building your company, or your company’s thought leaders, as a trusted authority on particular topics still works and will largely continue to work.

Hopefully, this has been helpful, or at the very least sparked some ideas for you. Let me know if this resonates with you or if you have begun to experiment with these ideas.