The search landscape is undergoing its most significant shift since Google launched. With AI Overviews, AI Mode, and other generative platforms changing how people discover information, the rules of visibility have been rewritten. Click-through rates are dropping, zero-click searches are climbing, and personalization is reshaping every journey.
In our recent Winning AI Search webinar, Jeff Coyle (Chief Strategy Officer, Siteimprove) and Garrett Sussman (Director of Marketing, iPullRank) broke down what this means for marketers. Watch Winning AI Search to hear the complete conversation or keep reading for the distilled playbook (you can also download the AI Search Survival Guide for structured checklists, frameworks, and action steps).
Not all AI outputs are equal
AI-driven search isn’t a monolith. Two experiences dominate Google today:
- AI Overviews: Short, synthesized responses that sit above the blue links. They can include text, bullets, breakouts, or even e-commerce grids.
- AI Mode: A more conversational, assistant-like interface that decomposes queries into multiple sub-queries (“fan-out”), retrieves content at the passage level, and personalizes results.
Jeff emphasized that site-level authority now matters more than page-level optimization. Every page contributes to, or detracts from, whether your brand is cited.
Garrett added that this fan-out process can pull passages from pages that don’t even rank in the top 100, underscoring why coverage and interconnection across a site are essential.
“Infinite mediocre content is easy to generate. Infinite authoritative content is impossible without strategy, editorial integrity, and proof of trust.” — Jeff Coyle
The end of SEO loopholes
The days of keyword tricks, thin link schemes, and exact-match domains are over. AI search engines evaluate topical authority, semantic coverage, and editorial trust — what iPullRank has termed Relevance Engineering. The main shifts are:
- From keywords to topics: Search models think in entities and clusters, not individual phrases.
- From hero pages to full sites: Even low-traffic content now influences your authority footprint.
- From shortcuts to integrity: Generative systems reward well-structured knowledge, not content farms.
“It’s not about loopholes anymore; it’s about your whole corpus. If you’ve got off-topic or weak pages, they act like sandbags on your visibility.” — Garrett Sussman
Behavioral SEO: memory and personalization
AI search introduces a new dimension: behavior. AI Mode personalizes answers through embeddings, meaning two users searching the same phrase can get different outputs. Future systems, like Google’s Project Astra, will even carry memory across sessions, shaping results based on past queries.
Jeff framed this as a turning point:
“Search is no longer about the click; it’s about being remembered. The AI will carry your authority forward, or it won’t carry you at all.” — Jeff Coyle
Garrett added the enterprise lens, proving ROI gets harder when traffic drops but conversions rise:
“Fewer visits doesn’t necessarily mean less impact,” he noted. “The ones who do click through may be far more qualified.”
Practical tactics for the new search
So how should marketers adapt?
- Design for conversations, not just queries. Anticipate follow-ups and pivots.
- Structure for passage-level retrieval. Break content into coherent, extractable chunks.
- Prioritize human-in-the-loop editing. AI can accelerate outlines and drafts, but editorial rigor ensures authority.
- Prune and refresh. Remove or update weak pages that dilute topical clusters. An makes it easier to identify underperforming pages, monitor keyword coverage, and keep your content aligned with AI-driven search expectations.
- Diversify formats. Use multimodal assets (text, video, and structured data) to increase inclusion in query fan-outs.
Garrett emphasized content hub strategies and multimodal planning:
“Don’t just think about one product page ranking. Map the entire journey, from awareness to decision, with interlinked, multimodal content.” — Garrett Sussman
The enterprise playbook for Relevance Engineering
At scale, SEO becomes less about hacks and more about orchestration. Together, Jeff and Garrett underscored the importance of Relevance Engineering through:
- Semantic precision: Using triples (phrasing sentences as subject–predicate–object) and embeddings to structure knowledge.
- Cluster orchestration: Aligning editorial, SEO, and product teams around unified topical coverage.
- Scientific measurement: Tracking citation share, coverage ratios, and inclusion in AI memories, not just rankings or CTR.
“If you’re still splitting SEO, brand, and performance into silos, you’re already behind.” — Garrett Sussman
Risks and red lines
AI search is still immature, and yes, it can be manipulated. Some brands are tempted by gray-hat tactics like semantic mimicry or seeding “chunk bait.” Jeff’s warning: these shortcuts collapse quickly.
The only sustainable play is to build authoritative, extractable, trustworthy content.
“The winners aren’t those who generate the most content. They’re the ones who orchestrate content with precision, editorial integrity, and conversational foresight.” — Jeff Coyle
What to do next
The AI search era rewards precision, authority, and orchestration. Jeff and Garrett both recommended five starting moves:
- Audit your pages with a claim registry. What facts do you deserve to be cited for?
- Map fan-outs for critical journeys and ensure deep coverage.
- Monitor your embedding proximity to authoritative neighbors.
- Run corpus hygiene sprints to prune, refresh, and interlink.
- Shift leadership conversations from rankings to citations.
This joint perspective from and makes one thing clear: AI search is less about chasing algorithms and more about engineering trust. Brands that adapt now will own visibility when search becomes fully conversational.
Want the full framework?
Watch the webinar replay for the complete conversation with Jeff and Garrett, or download the AI Search Survival Guide for structured checklists, frameworks, and action steps.
