Google I/O made it official: AI-generated answers now sit at the top of search results, and most brands have almost no idea how those answers describe them to customers. That’s the takeaway from a recent episode of TechCrunch’s Equity podcast, where Rebecca Bellan spoke with Matt Thompson, VP of partnerships at Scrunch, a startup building tools for the AI search shift. If you’ve spent years optimizing for ten blue links, your strategy is now aimed at a search engine that’s fading out.
This is significant because the change isn’t cosmetic. The interface people use to find information is being rebuilt, and the optimization tactics that worked for two decades don’t map cleanly onto a world where an AI summarizes the answer before anyone clicks.
What’s actually changing
Three shifts stand out from the conversation, according to TechCrunch AI’s reporting.
- AI referrals convert far better. Thompson points to AI-driven referrals converting at roughly 400% higher than traditional organic search. Fewer visitors, but each one arrives with more intent. The person already got a recommendation from the AI, so they show up closer to a decision.
- Google isn’t the whole game anymore. ChatGPT still holds the lion’s share of AI search traffic. Optimize only for Google and you miss most of where the AI conversations are actually happening.
- Google’s own advice may mislead you. Thompson argues that Google’s published SEO best practices can steer marketers in the wrong direction for this new environment. The company optimizing its own search product isn’t necessarily optimizing for yours.
Why it matters now
The old model rewarded volume: rank for keywords, capture clicks, fill the funnel at the top. The new model rewards being the source an AI trusts enough to cite. Those are different problems.
When an AI answers a question, it picks which brands to mention and how to describe them. Most companies have zero visibility into that process. You can’t see what ChatGPT or Google’s AI Overviews say about you the way you can check your keyword rankings. That blind spot is the real story here.
Thompson uses the phrase “agent ready” to describe sites built for this shift, and notes most enterprise sites aren’t there yet. A site designed for human skimming and Google’s crawler isn’t necessarily structured for an AI agent to parse, understand, and quote accurately.
The Future Cast: where this goes
Over the next one to three years, expect a few things to harden into standard practice:
- A new category of monitoring tools. Just as brands track keyword rankings today, they’ll track how AI models describe them. Startups like Scrunch are betting this becomes a budget line, not a curiosity.
- Content built for machines first. Clean structure, clear factual statements, and machine-readable formatting will matter more than keyword density. If an AI can’t cleanly extract your value proposition, it’ll paraphrase a competitor’s instead.
- A platform split. Optimizing for ChatGPT, Google’s AI Overviews, Perplexity, and others won’t be identical work. Marketers will need to think across multiple AI surfaces, not one search box.
A fair caution: Thompson sells a product in this space, so the 400% figure and the urgency come from someone with a clear interest in the shift being big. Treat the direction as solid and the specific numbers as worth verifying against your own data.
What to do this quarter
- Run your own brand through ChatGPT, Perplexity, and Google’s AI Overviews. Read what they say about you. That’s your new baseline audit.
- Check whether your highest-value pages state facts clearly and are structured for extraction, not just for human readers.
- Stop treating Google as the only destination. Map where your buyers actually ask questions.
- Watch conversion quality, not just traffic volume. If AI referrals convert better, raw pageviews become a weaker metric.
The brands that win the next phase won’t be the ones with the most backlinks. They’ll be the ones AI models understand and trust enough to recommend. You can hear the full discussion on TechCrunch’s Equity podcast.