The GEO Tactic Half of Marketers Use Is the Highest Risk — What Trust Data Says to Do Instead

FAQ optimization is the most popular generative engine optimization tactic — 49% of marketers use it. It is also the highest risk. Fractl's Q2 2026 study of 1,008 consumers and 150 marketers categorized it that way for a reason: AI can replicate generic FAQ content instantly, and your competitors can too. Meanwhile, consumer trust in AI search dropped from 82% to 54% in a single year, and a 19-market YouGov survey found only 28% of Americans trust AI-generated answers at all. If your entire GEO strategy is FAQ pages, you are building on a surface that consumers already distrust and that AI models can generate without you.

What Two Studies Actually Show About AI Search Trust

The trust numbers are converging from multiple directions. Fractl's Q2 2026 consumer study, presented by cofounder Kelsey Libert at SMX Advanced, tracked a 28-percentage-point decline in consumers who found AI search "more helpful" than traditional results — from 82% in 2025 to 54% in 2026. The skeptic camp grew sixfold in the same period.

A separate YouGov survey across 19 markets, analyzed by Greg Jarboe at Search Engine Journal, sharpened the picture further. Only 28% of U.S. online searchers trust information from an AI assistant, versus 70% who trust a search engine. The U.S. has the lowest AI search adoption rate in the study at 48%, compared to 89% in India and the UAE. Americans are the global outlier — slower to adopt and harder to convince.

The operational signal is clear. AI search is not replacing traditional search behavior. It is layering on top of it, and consumers are not treating AI answers as settled.

Why FAQ Optimization Is the Riskiest GEO Move

Fractl's GEO hierarchy breaks tactics into three tiers, and the distinctions should change how you allocate budget this quarter.

High risk: FAQ optimization (49% of marketers). The content is easily replicated by both AI and competitors. On its own, it offers no defensibility. You can strengthen it with proprietary data or expert sourcing, but without those additions, you are competing with something the model already knows.

Table stakes: Building brand mentions (43%), establishing topical authority (36%), implementing structured data (30%). Necessary, but not sufficient to build a position that lasts.

The moat: Original data and proprietary studies (35%) and digital PR (24%). These are harder for AI to replicate, which is what makes them the only durable position. As Libert put it at SMX: "LLMs and Google need fresh content to pull into timely and common RAG queries." Proprietary data gives them something they cannot generate on their own.

The math backs this up. Ahrefs research analyzing roughly 75,000 brands across ChatGPT, AI Mode, and AI Overviews found that branded web mentions and YouTube impressions showed the strongest correlations with AI visibility — 0.50 to 0.74 on the Spearman scale. Backlink count and ad spend fell below 0.30. AI systems are rewarding brand presence and earned mentions, not link-building volume.

Consumers Validate After the AI Answer — Not Before

The YouGov data reframes what AI visibility is actually worth. Among people who use AI assistants for search, only 16% call AI their starting point. Thirty-two percent use it after trying other sources first. Another 27% use it only for specific questions where they already suspect a direct answer exists.

And when AI does deliver an answer, only 17% stop searching. Twenty-two percent click through to the supplied links anyway. Among daily AI searchers — the heaviest users — that click-through rate climbs to 33%.

The Fractl data adds another layer: consumers now check an average of 2.4 platforms before validating a purchase. Google still leads AI tools three to one for trusted product recommendations, at 39% versus 14% for AI tools. Reddit sits at 15%, right between them.

I see two things happening here. First, an AI citation is not replacing a search result — it is riding on top of it. Second, the operators who only measure whether they appear in the AI answer are tracking the wrong part of the chain.

What to Measure and Change This Quarter

If you are running a GEO program, here is how I would reweight based on this data:

1. Audit your FAQ dependency. If FAQ optimization is your primary GEO investment, you are in the high-risk tier. Layer proprietary data, named expert sourcing, or original research into every FAQ asset. If you cannot add defensible substance, redirect the budget.

2. Track branded mentions, not just backlinks. The Ahrefs correlation data is straightforward: branded web mentions correlate 2x–3x more strongly with AI visibility than backlinks or ad spend. Start tracking mention velocity on high-authority, niche-relevant publishers. If you are not using a media monitoring tool to count this, start.

3. Measure the validation chain. An AI citation is the beginning of the buyer's journey, not the end. Track what happens after the AI answer: do users click through to your site, search your brand on Google, check Reddit? If you are only measuring "appeared in AI answer," you are missing 83% of what consumers do next.

4. Fund digital PR over content volume. Only 24% of marketers invest in digital PR as a GEO tactic. The research says it is the moat. Earned placements on publishers that AI engines index create the kind of brand-entity signals that generic content production cannot.

Where Machine Relations Fits

This data validates something I have been tracking for a while. AI engines do not build authority from your website alone — they build it from what trusted third-party sources say about you. The Ahrefs correlation data makes this concrete: branded mentions on authoritative publishers are the strongest AI visibility signal measured at scale.

That is the core Machine Relations premise. The publications AI engines trust are the same ones that built credibility with human readers for decades. What changed is the reader, not the source. The brands that earn placements in those publications — through real editorial relationships, not cold outreach and content spam — are the ones AI systems cite and recommend.

The trust decline makes this more urgent, not less. If consumers validate AI answers across 2.4 platforms before acting, your brand needs to be findable on every surface they check. That means earned media placements that show up in Google, in Reddit threads, in YouTube results, and in the AI answer itself. GEO without that foundation is a content exercise. Machine Relations with it is a compounding visibility position.

FAQ

Is FAQ optimization useless for GEO?

Not useless, but high risk when it is your only play. Generic FAQ content is easily replicated by AI and competitors. Strengthen FAQ assets with proprietary data, named expert commentary, or original research to make them defensible. Without those additions, reallocate budget toward digital PR and original data — the two tactics Fractl's research categorizes as the moat.

How many platforms do consumers check before acting on an AI answer?

Consumers now validate purchases across an average of 2.4 platforms, according to Fractl's consumer study. Google still leads AI tools for trusted product recommendations at a 39% to 14% trust margin. Even among daily AI users, 33% click through to supplied links and only 17% stop searching after the AI answer.

What signals correlate most with AI visibility?

Branded web mentions and YouTube impressions showed the strongest correlations with AI visibility (0.50–0.74 Spearman) across 75,000 brands analyzed by Ahrefs in ChatGPT, AI Mode, and AI Overviews. Backlink count and ad spend ranked in the weakest tier, below 0.30. Brand presence matters more than traditional SEO scale metrics for AI citation.