According to Visium, Front Row's proprietary AI visibility tracking tool, blog-style content is cited roughly 5x more often than product pages across the prompts we track. The data is a glimpse of what large language models are actually built to reward, and it’s not just product optimization.
Ask ChatGpt for the best curly hair conditioner, and it won’t send you a product page. The process is AI will give you an answer, with a citation attached, and that citation is what’s doing the selling before you’ve clicked anything.
This is the part of the AI shift that a lot of brands are still catching up to. The type of content earning citations has changed, and product pages are losing to blog posts by a wide margin.
According to Visium, Front Row's proprietary AI visibility tracking tool, blog-style content is cited roughly 5x more often than product pages across the prompts we track. The data is a glimpse of what large language models are actually built to reward, and it’s not just product optimization.
The consumer shift is already here
The numbers on how people shop now aren't subtle. Across a recent Accenture consumer survey, 91% of shoppers said AI influenced or validated a purchase decision, 8 in 10 said they trust AI recommendations as much as brand messaging, and AI assistance cut time-to-purchase by 47%. AI isn't a side channel anymore. For a growing share of shoppers, it's the first stop.
Which raises the real question: when an AI model answers "what's the best X," why does it point to a blog post instead of the manufacturer's own product page?
SEO earns a ranking while GEO earns the answer
Traditional SEO and generative engine optimization (GEO) are solving different problems, even though they run on the same foundation.
Search someone's exact query in Google the old way, "best night face moisturizer," and you get a ranked list. Three blue links, three brands competing for a click. The user still has to open each one, scan it, and decide for themselves. SEO gets you on the page. The user does the rest.
Ask an AI assistant the same thing conversationally and something different happens. It skips the list and gives a direct answer, complete with a specific product name and the reasoning behind the pick, sourced from a couple of citations. No click required. The AI has already made the comparison the user would have made. If your content is one of those citations, you are the answer. If it isn't, you don't exist in that conversation at all.
That's the entire premium of GEO: less about visibility and more about being the source the model trusted enough to repeat.
Why product pages don't make the cut
Product pages are built to sell. They're built around a single SKU, and they read like marketing copy because that's exactly what they're for; benefit statements, brand voice, calls to action. While LLMs can extract marketing copy, they’re really built to extract answers, and the content that maps most cleanly to an answer is the content that’s already structured like one. This means having a clear answer, a direct response, comparative reasoning, and real specifics. That’s a blog post format in comparison to a DTC product page.
A blog post titled "Best exfoliating cleansers for sensitive skin" can compare five products, explain what exfoliation actually means, and answer the follow-up questions a shopper would naturally have. A product page can't do any of that without stopping being a product page. It's optimized for a shopper who has already decided. It was never built to convince someone who hasn't.
Blogs have always been where brands could go deep: answer real questions, walk through comparisons, build authority in a category. What's new is that this exact structure, question-driven, problem-solution, specific, is precisely the format AI models are built to lift and cite.
Blogs are still the connection point between search, trust, and organic visibility. With AI in the mix, that connection just got more important.
What this looked like in practice
We saw this play out directly with Canopy, a home wellness brand whose entire category (humidifiers, filtered shower heads, diffusers) maps almost perfectly onto the questions people now ask AI: "best dishwasher-safe humidifier," "best filtered shower head for dry skin," "best baby bath products."
Canopy already had a strong SEO and paid media foundation. What they didn't have was a strategy for AI-generated answers, or a way to measure whether they were showing up in them.
We audited every collection page for GEO gaps, mapped target prompts across the major LLMs, and wrote FAQ content for each page grounded in real consumer pain points: mold resistance, dishwasher-safe design, filter change frequency. Then we tracked the results using Visium from day one.
In 24 days, Canopy saw a 163% increase in mention rate and a 190% increase in citation rate across ChatGPT, Gemini, and Claude. That's how fast the right content, in the right format, can move the needle on AI visibility.
Is your brand visible to AI agents?
The shoppers asking AI for recommendations aren't going away, and the brands earning those citations right now are the ones treating their blog as infrastructure, not an afterthought. If your product pages are doing all the talking and your blog isn't built for the questions people are actually asking an AI, you're invisible at exactly the moment it matters most.
Learn more about our SEO/GEO capabilities here.
