Building product authority in a GEO world
- Teresa Sperti
- Aug 4
- 7 min read
Search as we know it has changed.
The introduction of Google’s AI overviews has caused clicks to fall by 30%, providing summaries of search results at the top of the SERP to provide a quick, concise answer to your search queries.
And then came social search as the new Google, with 51% of Gen Z turning to social to find out more information on product brands and services. And the latest announcement from Instagram means your social posts are now visible directly in Google’s SERP’s.
Then ChatGPT (literally) entered and with it the conversational era began to take flight, with referral traffic to news sites reaching 243.8 million visits in April 2025, up 98% from 123.2 million in just 4 months. And with the LLM already handling over 1 billion queries a day, it is fast catching up Google.
So, after years focused on SEO: keywords, backlinks, domain authority – there’s a new kid on the block.
Generative engine optimisation, or GEO.
Where SEO helped you rank when users searched, GEO helps you show up in response to questions, problems and more. Instead of search engines returning ten blue links and 10+ pages of results (who honestly ever looks past the first page), generative engines (think ChatGPT, Google’s AI Overviews, Perplexity) synthesise a direct answer, and they choose which brands and products to mention based on trust signals, quality content, and online reputation. It’s a huge shift for eCommerce teams, especially those thinking about the digital shelf.
The question is no longer ‘how do you get to the first page of Google?’. It’s ‘how do you make sure your product or brand mentioned in the answer?’
In traditional SEO, you needed link authority - backlinks from high-quality, reputable sites - to earn your place at the top. In a GEO world, it’s about what we are calling product authority at Arktic Fox.
Product authority in the context of GEO refers to the ability for a product or its associated content to be recognised and utilised by AI-powered search engines as a reliable and authoritative source of information. This means that when a user asks a question related to your product or a topic it covers, the generative engine is more likely to surface your content as a relevant and trustworthy answer.
Think of it as the generative engine’s version of trust. Instead of relying on links between websites, LLMs look for signals of credibility, accuracy, and relevance specifically about your product.
And it’s not just about your PDP anymore - it’s about the entire online footprint your product leaves.
Are people talking about it in forums and reviews?
Is it mentioned by reputable sources?
Does the content about it show clear EEAT? (Experience, Expertise, Authoritativeness, and Trustworthiness)
Product authority is what makes your brand not just visible in GEO - but featured. And the more authority your product has, the more likely it is to show up in the results.
In the world of GEO, product authority is in the eyes of the LLM. These models aren’t just scanning your PDP or pulling data from your sitemap, they’re making complex inferences based on patterns across the entire web.
They’re trained on billions of data points: content, reviews, news articles, blog posts, Reddit threads, YouTube transcripts, even brand mentions in passing. All of that forms the model’s understanding of what your product is, how it’s perceived, and whether it’s trustworthy enough to be cited in an answer.
And as LLMs are used more as part of the discovery journey, to lift the load from shoppers, they’re becoming the new gatekeepers to product visibility.
So, if product authority is the goal, the model is the judge.
And what it comes down to, is share of model.
A brand’s share of model details how often, how prominently and how favourably it is surfaced by LLM’s compared to other brands. It’s the AI version of share of search (how often do people search for my brand compared to competitors?) and share of voice (how often do people talk about my brand compared to competitors?).
Share of model offers the lowest brand control compared to share of voice and share of search. Unlike traditional SEO where you can optimise your keywords, or media and social where you can invest in paid reach, SOM is built on what the model already knows about you.
LLMs make decisions based on their training data: what’s said about your product, where it’s said, and how often it’s said. If you’re not showing up in high-quality, trusted sources or being talked about with the right signals, you’re invisible to the model, no matter how great your PDP is.
Generative engines don’t just surface a single source, they surface the most mentioned, most trusted, and most frequently reinforced options. That means it’s not enough to show up once. You need to show up consistently across different transactional sources, retailer sites, marketplaces, buying guides, and expert reviews to increase your chances of being featured in one of the few available result slots. In our own testing, we’ve seen that when a consumer asks for “the best product” in a category, LLMs often surface just three to six specific brand or product recommendations. If you’re not showing up in multiple credible sources, you risk being left out of that short list entirely.
The 7 steps to achieve product authority
1. High-quality content (that answers real questions)
In a GEO world, your content needs to do more than describe - it needs to educate. Generative engines favours product content that’s clear, accurate, and genuinely helpful as it supports response generation: specs, comparisons, FAQs, benefits, and buyer guides all matter.
Structure also counts. LLMs prefer well-organised, scannable content with natural language and clear headings. As Contentful explains, think beyond PDPs and focus on content that answers real-world queries. If your product content isn’t helpful to a human, it won’t be helpful to a model.
And it’s not just about answering questions generally, it’s about answering them for different types of buyers. Consumers use the same product for different reasons, and LLMs are increasingly picking up on those context-specific signals. Tailoring your content to speak to these different use cases - whether it's expert-level detail for enthusiasts, beginner guides for first-timers, or problem-solving content for niche needs - helps expand your product's relevance. The more angles your product is covered from, the more likely it is to be considered useful in a variety of LLM-generated answers.
2. Optimise for E-E-A-T: Experience, Expertise, Authoritativeness, and Trustworthiness
E-E-A-T - Experience, Expertise, Authoritativeness, Trustworthiness - is foundational to how LLMs assess content. These signals tell the model that your product is safe and credible to recommend.
This means using expert-led content, citing credible sources, earning mentions on reputable sites, and collecting trusted reviews. As Salsify notes, optimising for E-E-A-T makes your brand more valuable as a generative resource, and more likely to appear in AI-generated responses.
3. Invest in Digital PR
If you want generative engines to mention your product, you need to be talked about in the right places.
That’s where digital PR comes in.
Getting your product featured on trusted, authoritative websites through product roundups, expert reviews, or thought leadership helps build the external signals LLMs rely on. These mentions not only boost your credibility, they train the model to associate your brand with specific categories or queries.
Being visible in the places AI learns from (news sites, forums, blogs, Wikipedia, reviews) helps ensure your product becomes part of the model’s knowledge.
4. Leverage influencer marketing
Influencers aren’t just invaluable for awareness; they’re increasingly shaping what LLMs learn. Reviews, tutorials, and mentions across YouTube, blogs, TikTok, and even podcast transcripts contribute to the digital footprint AI models absorb.
Collaborating with credible creators helps build social proof and trust, especially when content includes detailed product experiences. Just like a backlink once boosted your domain authority, an honest, high-engagement mention from a respected voice can strengthen your product’s authority in generative search.
5. Encourage and amplify customer reviews
Customer reviews are gold in the world of GEO. They’re real, relevant, and rich in natural language - exactly the kind of content LLMs love.
Whether it’s on Google, Trustpilot, Amazon, or niche retail sites, these reviews help build trust signals that feed into a product’s perceived credibility. A steady stream of detailed, authentic feedback shows that people are engaging with your product and gives the model more context to draw from when generating answers.
6. Engage in relevant communities
Reddit threads, Quora answers, Discord chats - these are the corners of the internet where people ask questions and share real opinions. When your brand (or people genuinely talking about your product) shows up in these spaces, it adds depth to your product’s online footprint. Participating (but not promoting) can help build reputation, surface long-tail keywords, and shape how the model understands your brand within specific contexts. Plus, it gives you a direct line to what people actually think — how they use your product, what they love, and what they’d change. It’s insight you can act on, like Refy Beauty did when they reformulated their brow sculpt after customers flagged an issue with the formula in reviews.
7. Ask the AI
Understanding how AI perceives your brand is also important to consider. Sometimes the best way to understand what the model values... is to just ask it.
Prompt tools like ChatGPT, Perplexity, or Claude with queries your customers might use – like:
“What’s the best low-VOC paint for interior walls?”
“Which red wine pairs well with roast lamb?”
“Best hiking shoes for flat feet Australia”
“What to make with leftover hot cross buns?”
See which products or brands show up, what sources are cited, and how answers are phrased.
This can give you clues about where LLMs are pulling data from and where you might want to target content, outreach, or digital PR. Think of it as competitor research, but through the eye of the LLM.
The way products show up online is changing. It’s no longer about keywords, but credibility. That means what shoppers see is increasingly shaped by how much trust platforms place in your brand, your content, and your product coverage. Getting ahead in this new environment isn’t about playing the algorithm, it’s about showing up consistently, being accurate, and building real signals of authority.
Building product authority isn’t a one-time optimisation tactic. It’s an ongoing strategy to earn trust, demonstrate relevance, and show up consistently in the digital spaces where LLMs learn.
As generative engines become the new front door of the shopping journey, understanding and influencing your share of model will define whether your products are seen or skipped. Now is the time to adapt, build, and invest in the signals that matter most to machines - and to people.
At Arktic Fox, we partner with CPG and FMCG brands to build stronger product content strategies so you can show up where it matters, and stand out when it counts. Want to learn more? Let’s chat.
