The brands winning in AI search right now are the ones that have built verifiable entity presence across the web, published original data, and structured their content for extraction. I have watched Velar go from zero to 100k followers on X and ranked RemoteStack number one for its core keyword on a new domain with zero backlinks. The same principles apply to GEO: AI citation is earned, not bought.
Three brands that dominate AI search right now are Ahrefs, Semrush, and HubSpot. Each of them follows a pattern I have reverse-engineered from Google's AI Overviews, Perplexity, and ChatGPT outputs. They are not the only winners, but they are the most instructive.
Ahrefs: Original Data That AI Cannot Ignore
Ahrefs publishes its own SEO statistics and industry reports. When a user asks ChatGPT "what is the average click-through rate for position one?" the AI cites Ahrefs because Ahrefs owns that number. No one else has the same dataset.
This is the core of entity-based citation. Ahrefs is not just a blog. It is a consistent entity with a defined expertise area, backlinked by thousands of sites, and referenced in third-party research. Google's Knowledge Graph and AI models treat Ahrefs as an authority on SEO metrics because the web agrees on that fact.
Actionable takeaway: Publish one piece of original research per quarter. Survey your audience, analyse your internal data, or compile industry benchmarks. That dataset becomes your citation anchor.
Semrush: Comparison Content That Gets Extracted
Semrush consistently ranks in AI responses for comparison queries. Search "Semrush vs Ahrefs" in any AI tool and Semrush's own comparison page is often cited. Why? Because Semrush wrote the definitive comparison with structured tables, pros and cons, and pricing breakdowns.
Search Engine Journal has noted that comparison content performs well in AI overviews because it directly answers evaluative queries. AI models prefer content that presents multiple options with clear trade-offs. Semrush does this better than most.
Actionable takeaway: Write comparison pages between your product and two competitors. Use a table format. Include specific numbers, feature differences, and use cases. Do not be vague. AI extracts specificity.
HubSpot: FAQ Structure and Entity Consistency
HubSpot's blog is structured around FAQ blocks and clear headings. When an AI model needs an answer to "what is a CRM?" it pulls from HubSpot because HubSpot has defined that term consistently for years across its own site, Wikipedia references, and third-party articles.
HubSpot also maintains a FAQPage schema on its help articles. According to Google's documentation, structured data helps search engines understand content. AI models trained on web crawl data benefit from the same signals.
Actionable takeaway: Add FAQ sections to your core content pages. Use FAQ schema markup. Answer one question per paragraph. Keep each answer between 40 and 60 words. That is the length AI models most reliably extract.
The Common Thread: Entity, Specificity, and Third-Party Proof
Every brand winning in AI search has three things in common. First, they are verifiable entities. You can find them on Wikipedia, Crunchbase, or industry directories. Second, their content is specific. They use numbers, names, and timeframes. Third, they have third-party mentions. Other authoritative sites link to them, cite their data, or reference their brand.
I ranked RemoteStack for its core keyword on a new domain with zero backlinks by building entity presence first. I created a Crunchbase profile, a LinkedIn company page, and got mentioned on three industry roundups. Within weeks, Google surfaced RemoteStack in AI overviews for remote work queries. The backlinks came later.
Entity before content. Be a verifiable, consistent entity across the web before you write a single word of GEO-optimised content.
What About Smaller Brands? Can You Win Without a Budget?
Yes, but you have to be smarter. You cannot outspend HubSpot. You can out-specialise them. Pick a narrow niche, publish original data from a survey of 100 people, and get mentioned on one mid-tier industry blog. That is enough to start earning AI citations.
I have seen a solo founder in the B2B SaaS space get cited in Perplexity for a blog post titled "average email open rates for SaaS in 2024" because they surveyed 50 founders and published the raw numbers. No backlinks. Just specificity and a unique dataset.
Actionable takeaway: Pick one question your audience asks ChatGPT. Survey 50 to 100 people in your network. Publish the results with a clear methodology. That post becomes your GEO anchor.
Reverse-Engineer the AI Answer
The most effective GEO strategy is to imagine what a buyer asks ChatGPT, then be that answer. If your buyer asks "what is the best project management tool for remote teams?" your content should be the definitive answer to that exact question. Do not write a generic blog post. Write the answer AI would extract.
I use Perplexity to reverse-engineer this. I type my target query, see which sources AI cites, and analyse why. Usually the answer is: original data, clear structure, and entity authority. Then I build content that fills the gap.
Actionable takeaway: Every week, pick three buyer questions. Search them in Perplexity or ChatGPT. Note which sources appear. Build content that matches their structure but adds your own unique data.
Backlinks Are Grunt Work, Not Budget Work
Brands winning in AI search do not buy links. They earn citations through original research, guest contributions on authoritative domains, and consistent entity signals. I have never bought a backlink for any site I ranked. Every citation came from being useful first.
Focus on getting mentioned in one roundup post, one Wikipedia reference, and one industry report. That is enough to signal entity authority to AI models. The rest is content structure and specificity.
Actionable takeaway: Identify three authoritative sites in your niche that publish roundup posts or resource lists. Pitch them a specific data point from your original research. One mention is worth ten guest posts.
The Bottom Line
Ahrefs, Semrush, and HubSpot win in AI search because they built entities, published specific data, and structured content for extraction. You can do the same at any scale. Start with one original dataset, one FAQ page, and one comparison post. That is your GEO foundation.