Most GEO Advice Is Just Repackaged SEO, According To Harpal Singh

“Most ‘GEO’ advice misses how brands can become visible beyond AI and is fundamentally repackaged traditional SEO.”

 

People have always disagreed with one another. That’s hardly a new phenomenon, nor is it necessarily a bad thing. Different perspectives, competing ideas and opposing viewpoints are often what lead to better decisions, stronger arguments and more interesting conversations.

The problem, however, is that genuine debate seems to be getting harder to find.

Today, it’s remarkably easy to exist inside an environment where most people think much the same way we do. The content we consume, the communities we join and the algorithms that shape our online experiences often reinforce our existing beliefs rather than challenge them. And over time, that can create the impression that our perspective is the obvious one, shared by everyone who matters and supported by the overwhelming weight of opinion.

The result isn’t greater understanding. Rather, it leads to entrenched positions, fewer meaningful discussions and less willingness to engage with ideas that sit outside our comfort zones.

That’s why we started Quite Contrary, a new TechRound series built around challenging assumptions. Each edition begins with a deliberately provocative or unconventional viewpoint designed to spark discussion. From there, we invite the wider TechRound community to weigh in, disagree, agree or offer a completely different perspective.

 

This Week’s Contrarian

 

This week, the hot take comes from Harpal Singh, AI SEO and GEO Consultant, Fractional CMO and Founder of blimpp. Singh believes strongly that what many people are calling GEO in the modern day is actually pretty much just old-school SEO with a new, fancy label, and it isn’t really providing the added benefit that they think it is.

Do you think Harpal has identified a serious problem and truth within the industry, or is his assessment a little too bold?

 

Harpal Singh, AI SEO and GEO Consultant, Fractional CMO and Founder of blimpp

 

 

Harpal Singh is an AI SEO and GEO consultant, fractional CMO and founder of blimpp based in the UK. He has 12+ years in performance marketing for Google, Meta and Amazon.

He enables DTC, B2B and SaaS brands enhance their presence in ChatGPT, Perplexity and Google AI Overviews by evaluating the pages, buyer guides, comparison sites, and community conversations that AI engines reference. He is also the founder of LeoSEO and UpRate.io.

 

 

Harpal’s Hot Take 

 

“Most “GEO” advice misses how brands can become visible beyond AI and is fundamentally repackaged traditional SEO.”

 

“Copying and pasting your website content for ChatGPT isn’t going to get your brand’s AI answer visibility. AI systems frequently default to some combination of comparison sites, reviews, guides, posts, discussions, and threads when deciding which brands to respond with.

“Thus, it is entirely possible for a brand to dominate the top positions of Google and still be AI answer invisible. The real opportunity here isn’t to ‘optimise for AI,’ but rather to target the sources that impact AI answers and gain a trusted spot within them.

“I’ve noticed that brands have a tendency to spend large amounts of money on developing additional posts, content, or resources while ignoring the sources that AI tools trust. Most of the time, one solid citation in a comparison thread, post, or discussion is worth a hundred posts or resources.”

 

Responses from the TechRound Community

 

• José Rolão: Digital Marketer and Social Media Specialist
• Varun Gazala: AI Innovation Engineer at Kno2
• Zackery Nienaber: Owner and Field Operations Manager, Nienaber And Sons Lawn Care L.L.C.
• Rihards Buks: Founder and CEO at KPIKIT
• Ginger Petrus: Content Marketing Manager at Beacon Nonprofit
• Brad Perniciaro: Owner and Publisher at Living Coastal
• Bryce Wilcock: Account Director at Creo Comms
• Liam McLaughlin: Managing Partner, Europe at Clarity Global

 

José Rolão, Digital Marketer and Social Media Specialist

 

“In fact, from my point of view, GEO is not an extension of SEO, it is like a younger brother, with a large percentage of similar DNA, but with enough of a percentage to differentiate them.

“Even the SEO that is done today is different from what was done 2 years ago, and the SEO from 2 years ago is different from what was done in the 2 years before that. Google’s algorithm, for example, has undergone changes over the years, so the way SEO is done has also changed.

“Answering the question more directly, and of course without going beyond my thesis, the brands whose pages appear well positioned in the SERP but not in AI-generated answers are precisely those that benefit from their age, notoriety, good link work (backlinks, interlinks, references on credible websites, etc.) and from complying with accessibility laws, among other points. All of this counts towards the ranking of a brand’s page.

“In turn, in my experience and observation, the brands that are more frequently suggested by LLMs are those that are more popular and engaged on social media, forums and are mentioned on authoritative websites. And those that have or appear in content optimised for web content, such as blogs, “top 5 of x”, tutorials, tips and some shopping content, for example. In other words, it goes more towards what we usually call consumers’ “pain points”, or the attempt to respond to everything that the consumer/user may need.”

 

Varun Gazala, AI Innovation Engineer at Kno2

‘

“I completely agree with this assessment from a purely architectural standpoint. As an AI engineer building Retrieval-Augmented Generation (RAG) pipelines and multi-agent systems, I can confirm that LLMs do not scrape the web looking for keyword density the way traditional crawlers do. Instead, they look for information density and consensus.

“When an LLM generates an answer, it synthesises context from sources that establish relational truth, often third-party reviews, Reddit threads, or aggregated comparison sites. A brand’s proprietary blog post might have perfect SEO, but to an AI’s retrieval pipeline, it lacks the multi-source consensus required to be weighed as a highly factual entity. Traditional SEO hacks crawler algorithms using localised backlinks and keyword matching, whereas AI engines rely on semantic retrieval architectures that prioritise information consensus.”

 

Zackery Nienaber, Owner and Field Operations Manager, Nienaber And Sons Lawn Care L.L.C.

 

“This take is spot on. In the local service industry, I see too many companies blowing their budgets on endless blog posts and keyword tricks, thinking they can algorithm their way to the top. But when someone asks an AI tool for a recommendation on who to hire, the engine isn’t reading our service pages—it’s pulling from community discussions, local forums, and independent review sites where real neighbors talk about who actually showed up and did a good job.

“If your business isn’t part of those trusted third-party conversations, traditional SEO isn’t going to save you. One genuine recommendation in a local thread carries more weight with an AI model than a hundred self-published pages on your own site. Winning the AI era isn’t about gaming a search bar; it’s about building real-world reputation where people and algorithms actually look for validation.”

If GEO is merely an offshoot of SEO, what accounts for brands that have a strong search presence, yet are seldom cited or endorsed by AI engines? 

“Brands struggle with AI visibility despite strong search rankings because search engines reward keyword optimisation and owned content, whereas AI models heavily rely on third-party consensus, reviews, and external validation that a company’s own website cannot provide.”

 

Rihards Buks, Founder and CEO at KPIKIT

 

“I agree, and our own data adds a proven detail: for most brands, access is not what holds them back. KPIKIT checks how readable a homepage is to AI systems. Of 4922 sites we audited since mid-September, 2334 did well on the search basics, such as clear titles and being open to indexing. Most of them (87%) also allowed AI crawlers in. Even so, about eight in ten of the sites that were both well optimized and open to AI showed little visible proof of reputation on their own pages, such as named clients, certifications or press mentions.

“So the technical work angle is largely done. What is missing is evidence that other people really trust the brand. Our data cannot measure citations directly, but clearly states the pattern: being easy to find and being recommended are different problems. Before writing another post, I would earn one mention where buyers compare options.”

If GEO is merely an offshoot of SEO, what accounts for brands that have a strong search presence, yet are seldom cited or endorsed by AI engines? 

“Search rewards what a website says about itself, while AI answers lean on what other people say about it. In our scans, most sites with strong SEO and fully open AI – crawlers still showed little visible proof of reputation, so the bottom line: the sites mostly are easy for an AI to read but give it little reason to recommend them.”

 

Ginger Petrus, Content Marketing Manager at Beacon Nonprofit

 

“I agree that ranking well in search doesn’t automatically translate to brand visibility in AI. We’re seeing that firsthand at Beacon Nonprofit. We have content that ranks on the first page of Google for relevant searches, and our AI citations have grown from 333 in June to 775 across 95 cited pages; however, direct brand mentions remain comparatively low.

“That distinction has changed how I think about GEO. Creating content that AI systems can find, understand, and cite is important, but being used as a source isn’t the same as being recognized as a brand worth mentioning or recommending. That’s where earned media, third-party mentions, and authority beyond your own website become especially important. Traditional SEO still matters, and probably always will matter, but if you’re only looking at rankings and your own content, you’re missing part of how AI systems discover and evaluate brands.”

If GEO is merely an offshoot of SEO, what accounts for brands that have a strong search presence yet are hardly ever cited or recommended by AI engines?

“AI engines have plenty of sources to choose from, and ranking well in Google doesn’t guarantee your content is the easiest for them to access, understand, or trust. Search visibility gets you in the running, but AI visibility also depends on technical accessibility and authority beyond your own website, including credible third-party mentions, citations, and earned media.”

 

Brad Perniciaro, Owner and Publisher at Living Coastal

 

“I agree with half of it. One citation in a trusted comparison thread is worth more than a hundred new posts. But the opinion stops one step short. AI tools repeat whoever holds the answer, and most brands hold nothing a competitor couldn’t copy. The lasting play is owning data nobody else has.

“On our site, we stopped repeating national estimates for flood insurance and went straight to the federal policy records, 28,552 of them for our stretch of coast. A comparison thread can’t beat the original source, and an AI has a hard time answering that question well without it.

“I’d also drop the old on-page checklist. Keyword density and keyword-stuffed image tags do nothing for an AI answer. Target the questions AI can’t answer yet, then be the only one who can.”

If GEO is merely an offshoot of SEO, what accounts for brands that have a strong search presence yet are hardly ever cited or recommended by AI engines?

“Ranking and being cited reward different things. Google rewards the page that best matches a search, an AI engine repeats the sources it trusts, and a brand that only rewrote what everyone else already says gives it nothing new to repeat.”

 

Bryce Wilcock, Account Director at Creo Comms 

 

“The vast majority of Large Language Models (LLMs) prioritise authoritative, trusted sources when scraping information, so it is vital that companies wanting to improve their AI answer visibility continue to be talked about on platforms other than their own websites and social channels.

“Reviews, Reddit threads, comparison sites and contributing to research and reports can all help improve a brand’s visibility on tools such as ChatGPT etc, as too can earned media coverage, which in recent times has been music to the ears of those working in the PR industry.

“Trade, regional and national news titles are heavily cited by LLMs therefore earned PR should underpin any GEO strategy, with brands ignoring it at their peril. Those who solely focus on content for their own website and socials risk being massively left behind by the competition.”

 

Liam McLaughlin, Managing Partner, Europe at Clarity Global 

 

“Ranking on Google is no guarantee of showing up in AI answers, but the fix isn’t chasing citations in threads and comparison sites. If a company hasn’t clearly put its own story into the world, and been consistent with that story across all of its channels, AI has very little authoritative material to draw on when someone asks about it.

“AI engines build their view of a brand from what credible third parties say, so earned media and a clear, consistent narrative matter more than ever. That’s why brands should spend less on volume content and more on earning credible coverage. It also means breaking down silos. Marketing, SEO and PR teams need to align on one story, shared budgets and common goals. They should treat AI search as a new channel rather than an extension of SEO, and take a test-and-learn approach while measurement catches up.”