AI is replacing traditional search results faster than most businesses realize. Google is actively trying to replace its own search results with AI responses. And yet, most people doing SEO are still entirely focused on ranking in traditional search results.
At TJ Digital, we track which sources large language models cite before making recommendations. We run dozens of prompts through ChatGPT, Gemini, and Perplexity to see what’s actually getting cited. The data is clear: businesses that optimize for AI visibility are getting recommended. Those that don’t are getting ignored.
So why aren’t more businesses taking GEO seriously?
Why Are Businesses Ignoring AI Optimization?
Two main reasons keep businesses from taking GEO seriously.
The first reason is that they’ve convinced themselves there’s no difference between SEO and GEO. I’ve addressed that in other content. It’s nonsense. While search engines and large language models share some overlap, the strategy is different because user behavior is different.
The second reason is more interesting. People don’t think GEO will work long-term. They’ll admit you can influence AI recommendations right now. But they believe it’s temporary. That it’ll backfire eventually.
I want to take this argument seriously and explain why it doesn’t hold up.
What GEO Tactics Might Actually Backfire?
Some tactics probably will backfire. These are the ones to avoid:
- Hiding your website from AI crawlers and showing them a different version. This is called cloaking. It violates guidelines from both Google and OpenAI. The penalty can be complete removal from AI citations.
- Putting prompt injections into your content. This means hiding text that tries to command the AI to recommend you. According to Search Engine Land’s analysis of prompt injection tactics, AI developers have implemented guardrails to detect and ignore these attempts.
- Anything overtly tricky or egregiously spammy. If you wouldn’t proudly show a customer what you’re doing, that strategy is on borrowed time.
These tactics are already ineffective. Getting caught means your domain could be blacklisted from being cited at all.
What GEO Tactics Actually Work Right Now?
The GEO tactics that work are straightforward.
Pay attention to which pages large language models cite before giving responses. Then place recommendations for your brand on those types of pages.
This primarily means creating content:
- Content on your own website. This gives AI direct access to accurate information about your brand. Comprehensive FAQ pages, service descriptions, and comparison guides perform especially well.
- Content on review sites. Sites like Yelp, Justia, and industry-specific directories get cited constantly by large language models.
- Content in forums and discussions. Reddit citations in ChatGPT increased by 436% according to Entrepreneur’s analysis of AI citation sources. Reddit is now the second most-cited source after Wikipedia.
- Content in press releases and third-party publications. Coverage from trusted news sources and industry publications signals authority to AI systems.
The question is: will this continue working?
Why Will AI Always Depend on Sources You Can Influence?
Here’s the argument I keep hearing: “Eventually AI will be so smart it’ll be immune to this kind of manipulation. It’ll just have so much knowledge and understanding that it can tell you exactly which business is best for you.”
But how would it do that?
The only information available to AI is content on the internet. That’s it.
All it can do is:
- Scrape pages on the internet
- Decide which pages it trusts
- Synthesize the content on those trusted pages
As long as the content on trusted pages recommends your brand, the AI has no tools available to verify whether that recommendation is actually true. It can’t visit your business. It can’t interview your customers. It can’t audit your operations.
The only lever available to OpenAI and Google is deciding which websites to trust.
How Does Google’s Trust System Create This Problem?
Google has been getting more selective about which websites it trusts. ChatGPT, on the other hand, will cite almost any website.
But here’s the thing: it doesn’t matter how selective Google gets. They will always have to rely on websites that can be influenced through legitimate means.
| Source Type | Why AI Trusts It | How It Can Be Influenced |
| Review sites (Yelp, Justia, etc.) | User-generated trust signals | Paid placements, review campaigns, responding to reviews |
| Forums (Reddit, Quora) | Authentic peer recommendations | Genuine participation, helpful answers to real questions |
| News publications | Editorial authority | PR outreach, newsworthy content, expert commentary |
| Your own website | Direct source of brand information | Content creation, clear messaging, comprehensive FAQs |
Large language models cite review websites constantly. Sites like Yelp or niche directories like Justia appear in AI responses all the time. You can pay these review sites for premium placement at the top of their lists. When you do, large language models start recommending you more.
Now, you might think: “Eventually AI will be smart enough to ignore sponsored results.”
But if AI ignored sponsored results, you’d actually get penalized for paying for placement. That’s not a viable solution.
And if AI only looked at reviews instead of placement order? Then you’re just rewarding the businesses that are best at getting reviews. That typically means larger businesses with more customers.
Why Can’t Google Just Recommend Only Big Brands?
Google could theoretically avoid recommending all but the most established brands. That would make it impossible for any small or boutique business to get recommended in AI results.
But Google understands this would create a terrible search experience.
Big brands already have an inherited advantage. Google is already going out of their way to promote smaller, niche brands to avoid monopolizing results.
The problem is there isn’t a reliable corpus of trusted information on smaller brands. Often the only way AI can get accurate information about a brand is from that brand’s own website.
This is exactly why creating content on your own website is the most effective GEO strategy for most businesses.
Does This Problem Apply to All Source Types?
These challenges apply across every source type AI relies on.
Discussion forums: Is Google going to stop citing Reddit? Reddit’s data is officially licensed to OpenAI. The citations keep increasing.
Press releases and news: Large publications still cover businesses, and AI cites them. You can influence coverage through PR and creating newsworthy content.
Third-party websites: Google could limit citations to only the most trusted sites. But then they’re giving a small handful of websites all the power over AI recommendations.
The closer you look, the more you see there’s no way out of this puzzle. For the foreseeable future, AI will be dependent on websites that can be influenced through legitimate content and marketing efforts.
What Does This Mean for Good Businesses?
Here’s what matters: it’s still easier to be recommended if you’re running a good business.
Good businesses get more reviews naturally. Good businesses have people talking about them online. Good businesses create content that genuinely helps their customers.
But here’s the catch.
If your competitor is running an effective GEO campaign and you’re not, it probably won’t matter how good your business is. They’re going to get recommended more often.
How Should You Get Started with GEO?
If you’re convinced GEO matters, here’s what to focus on:
Track your AI visibility. Use tools like Peec.ai to see which sources large language models cite before recommending businesses in your industry. This tells you where to focus your efforts.
Create content on your website. Comprehensive, well-structured content that directly answers questions your customers ask. FAQ sections perform especially well for AI citations. Comparison pages help too. Detailed service descriptions give AI accurate information to work with.
Build presence on the sites AI trusts. This means review sites, industry directories, forums where your customers ask questions, and publications that cover your industry.
Be consistent across all platforms. AI builds understanding of brands as entities. If your brand is described differently across different sites, AI gets confused about who you are and what you do. Use the same name, same descriptions, same key facts everywhere.
Don’t try to trick the system. Focus on being genuinely helpful. Create content that would hold up if a human reviewed it. The tactics that work are the ones that make the internet better, not worse.
Frequently Asked Questions About GEO
Is GEO the same as SEO?
No. While there’s overlap, the strategy differs because user behavior differs. Large language models use search engines differently than humans do. They often look at dozens of results and synthesize information across multiple sources. Humans typically click on one or two results. This changes which content performs best.
Will AI eventually be immune to optimization?
Not in any foreseeable future. AI can only work with the information available on the internet. It has no way to verify whether recommendations on trusted websites are accurate. As long as trusted websites can be influenced through legitimate content creation, GEO will remain effective.
Can you pay to show up in AI results?
Not directly. There’s no equivalent of buying a Google Ad for ChatGPT. However, you can pay for placement on sites that AI frequently cites, like review directories. This indirectly influences AI recommendations.
Should I stop doing SEO and only do GEO?
No. The content that performs well for AI visibility largely overlaps with what performs well for traditional SEO. Creating high-quality, comprehensive content helps with both. Think of GEO as an extension of SEO, not a replacement.
The Bottom Line
GEO isn’t going anywhere. The fundamental mechanics of how AI retrieves and synthesizes information mean there will always be ways to influence recommendations through legitimate content creation and marketing.
The businesses that recognize this and adapt will capture visibility in AI results. The businesses that wait for “best practices” to emerge will find those practices have changed by the time they hear about them.
At TJ Digital, we track AI citations and test GEO tactics continuously. If you want to see how your business currently shows up in AI results and what you can do about it, contact us for a free audit. We’ll show you exactly what large language models are saying about your brand and your competitors.