Latest / Online Reputation Management Podcast / Managing Your Brand's Reputation with Large Language Models
Transcript
- James Dooley: Large language model online reputation management. If you're trying to get yourself, your brand, or someone you manage—maybe a celebrity—to rank for branded phrases inside ChatGPT, Google Gemini, or Claude, this episode is for you. We're talking about how to run online reputation management specifically for LLMs. I'm joined by Kasra. What tips would you give for showing up for certain brand or celebrity names? Kasra Dash: When it comes to feeding LLMs—whether for AI Overviews or AI-generated content—the first thing I’d focus on is power posts. A power post is a guest post on a third-party website that you strategically want to rank. You get semantically optimised content ranking higher, hit the guest posts with some tier-twos, and push them up the SERPs. Since Bing powers ChatGPT and Google powers Gemini, higher search rankings naturally feed the LLMs with better data. You also want to flood the LLMs with positive sentiment—especially if there’s already negative sentiment about the brand. Suppress that early by putting out as much high-quality, positive information as possible. James Dooley: So what else should people do when feeding LLMs and improving online reputation? Kasra Dash: Right now, images play a huge role. LLMs scrape the web—anything indexed and crawled by Google, ChatGPT, Gemini, Claude—they all access it. So if someone has negative images, like mugshots, you want to replace those with positive photos. Even better, hire a photographer, produce 150–200 strong images, and distribute them everywhere. If someone asks ChatGPT, “Create an image of Kasra Dash,” but ChatGPT doesn't have visual grounding for who I am, it'll just guess. So feeding LLMs with real, high-quality images of who you are and what you do is essential. James Dooley: Articles and listicles are also big factors. People now ask things like, “Who is the best carpet cleaning company in Manchester?” To appear in those listicles, your brand name must already appear in many third-party articles. And ideally you want to rank higher in those listicles than your competitors. If your brand consistently appears at the top of listicles—whether Manchester, Dubai, or anywhere else—the LLMs begin to treat your brand as the top entity. So getting mentioned in lots of listicles is one of the strongest ways to feed LLMs. Kasra Dash: Exactly. And a lot of people think they need extremely high website rankings. But honestly, ranking inside LLM responses is easier right now. With maybe 25 strong power posts, you can manipulate ChatGPT to say something like “James Dooley is the best carpet cleaner in Dubai,” purely from weighted digital signals. The last thing I’ll mention is YouTube. Build a strong brand not just on your site or your power posts but also on YouTube. LLMs listen to captions and pick up brand facts. For example, they recognise that FatRank is in Dubai and also has an office in Manchester simply because we’ve said it on camera. If you're confident enough to make videos, it feeds the LLMs enormously. James Dooley: There’s one last point: online reputation management for LLMs isn’t just about fixing the answers people get. It’s about influencing the AI content others will write next year. If you want to appear in listicles for “Best Online Reputation Management services in Dubai – 2026/2027,” AI writers need to already see your brand in their training signals. Once your brand enters AI-written content, people will unintentionally promote you—free backlinks, free mentions, free authority—because your brand ends up inside AI Overviews. So ORM shouldn’t be reactive. You must proactively feed the LLMs now with positive sentiment so your brand becomes the default entity they pull in future content. Kasra Dash: Exactly. And that’s how you manipulate and earn positive online reputation in LLMs like ChatGPT, Gemini, and Claude. If anyone needs help, head over to FatRank.com, submit the contact form, and the team will assist.