There’s a lot of conflicting information about how to rank in ChatGPT and other AI search platforms.
Just this week, I’ve seen people say that ChatGPT hates press releases. In my experience, they love press releases. I’ve heard that ChatGPT never cites Reddit and that ChatGPT is always citing Reddit. Another person I respect said that large language models don’t like LinkedIn, but in my experience, they love LinkedIn.
So who’s right? They’re all right. What’s not being talked about enough right now is that the websites these large language models cite depend entirely on the prompt itself and the industry context.
We currently track about 4,000 different prompts across 25 industries. We run all these prompts every day across ChatGPT, Perplexity, and Google’s AI mode. Whenever someone asks me which websites these large language models cite most often, the only coherent answer is: it depends on the industry.
The One-Size-Fits-All Myth
The biggest mistake people make with AI optimization (AIO) or Generative Engine Optimization (GEO) is believing in universal rules.
You’ll see articles claiming “ChatGPT loves Wikipedia” or “Reddit dominates AI citations.” These statements are both true and meaningless.
Recent research from Yext analyzed 6.8 million AI answers across ChatGPT, Gemini, and Perplexity. They found that 86% of citations came from brand-controlled sources like websites, directories, and reviews. Forums like Reddit accounted for only about 2% of citations when considering local intent.
But here’s where it gets interesting. A separate study from SEMrush found Reddit was the single most cited source overall, outpacing Wikipedia and YouTube. Another analysis reports ChatGPT pulls nearly 48% of its citations from Wikipedia, with Reddit around 11%.
How can all these studies be true at once?
Context Is Everything
AI models tailor their source selection to the industry and intent of the question.
The Yext study breaks this down clearly:
Retail and Finance: Nearly half of AI citations came from brand-owned websites in these sectors (47.6% in retail, 48.2% in finance). When people ask about products or financial services, AI pulls directly from company sites and listings.
Healthcare: AI answers cite medical directories and Q&A sites most heavily. In healthcare queries, 52.6% of citations were to health sites like WebMD or Vitals. AI leans on recognized medical authorities rather than forum posts.
Food and Hospitality: User reviews and local listings matter here. Yext found that 13.3% of citations in food-service queries came from review and social sources. Think Yelp or TripAdvisor for restaurants.
Different industries see completely different authoritative sources. A query about medical advice triggers AI to cite health experts. A tech question might draw from Wikipedia or Stack Overflow. There’s no one-size-fits-all source list.
Why General AI Optimization Advice Fails
Broad “AI SEO” tips often miss the mark because generative search is highly dynamic and context-dependent.
Unlike traditional SEO, there is no fixed ranking algorithm for AI responses. Each large language model answers differently. As one analysis notes, “LLMs lack a unified or transparent ranking system” and their outputs can change drastically with phrasing, context, or which model is used.
A small tweak to your prompt or a different model can lead to completely different sources being cited.
AI answers often contradict traditional SEO wisdom. The SEMrush study found that ChatGPT and Google’s AI favored community-driven sources over official brand content in most industries. ChatGPT even cited Reddit finance threads far more often than expert bank sites.
The field is rapidly evolving. LLMs update their training data and retrieval methods frequently. New models emerge. Advice that was true for one model last month might not hold next month.
Blanket rules are risky. Effective AI optimization requires testing in your specific context.
What Actually Matters for Your Business
Ignore any information you hear about ChatGPT loving this website or hating that website.
All that really matters is what websites these models cite when people are asking about your product or service.
The only way to know that is to track it.
You need to identify the key content hubs in your vertical and ensure your information is available there. If your company is in finance, you benefit from being listed on financial directories or having press coverage cited by sites like Forbes or Investopedia, since AI will check those first. If you’re traveling, focus on travel guides and review sites.
In every case, the model uses the authoritative sources relevant to your niche.
How to Track What Actually Matters
To audit your brand’s visibility in AI search results, you need to simulate how real users might encounter your brand via AI answers.
Here’s the basic approach:
Identify Key Queries: List the questions or keywords most relevant to your brand. Include common variants and location-specific queries if applicable.
Query Each AI Platform:
Run your queries in ChatGPT (with browsing enabled), Perplexity, and Google’s AI mode. Note which websites are cited for each query.
Use Tracking Tools: Manual checks work for spot tests, but scalability requires tools. Specialized platforms like Peec.ai, Profound, or RankPrompt can automate this process. These tools let you enter your domain and target terms, then periodically run those prompts across multiple LLMs and track mention frequency, sentiment, and citation sources.
Analyze and Optimize: Use the collected data to identify gaps. If a competitor is frequently cited in your topic but you’re not, examine where AI got that info. You can then create or improve your content on those sources.
Peec.ai vs. Profound: Which Tool to Use
The tool I use and highly recommend is Peec.ai. It’s much more affordable than the leader, Profound, and does the exact same thing.
Here’s the breakdown:
Platform Coverage: Peec.ai supports a wide range of LLMs (ChatGPT, Gemini, Claude, Perplexity) and even newer engines. Profound covers many major systems but lacks some newer ones like current Gemini coverage.
Pricing: Peec.ai’s entry plan starts at €89/month (about $100 USD) for 25 prompts and basic analytics. Its higher tiers (Pro €199, Enterprise €499) add more prompts, languages, and features. Profound’s base price is around $499/month with custom enterprise pricing available.
Ease of Use: Peec.ai is more user-friendly for those new to AI tracking. It has a clean interface and built-in competitor benchmarking. Profound’s interface is more technical and customizable, with a steeper learning curve.
Analytics: Profound excels at deep analytics with advanced pattern recognition, customizable dashboards, and API integration. Peec.ai focuses on core visibility metrics (mentions, positions, sentiment) with straightforward historical trends by region and language.
For most small to medium businesses, Peec.ai offers better value with broader platform coverage on a tight budget. Profound delivers greater analytic depth for enterprises willing to invest more.
Stop Looking for Universal Rules
My advice is simple: ignore any information you’re hearing about ChatGPT loving this website or hating that website.
All that really matters is what websites they care about when people are asking about your product or service. The only way to know that is to track it.
Different verticals see different authoritative sources. Different prompts trigger different citation patterns. Different models have different preferences.
There are no AI optimization experts yet. This industry is brand new and changing rapidly. Anyone claiming they have a secret system is either lying or will be wrong in three months when the models update.
What works is data. Track your specific prompts in your specific industry. See what sources AI actually cites for your topic. Then make sure your information is available and authoritative on those sources.
That’s AI optimization in 2025.
Ready to Track Your AI Visibility?
If you want help setting up AI tracking for your business and ensuring you show up where your customers are actually searching, contact TJ Digital. We track thousands of prompts across dozens of industries every day and can show you exactly where you stand in AI search results.