77% of brands are invisible to AI platforms like ChatGPT. They score below 5 out of 100 on AI visibility. Meanwhile, the brands that do show up convert visitors at 2.3x the rate of traditional search traffic, with 4.4x higher visitor value. Here’s what the research says about how ChatGPT picks which brands to recommend, and the exact steps to become one of them.

How ChatGPT Actually Decides What to Recommend

Most advice about ChatGPT visibility is vague. “Create great content.” “Build authority.” That doesn’t help anyone. So I dug into the research to find the actual signals ChatGPT uses.

Here’s what the data shows. ChatGPT’s recommendation signals break down into a clear hierarchy:

  • Authoritative list mentions: 41% of recommendation influence. Industry rankings, expert roundups, and curated lists are the single strongest signal.
  • Awards and accreditations: 18%. Institutional recognition from programs like G2 Best Of, industry awards, or accelerator badges.
  • Online reviews: 16%. Third-party review coverage on platforms like G2, Capterra, and Trustpilot.
  • Entity recognition from training data. How often and how consistently your brand appears across the web.
  • Content freshness. 71% of ChatGPT’s citations come from content published between 2023 and 2025.

Traditional SEO signals (backlinks, domain authority, keyword optimization) have near-zero direct influence on ChatGPT recommendations. This is a completely different game.

The Numbers You Need to Know

A study running the same prompts through ChatGPT 100 times each (1,200 total interactions) revealed how the recommendation engine actually behaves:

ChatGPT recommends about 10 brands per response. That’s it. If you’re not in those 10 slots, you don’t exist for that query.

Only 11% of brands achieve dominant status. “Dominant” means appearing in 80% or more of responses for a given query. In competitive categories, that drops to just 7%.

ChatGPT draws from roughly 44 brands per category. But most of them are in the long tail, appearing in less than 20% of responses. In competitive categories, 72% of brands languish in this long tail.

Recommendations vary every time. ChatGPT rarely returns the same brand list in the same order twice. This is why checking once and assuming you know your visibility is misleading. You need to track it over time.

Each platform is different. ChatGPT favors Wikipedia (47.9% of top citation sources). Perplexity heavily cites Reddit (46.7%). Google AI Overviews prefer YouTube (23.3%). Only 11% of domains are cited by both ChatGPT and Perplexity. These are entirely different ecosystems.

Why Most Brands Are Invisible

The reason 77% of brands score near zero on AI visibility comes down to a few common problems.

No presence on authoritative lists. If you’re not on industry roundups or comparison lists, ChatGPT has no signal to pick up. This is the single biggest factor at 41%.

No review coverage. Companies with review scores below 70% are significantly less likely to be recommended. No reviews at all is worse.

Inconsistent brand information. If your brand name, description, or positioning is different across your website, social profiles, and directories, AI models can’t build a clear picture of what you are.

Content that’s not structured for AI. AI systems process content through chunk-level retrieval, breaking pages into sections and evaluating each independently. Content optimized for this approach is 50% more likely to be selected for AI answers.

The 6 Steps That Actually Work

Based on the research, here’s what moves the needle. These are in priority order.

Step 1: Audit Your Current Visibility

Before optimizing anything, you need to know where you stand. Open ChatGPT and ask 10-20 questions your customers would ask. Not keyword phrases, but full questions:

  • “What’s the best [your category] for [your audience]?”
  • “Can you recommend a [product type] for [use case]?”
  • “How does [your brand] compare to [competitor]?”

Run each prompt 3-5 times. ChatGPT varies its output, so a single check doesn’t tell you much. Document which brands appear, how often, and what ChatGPT says about each one.

Then do the same on Claude and Perplexity. Each platform draws from different sources and recommends different brands. CitedBy automates this across all three engines if you want to skip the manual work.

Step 2: Get on Authoritative Lists

This is the highest-leverage action you can take. At 41% of recommendation influence, authoritative list mentions are more than twice as important as any other signal.

What counts as an authoritative list:

  • “Best [category] tools 2026” roundup posts on industry blogs
  • G2, Capterra, or similar platform category pages
  • Expert comparison articles on sites like TechCrunch, HubSpot, or industry-specific publications
  • Curated lists on niche community sites or newsletters

How to get there:

  • Submit your product to relevant review platforms
  • Pitch industry bloggers for inclusion in roundup posts
  • Contribute guest content to publications that write comparison pieces
  • Sponsor or participate in industry awards programs

Step 3: Build Review Coverage

Online reviews account for 16% of ChatGPT’s recommendation signals. ChatGPT draws from review platforms when assembling its recommendations, and brands with strong review profiles surface more consistently.

The minimum bar: you need reviews on at least one major platform (G2, Capterra, Trustpilot, or an industry-specific equivalent). Aim for a score above 70%, because companies below that threshold get recommended significantly less often.

Ask every happy customer for a review. Make it easy. A direct link to your review profile page removes friction.

Step 4: Add Structured Data to Your Site

86% of AI citations come from sources brands already control, according to Yext’s study of 6.8 million citations. Your website is the most important source you control.

Implement these schema types:

  • Organization schema with your brand name, description, logo, and social profiles
  • Product schema with features, pricing, and category
  • FAQ schema on pages where you answer common questions
  • Review schema if you display testimonials on your site

Structured data helps AI models parse your brand information instead of guessing. Pages with structured data see measurable citation improvements within 2-3 weeks.

Step 5: Make Your Brand Information Consistent

AI models cross-reference multiple sources to build a picture of your brand. If your website says one thing, your LinkedIn says another, and your directory listing says something else, that’s noise, not signal.

Audit your brand presence across:

  • Your website (homepage, about page, product pages)
  • Social profiles (LinkedIn, Twitter/X, Facebook)
  • Business directories (Google Business, Crunchbase, AngelList)
  • Partner listings and integrations pages
  • Review platform profiles

Use the same core language to describe what you do. Same brand name, same category, same positioning.

Step 6: Track and Iterate

ChatGPT’s recommendations change over time. A brand that’s recommended today might not be next week. 46% of ChatGPT interactions now use integrated web search, which means fresh content and recent mentions carry increasing weight.

The average time from optimization efforts to measurable results is 89 days. This isn’t a one-time fix. It’s an ongoing process of building visibility, tracking changes, and adjusting based on what the data shows.

What Doesn’t Work

A few things I see brands waste time on:

Stuffing keywords into your site. ChatGPT doesn’t rank pages by keyword density. It evaluates brand authority across the web.

Building backlinks specifically for AI visibility. Backlinks have near-zero influence on ChatGPT recommendations. They help with Google, not AI.

Optimizing for one platform only. ChatGPT, Perplexity, and Google AI Overviews each draw from different sources. Only 11% of domains are cited by both ChatGPT and Perplexity. You need a multi-platform approach.

Checking ChatGPT once and assuming you’re fine. Recommendations vary every single time. Without tracking over time, you’re guessing.

The pattern is clear from the data. Brands that consistently show up in ChatGPT’s recommendations share three traits:

  1. They appear on multiple authoritative lists and comparison articles
  2. They have review coverage with scores above 70%
  3. Their brand information is consistent and structured across the web

None of this requires a massive budget. It requires focused effort on the signals that actually matter, and the discipline to track results over time instead of guessing.

The 77% of brands that are invisible to AI aren’t there because ChatGPT is unfair. They’re there because they haven’t done the work to be found. The good news is that the playbook is straightforward. The brands that start now will have a compounding advantage over those that wait.