ChatGPT

Provider: OpenAI Model family: GPT-4o Search: Uses Bing Search for browsing-enabled queries

How it finds brands

ChatGPT draws on both its training data (a large snapshot of the web) and live Bing search results when browsing is enabled. Brand recommendations often come from a combination of memorised associations and freshly retrieved web content.

How to optimise

Ensure your site is indexed by Bing (Bing Webmaster Tools)
Build brand mentions on publications that Bing indexes prominently (major trade publications, news sites)
Use clear entity signals: your brand name + category should be consistent across your site, Wikipedia presence, and social profiles

Claude

Provider: Anthropic Model family: Claude 3.5+ Search: Uses web search for grounded answers in Projects and API modes

How it finds brands

Claude uses training data plus web search for factual grounding. It tends to be more conservative in recommendations and cites information it can verify. Brand signals from authoritative, well-structured content are especially effective.

How to optimise

Prioritise structured data: Organization, Product, and FAQPage schema
Write clear, factual content — Claude is less likely to recommend brands with vague or marketing-heavy copy
Earn citations from credible third-party sources (reviews, case studies, press)

Gemini

Provider: Google Model family: Gemini 1.5+ Search: Deeply integrated with Google Search index

How it finds brands

Gemini has the strongest integration with Google's search index of any AI assistant. Your Google Business Profile, Google Search Console data, and traditional SEO rankings all feed directly into Gemini's brand awareness.

How to optimise

Maintain a complete, verified Google Business Profile
Fix all Google Search Console errors — crawlability issues directly affect Gemini's knowledge of your brand
Target featured snippet positions (Gemini frequently pulls from these)
Schema markup is especially important here given Google's investment in Knowledge Graph

Perplexity

Provider: Perplexity AI Model family: Various (Claude, GPT-4, internal models) Search: Uses its own web index with real-time retrieval

How it finds brands

Perplexity is the most citation-heavy of the four platforms — it explicitly lists the URLs it retrieved and summarised. A mention in Perplexity often comes with a direct link to your site, making it high-value.

How to optimise

Ensure your site is crawlable and fast (Perplexity crawls frequently)
Write comprehensive, well-structured pages on topics in your niche — Perplexity favours depth
Get mentioned on sites Perplexity trusts (Reddit, major publications, Quora for consumer topics)
Use robots.txt to explicitly allow Perplexity's crawler: PerplexityBot

Crawler access

All four platforms use their own crawlers to retrieve live web content. Make sure your robots.txt does not block them:

# Allow all AI assistants
User-agent: GPTBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: Google-Extended
Allow: /

User-agent: PerplexityBot
Allow: /

Surfaceable's indexability check flags if any of these crawlers are blocked.