How ChatGPT decides what to recommend
Unlike Google, ChatGPT doesn't crawl the web in real time to answer most queries. It draws on patterns learned during training - the sum of what high-quality web content said about a brand across millions of pages. Different rules apply.
Training data drives recommendations
ChatGPT recommends brands that appear frequently in high-quality, authoritative training data - not brands that pay for placement or optimize their own site for keywords.
Third-party sources matter most
What others say about a brand carries far more weight than what the brand says about itself. A Wirecutter review is worth more than 100 blog posts from the brand.
Updates take months
ChatGPT's training data is updated periodically, not in real time. Tactics that work today will show measurable impact in 3–6 months.
Entity clarity is foundational
ChatGPT recommends brands it clearly understands. If it can't consistently describe what a brand does and for whom, it won't recommend it confidently.
ChatGPT SEO tactics that actually work
Earn inclusion in "best of" articles in recognized publications
Highest impact"Best [category] tools for [use case]" articles in TechCrunch, Forbes, Wirecutter, G2's blog, and vertical trade publications are weighted heavily in ChatGPT training. When multiple high-authority sources include your brand, ChatGPT learns to recommend it.
Timeline: 3–6 months to show in ChatGPT
Build a Wikipedia article (if notability criteria are met)
Very high impactWikipedia is one of the most heavily represented sources in LLM training data. A well-cited Wikipedia article about your brand significantly improves ChatGPT's understanding and recommendation frequency.
Timeline: 3–6 months after publication
Maximize review volume on G2, Trustpilot, and Capterra
High impactThese platforms are high-authority sources for product recommendations. Strong review profiles - particularly on G2 for software - correlate with higher ChatGPT mention rates in relevant categories.
Timeline: 2–4 months
Earn expert media coverage and author bylines
High impactFounder profiles, executive bylines, and expert source quotes in recognized media build topical authority for the brand. ChatGPT is more likely to recommend brands that are consistently cited as authoritative voices.
Timeline: 3–6 months
Establish a clear, consistent brand entity
Medium-high impactConsistent brand name, description, and category positioning across the web - About page, Wikidata, Google Knowledge Panel, social profiles - helps ChatGPT understand exactly what the brand is, which increases recommendation confidence.
Timeline: 1–3 months
What doesn't work for ChatGPT SEO
✗ Publishing blog posts on your site
First-party content has minimal direct impact. ChatGPT can't index your blog in real time.
✗ Keyword optimization
ChatGPT doesn't rank websites. Keyword-stuffed content won't influence its recommendations.
✗ Social media activity
Twitter/LinkedIn activity doesn't directly feed ChatGPT recommendations.
✗ Technical SEO
Site speed and Core Web Vitals are irrelevant for ChatGPT. They matter for Google, not LLMs.
How to measure ChatGPT SEO progress
The primary metric is mention rate - how often your brand appears across a defined set of relevant queries. Track it across 20–50 queries representative of how customers search in your category. Run the same query set monthly to measure change.
Patience matters here. Because ChatGPT updates are slow, set realistic expectations: early ChatGPT SEO work often shows results in Perplexity first (weeks), then in ChatGPT later (months). Tracking both separately helps show clients that the work is moving the needle even before ChatGPT fully reflects it.
Measure your ChatGPT mention rate
ArtificialPulse tracks your ChatGPT mention rate across a custom query set - so you can measure what's working.