Your Brand Might Be Invisible to AI Search

 1|Your Brand Might Be Invisible to AI Search
 2|
 3|Last week I helped a friend check his SaaS product's presence. He ranks on page one of Google. Two years of solid SEO work.
 4|
 5|Then I asked ChatGPT: "What's a good tool for [his category]?"
 6|
 7|His product wasn't mentioned. Not once. All competitors.
 8|
 9|He couldn't believe it. "I've been doing SEO for two years."
10|
11|Here's the thing — AI doesn't use Google's ranking algorithm. It pulls from training data and real-time crawls. It looks for structured Q&A content, authoritative third-party mentions, comparison contexts. Not your keyword density or backlink count.
12|
13|A brand can rank #1 on Google and score 0% on AI discovery.
14|
15|I ran into this problem with my own product. So I built Anchor.
16|
17|Enter your brand name and a category keyword. Anchor queries ChatGPT, Claude, Gemini, DeepSeek, and Kimi simultaneously. It simulates real user questions — recommendations, comparisons, pain points, trust signals — covering every scenario.
18|
19|3 minutes later you get a report:
20|- 0-100 visibility score
21|- Per-model breakdown — which AI knows you, which doesn't
22|- Who your AI competitors are
23|- Which query scenarios you show up in, which you're missing from
24|- Actionable optimization roadmap
25|
26|Save it to your brand library. Run it again after you've made changes. Track whether your GEO efforts actually moved the score.
27|
28|SEO tells you where you rank on Google.
29|Anchor tells you where you rank in AI answers.
30|These are two completely different things.
31|
32|anchor.polis.ink — free, no signup, 3-minute report.

Comments

[toshihiro shishido]

@jenodebug the "rank #1 on Google, 0% on AI" framing is exactly what most teams haven't internalized yet. SEO and GEO (generative engine optimization) are starting to look like two different jobs.

A practical observation from running this for a few SaaS clients:

1. AI visibility moves on a different cadence from Google rankings. ChatGPT and Claude appear to refresh their training-blended answers on different cycles — sometimes you'll see a brand show up in Gemini and DeepSeek months before it surfaces in ChatGPT, or vice versa. So a single per-model breakdown like Anchor offers is more useful than people realize.
2. Comparison context matters more than direct mentions. "Best tool for X" queries tend to surface brands that get co-mentioned with established competitors in third-party content, even if the brand itself has lower domain authority.
3. Forum / community presence (Reddit, dev.to) seems to weight heavier in AI training corpora than traditional SEO ranks. So the channels you'd dismiss as "low DR" can be exactly what moves AI visibility.

Curious whether you're seeing the per-model variance widen or narrow over time in Anchor's data — feels like an interesting longitudinal signal.