We Audited Our Own AI Visibility. The Score Was 10/100.

In September 2026, we did something uncomfortable.

We took the same GEO audit methodology we use for clients — five real buyer search intents × multiple AI engines — and ran it on ourselves.

Vibe Lingan sells AI visibility solutions for cross-border brands. We help Chinese manufacturers get found and cited by AI engines like ChatGPT, Perplexity, Gemini, and DeepSeek.

Our own score? 10 out of 100.

What "10/100" actually means

It means this:

• When someone searches our exact brand name — "Vibe Lingan" — we show up. Congratulations. So does anyone with a domain name.
• When someone searches a category query — "best GEO service providers for Chinese brands exporting to Europe" — we are completely absent.
• When AI engines generate answers about our industry, our brand has been cited as a source exactly zero times.

The three-state breakdown:
- Source (cited as evidence by AI): 0%
- Mentioned (named but not substantiated): 10%
- Missing (competitor gets the recommendation instead): 90%

For a company that tells clients "AI visibility is your new front door," this was the equivalent of a locksmith locking themselves out of their own shop.

Why we published this

We could have quietly fixed it. Instead, we decided to document the entire recovery — publicly, every 30 days.

Here's why.

Every day, we tell brands: "Your buyers have already moved to AI search. If you aren't visible there, you're invisible to the people who matter most." If that statement is true, it should be provable — starting with ourselves.

So we're turning ourselves into a public case study. GEO Score 10 → ?. Category visibility 0% → ?. AI citations 0 → ?.

Every 30 days, we'll publish: what we changed, what worked, what didn't, and the updated numbers.

What we found in the diagnosis

Four root causes — all fixable:

1. Our naked domain doesn't resolve. vibelingan.com has no DNS record. Only www.vibelingan.com works. AI crawlers treat these as two separate entities.

2. No llms.txt. AI crawlers have no "manual" to read about who we are. They're guessing.

3. Zero presence on the sources AI engines actually cite. Baidu Baike (Chinese Wikipedia equivalent)? Blank. Zhihu (China's Quora)? Blank. WeChat official account? Blank. Deep pages on our own site? Minimal. These four sources are where Chinese AI engines pull their citations from — and we have nothing on any of them.

4. Our core product — Dinghai — doesn't exist in structured data. The relationship "Vibe Lingan = company, Dinghai = product" is understood by humans but invisible to machines. No Product schema. No SoftwareApplication type. AI engines don't know Dinghai is a thing we built.

The 90-day plan

We're not guessing. We're executing against our own methodology:

• Days 1–30 (Foundation): Fix DNS, add llms.txt, deploy Product schema for Dinghai, add hreflang tags
• Days 31–60 (Source building): Launch Zhihu institutional account, WeChat column, LinkedIn presence — create citable content on the sources AI engines actually read
• Days 61–90 (Scale + remeasure): Publish video content, expand distribution, re-run the original 5-intent audit

Why this matters for your brand

If a GEO company scores 10/100 — what do you think a furniture manufacturer, an electronics exporter, or a SaaS company scores?

Probably lower. Because at least we know what to measure.

The uncomfortable truth of 2026: your B2B buyers in Europe are already using AI engines for supplier discovery. They ask ChatGPT and Perplexity about manufacturers in your category. If your brand isn't in the answer — you were never in the running.

The good news: this is fixable. And we're about to prove it.

Follow the experiment

We'll post update #2 in 30 days.

If you want to know your own score — the same five-intent audit we ran on ourselves — that's Door 1. 48 hours. Your data. No commitment beyond that.

Dinghai by Vibe Lingan. Shanghai · Poland · Germany. vibelingan.com