# GeoXylia
> We ran our own GEO audit on geoxylia.com: 50/100, below the 476-site median of 75. The brand gap we measure applies to us too. Honest breakdown.
← Blog/Measurement & Benchmarks/We Audited Our Own Site With Our Own Engine, and Scored Below the MedianMeasurement & BenchmarksIntermediate

## We Audited Our Own Site With Our Own Engine, and Scored Below the Median

GeoXylia's own engine gives geoxylia.com a 50/100, below the 75.0 median of the 476 sites we benchmarked. The diagnosis confirms our benchmark's headline finding: the brand gap is real, and it applies to us too.

Ethan Lim2026-08-039 min

Last updated: 2026-08-15

Share:

Measurement & Benchmarks hubExplore all Measurement & Benchmarks guides →

GeoXylia's own audit engine scores geoxylia.com at 50/100, below the 75.0 median of the 476 sites in our benchmark. The homepage scores 50 (D), the /audit tool page 46 (F), and this benchmark post 54 (D). The diagnosis our engine returns is exactly the headline finding of our dataset: the brand gap is one of the widest gaps in GEO, and it applies to us too.

We publish this for the same reason we publish the full benchmark dataset: radical transparency is one of our stated values, and a tool that audits websites should survive its own audit. It does, barely. Here is the full, honest breakdown.

## What scores did our own engine give geoxylia.com?

PageOverallGradeBrandSocial proofExpert authLLMOTechnicalHomepage50D34493782/audit (tool page)46F34483072Benchmark post54D344103997Dataset median (n=476)75.0-78-—6694

The dataset is available in full as a public CSV, and this audit ran on the same engine that produced it, so the comparison is like-for-like, with the methodology caveats below.

## What the engine actually found

1. The brand gap, demonstrated on ourselves. Brand signals score 34/100 on all three pages, the same dimension where the top decile of our dataset scores 93 against 65 for the bottom half. Our benchmark called this one of the widest gaps in GEO; our own site is living proof. The brand dimension measures unlinked mentions, citation context, and entity consistency across the web, and we have not yet built the third-party layer (Wikidata, Crunchbase, company profiles, mentions) that feeds it.

2. Zero social proof (4/100). No third-party reviews, ratings, or social validation signals detected. This is the 89%-earned-media reality (Ahrefs, 2026) from our research applied to us: citations flow to brands the web already vouches for, and we have not earned that vouching yet. The dataset backs this up from the other direction: branded web mentions correlate 0.664 with AI visibility, roughly three times the correlation of backlinks (Ahrefs), and our own site has no mention layer to speak of.

3. Weak author-credential signals (8-10/100). Every post has a byline, but the engine measures credential depth: linked professional profiles, verifiable expertise markers, and we score in single digits. We are fixing this with the same entity work we prescribe to clients.

4. Statistics without attribution (48-54%; GeoXylia audit 2026). Across all three pages, roughly half of the statistics in our content lack an inline source. Our own Princeton-aligned checks flag this as a warning: unattributed statistics do not earn the citation lift that attributed ones do. The irony is not lost on us; this is the exact pattern we tell clients to fix.

5. Keyword density creeping on /audit. The term "geoxylia" appears 3.3% of body content (our audit engine reading, 2026) on our tool page (93 times in 2,849 words), the engine flags this as approaching stuffing territory. Marketing pages repeat the brand name; the engine treats repetition as a signal problem, not a branding win.

6. The citation dimension is 0 by design. We run the audit in heuristic mode: without live API keys for citation sampling, the citation dimension reports 0 rather than guessing. We would rather show an honest zero than a fabricated estimate.

## What this confirms about the benchmark

Our dataset's three structural findings: the 28-point brand gap, the small vertical spread, and the dominance of technical and content fundamentals, are all visible in our own audit:

- Technical and content fundamentals are strong: technical 72-97, content 49-63, claim quality 100 on all pages. The architecture work we did (server rendering, schema, answer-first formatting) shows up exactly where the dataset says it should.
- 
- The brand layer is the ceiling: brand 34, social proof 4, expert auth 9. This is the onsite-only ceiling we document in the benchmark, the layer that separates the top decile from the median cannot be built on-page.
- 
- Freshness works: the benchmark post scores 82 on freshness, the highest of our three pages, because it carries visible, honest dates.
- 
- The correlations hold on ourselves: in the dataset, citability correlates 0.961 with the overall score, technical foundation 0.944, and content quality 0.937. Our own audit reproduces the pattern: the pages with the strongest technical and content scores are the highest overall, and the page with the weakest (the 
## Links
- [GXGeoXylia](/)
- [Features](/features)
- [Pricing](/pricing)
- [Methodology](/methodology)
- [Blog](/blog)
- [FAQ](/faq)
- [About](/about)
- [Free Audit](/audit)
- [Measurement & Benchmarks](/blog/geo-measurement)
- [public CSV](/data/geoxylia-benchmark-476-2026-08.csv)
- [Ahrefs](https://ahrefs.com/blog/ai-overview-brand-correlation/)
- [Princeton-aligned checks](https://arxiv.org/abs/2311.09735)
- [free, 60 seconds, 9 dimensions](https://geoxylia.com/audit)
- [476-site benchmark](/blog/state-of-geo-2026-benchmark-476-sites)
- [Benchmark methodology](/methodology#benchmark)
- [The Missing Signal: What AI Engines See That Google Doesnt8 min](/blog/geo-saas-missing-signal)
- [The 38% Visibility Gap: Why Your SEO Traffic Is Disappearing Into AI Search11 mi](/blog/the-38-visibility-gap)
- [Building Your Brand's Knowledge Graph From Scratch: The 2026 Enterprise Pla](/blog/building-your-brand-s-knowledge-graph-from-scratch)
- [What Is GEO? The Definitive Guide to Generative Engine Optimization (2026)20 min](/blog/what-is-geo-2026-definitive)
- [Zhenliang Lim — Founder on LinkedIn](https://www.linkedin.com/in/zhenlianglim)
---
Generated by [GeoXylia](https://geoxylia.com) — AI Visibility Platform