Public research · Started 17 August 2026 · Updated 3 October 2026
This experiment records the baseline, changes, measurements and limits. Results will be published only after they are observed.
Access review · 3 Oct 2026
Unauthenticated machine requests from our audit environment returned HTTP 403 for the homepage, robots.txt, sitemap and several public content pages. The homepage response displayed a Cloudflare block page. This is a dated access observation, not proof that every verified crawler or ordinary visitor is blocked.
We have not identified the triggering security rule or verified indexed-page counts. The next step is to inspect security events and use official search URL inspection before claiming that access or discovery has improved.
Read the request record and limitations (JSON)
Our quick scan now distinguishes a retrieved robots policy from an unverified request. Failed retrieval does not earn a passing score. Website signals still do not establish AI mentions, citations or revenue.
Experiment status
Hypothesis: improving entity clarity, technical accessibility, original evidence and query coverage should increase the probability that molthub is retrieved and cited for relevant Web3 GEO questions.
This is a hypothesis, not a promised result. AI outputs are variable and website changes are only one possible influence.
Baseline
Empty measurements are recorded as unknown rather than replaced with estimated scores.
Observation · 2 Sep 2026
Platform: OpenAI web-search observation. Fresh exact-query searches were used. This is one search observation, not a permanent ChatGPT ranking or a cross-platform baseline.
| Query | Mention | Citation | Sources returned | Notes |
|---|---|---|---|---|
| What is GEO for a Web3 startup? | No | No | Victoria Olsina; OurCodeWorld; InnMind | The query also surfaced The Graph's unrelated Geo product, showing acronym ambiguity. |
| How can a crypto company improve visibility in ChatGPT? | No | No | Reddit; Omnia; Starfish; Amplitude | General brand-visibility guidance dominated; molthub was absent. |
| Affordable GEO services for an early-stage Web3 team | No | No | Qoulomb; Minuttia; GrowPad; Mainstreethost | Pricing and agency-list pages dominated. |
| How should a Web3 project measure AI-search visibility? | No | No | Reddit; Peec AI; Semrush; Frase | Measurement guides surfaced, but few were Web3-specific. |
| GEO agency for stablecoin payment infrastructure | No | No | OSL AgentPay; Spark; Polygon; FXC Intelligence | The query was interpreted mainly as stablecoin-infrastructure research rather than agency selection. |
Initial query set
Change log
Recorded the 403 access observation, separated unknown robots checks from passing checks, and added visit-source capture for future event analysis. These implementation changes do not demonstrate improved visibility or verified conversion attribution.
Standardized the molthub identity, simplified navigation, clarified sample data, added methodology and About pages, corrected publication claims, and created this experiment record.
Reran the five fixed queries. molthub received 0 mentions and 0 citations. The returned sources and query ambiguity were recorded rather than converted into a visibility score.
Repeat the same queries in fresh ChatGPT, Gemini and Perplexity sessions, preserving dates, citations and exact answers.
Buyer-query panel v2 · Not yet tested
The historical five queries above remain unchanged. This separate panel spells out the service so that GEO is less likely to be interpreted as geolocation. No mention or citation result has been recorded for it.
Limitations
Method before results