The agency

A more careful way to investigate AI search visibility.

MentionGauge is an independent AI visibility consultancy focused on a simple principle: businesses deserve to see the evidence behind any optimization recommendation.

Why MentionGauge exists

AI search advice needs inspectable evidence.

When a buyer asks an AI assistant for a provider, the answer may mention brands, recommend alternatives or cite webpages. Those are different observations, and none alone proves a lead or an attributable sale.

MentionGauge focuses on human-reviewed research and practical website improvement. We investigate the questions customers ask, preserve observed responses, open the sources that are actually displayed and compare those findings with accessible business information.

Our approach is deliberately different from claiming that a single proprietary score reveals a universal AI rank. Responses vary with the question, platform, interface, language, time and available sources. We cannot see a third-party model's private reasoning, and we do not sell guaranteed recommendations.

We aim to give clients usable decisions: what the evidence supports, which site or content defects are independently verifiable, what remains uncertain and what to measure after a change.

Our working principles

  • Real buyer questions before vanity prompts
  • Human review of named brands and linked sources
  • Original and independently verifiable business facts
  • Technical SEO and editorial clarity before shortcuts
  • Explicit limitations, test dates and sampling scope
  • Responsible handling of enquiry and research data
What we actually provide

A diagnostic-first service, with separate implementation and monitoring work.

We distinguish the initial investigation from later changes, so clients know what they are requesting and what they will receive.

Diagnose

Manual AI visibility audits

Define relevant commercial questions, observe sample answers and compare mentions, recommendations, cited sources and competing providers. Findings distinguish confirmed problems from plausible explanations.

Review audit scope
Improve

AI search, GEO and AEO support

Investigate crawlability, structure, service clarity, answer quality and legitimate source corroboration. Each workstream has its own scope and acceptance criteria.

Review implementation services
Recheck

Comparable observation panels

Repeat documented buyer questions, inspect changes in citations or competitor mentions and report what is observed without treating one response as model-wide market share.

Review monitoring
Our research standards

How a human-led investigation is carried out.

A useful report should allow another reviewer to follow the question, the observed evidence and the proposed next action.

  1. Define the actual buyer decision

    Choose target services, markets, real discovery and comparison questions, and the AI interfaces that are relevant to the business.

  2. Capture what was returned

    Record dated observations with context, including whether the brand was mentioned, recommended, cited or absent.

  3. Verify the sources and the website

    Open displayed citations, check relevant facts and inspect site access, important pages and content clarity for verifiable defects.

  4. Separate findings from hypotheses

    Distinguish confirmed broken links or inaccurate descriptions from an unproven theory about why an AI answer selected a provider.

  5. Deliver priorities and repeatability

    Document useful changes, responsible owners and a fair comparison method for later checks, including uncertainty and attribution limits.

What clients can examine

The research methodology is published openly. It explains sample design, source validation, observation logs, reviewer checks, evidence limits and handoff requirements.

See our detailed methodology and how to measure AI visibility for definitions and examples. For a lightweight DIY approach, see the free checking guide.

Transparency

What we do not claim.

Credible AI visibility work is specific about its boundaries, particularly when systems are outside an agency's control.

No secret ranking access

We do not claim privileged access to ChatGPT, Claude, Gemini, Perplexity or another platform's private recommendation algorithm.

No guaranteed placements

We do not promise a fixed AI recommendation position, a universal score, citation inclusion or an automatic increase in enquiries.

No invented proof

Illustrative demonstrations are identified as examples. We do not present fictional clients, testimonials, case outcomes or affiliations as verified achievements.

Third-party AI platform names on our website identify experiences we may investigate; they do not indicate partnership, endorsement or privileged access. Project scope depends on tool availability, customer requirements and approved research conditions.

Working with MentionGauge

What to expect before an engagement begins.

Start by sharing your public website, the services you provide, your target market and a few buyer questions. We use that information to discuss a research scope and clarify what information is needed.

Agree on a scope and boundaries

Before paid work, confirm the platforms, buyer-question sample, markets, review depth, deliverables, responsibilities, access permissions and any work explicitly excluded. If implementation or recurring monitoring is required, define it separately.

Keep your information under your control

A first enquiry should not contain passwords, confidential client records or sensitive credentials. Additional information, if required, should be requested through an agreed process. See our privacy notice for the site's stated data practices.

Evaluate us against alternatives

Use the same evidence standard for every supplier. Our agency selection checklist explains the questions to ask about raw observations, deliverables, price and uncertainty.

Need a specific answer?

Tell us which customers you serve and where AI-assisted discovery appears relevant to their decision. We can discuss whether a scoped manual audit, an implementation review or a repeat-test project is the appropriate next step.

Contact MentionGauge about a scoped project
Start with the evidence

Know what AI answers actually show.

Request a scoped manual audit to understand how your business appears across relevant AI-assisted search experiences and what can be improved.