Neutral discovery
Would a buyer discover the company without supplying its name?
“Best skills assessment platform for a high-volume hiring team”
Methodology
We do not count a name drop as a citation.
AI visibility is not one thing. Outloud records what the answer said, which sources it showed, and how the customer and competitors appeared—then reports those signals separately.
Question design
Three prompt groups answer three different questions.
Competitor-led
Does the company appear when a buyer compares known options or asks for alternatives?
“Alternatives to [competitor] for skills-based hiring”
Branded controls
Can the system identify the company accurately when asked directly?
“What does [company] do and who is it for?”
TWO MEASUREMENT SYSTEMS
Organic search and AI answers overlap. Their outcomes should not be blended.
SEO establishes whether pages can be discovered, indexed and found through search. Outloud adds a controlled view of whether those pages and supporting sources are being used inside AI-assisted buyer research.
Search visibility
- Crawl and indexing status
- Search impressions and clicks
- Queries and landing pages
- Technical and content issues
Observed through site checks, Google Search Console and Bing Webmaster Tools where access is available.
AI answer visibility
- Neutral and comparison prompts
- Verified domain citations
- Brand mentions and recommendations
- Competitor and source coverage
Observed through a documented, dated prompt sample with the raw evidence preserved.
Evidence ladder
Visibility becomes more valuable as the evidence strengthens.
These are related signals, not interchangeable labels.
- 01
Retrieval triggered
The platform searched or retrieved external material.
- 02
Source selected
A page or domain appeared in the platform's source set.
- 03
Verified domain citation
A visible citation resolves to the customer's domain.
- 04
Brand mention
The answer names the company, with or without a direct citation.
- 05
Recommendation
The company is presented as an option for the buyer's need.
- 06
Supported claim
The answer makes a relevant claim and provides traceable supporting evidence.
Working loop
Measure, diagnose, improve, repeat.
- 01
Audit
Run a dated, documented prompt set and preserve the raw outputs.
- 02
Analyse
Separate content gaps, entity ambiguity, technical barriers and source gaps.
- 03
Implement
Make the highest-value site and authority improvements first.
- 04
Repeat
Re-run comparable questions and explain what changed.
Important limitations
A snapshot is evidence, not a permanent truth.
- Answers can vary by platform, model, location, date and browsing behaviour.
- A sampled prompt set cannot represent every question a buyer may ask.
- “Share of voice” means share within the documented sample—not the whole market.
- AI visibility cannot be inferred from organic rankings alone, and prompt sampling does not replace SEO reporting.
- A citation can be present without producing a useful or accurate recommendation.
- Improvement work raises the quality of discoverable signals; it cannot guarantee an output.
See your evidence
See the method applied to your company.
Request a free baseline
Or just email jason@outloud.au