Worked sample · illustration-1

Inspect the evidence before the recommendation.

A three-part teaching example of a Visibility Preview. Every record below is invented so the counting and interpretation can be checked without implying a client outcome.

Entirely synthetic teaching example. No live platform was queried. No client, performance result, market estimate or causal claim is represented.

Sample page 1

Scope and transparent counts

Business: Example Business (fictional)

Three fictional buyer questions, two fictional platforms, two runs each. All twelve records are valid by construction; no failures or exclusions. This is not a market sample.

The platform labels are fictional. There are no real collection dates, source URLs, transcripts or customer locations behind this demonstration.

Brand mentions

6 of 12

One mention flag per response, even when a response could name the business more than once.

Explicit recommendations

4 of 12

Counted separately from mentions. Recommendation flags imply a mention in this example.

Owned citations

3 of 12

Responses flagged as citing an owned source. This is response coverage, not the total number of links.

The fictional peer has 8 mention-bearing responses. The example business accounts for 6 of 14 counted brand-response appearances (42.9%). This is a share within two fictional brands and twelve invented records—not market share.

Read the denominator and aggregation notes →

Sample page 2

All twelve coded evidence records

These are synthetic coded records, not transcripts of real AI outputs. Every row is available in the downloadable JSON. No failed observations are concealed.

Q1

Which providers offer the service in Example Market?

Q2

Which providers explain how this service is delivered?

Q3

Which providers should a buyer compare for this service?

Invented observation set — all records
IDQuestionPlatform / runMentionRecommendationOwned citationPeer mention
01q1Fictional A / 1YesYesYesYes
02q1Fictional A / 2YesYesNoYes
03q1Fictional B / 1YesNoYesYes
04q1Fictional B / 2NoNoNoYes
05q2Fictional A / 1YesYesYesYes
06q2Fictional A / 2YesNoNoYes
07q2Fictional B / 1YesYesNoYes
08q2Fictional B / 2NoNoNoNo
09q3Fictional A / 1NoNoNoYes
10q3Fictional A / 2NoNoNoNo
11q3Fictional B / 1NoNoNoNo
12q3Fictional B / 2NoNoNoNo

Sample page 3

Interpretation, uncertainty and the next action

Observation

The fictional business appears in six records and in none of the four q3 comparison records. The peer appears in one of those q3 records. The totals alone do not explain why.

Interpretation to investigate

Review whether the comparison question matches the actual offer and whether relevant commercial information is available. The invented rows do not establish a technical, content or authority cause.

Proposed next step

In a real engagement, an analyst would review the relevant comparison pages and observed sources with the business owner. The output would be a specific page change or a better-defined follow-up test, not an automatic retainer recommendation.

Ownership and review

The business confirms the facts. An agreed content or development owner makes approved changes. The analyst repeats the relevant observations under documented conditions. This demonstration includes no before-and-after result.

Limitations: entirely invented data; two fictional platforms; three questions; no real customer demand weights; no claim of statistical representativeness, causation, Google rankings or revenue. It illustrates the structure of a preview, not the depth of a paid competitive report.