AI Visibility Analysis

AI Visibility Analysis turns completed monitoring evidence into a saved report with prioritised actions. Use it to understand why visibility differs by audience, platform or journey stage, how competitors are performing, what evidence AI platforms cite, and what to improve next.


The analysis reviews monitoring results already held by Wheremore. It does not run the selected monitoring questions again.


Before you begin

You need:


  • an AI Visibility project with at least one name saved under Your Brands;
  • completed, processed monitoring results in the selected period; and
  • enough eligible conversations for the chosen persona.

A persona needs at least three eligible conversations in the selected scope. All personas can analyse up to 12 eligible personas. Wheremore needs at least two completed persona reports to create the final cross-persona report.


Open the analysis drawer

  1. Open AI Visibility Dashboard.
  2. Select the project and dashboard scope you want to investigate.
  3. Select Analyse with AI.

The AI Visibility drawer has two tabs:


  • New analysis sets the scope and starts a report.
  • Recent lists saved reports for the selected project.

Set the analysis scope

On New analysis, review:


Choice What it changes
Project Selects the project and its available monitoring evidence.
Persona Analyses one configuration or all eligible personas.
Platform Uses all eligible platforms or one selected platform.
Journey stage Uses all stages or Awareness, Consideration or Decision.
Provider depth Includes Your Brands anywhere in the provider list, in the top three, or at number one.
Period Uses a preset period or a custom range of complete UTC days.

Preset periods end yesterday, so the analysis does not use a partially completed current day. A custom period can contain up to 180 complete days and cannot include today or a future date.


This date behaviour differs from the dashboard charts, whose preset periods can include today. Match the displayed dates when comparing a saved report with the dashboard.


Choose Standard or Enhanced

Choose the analysis mode before starting:


  • Standard analyses the available monitoring data and saved evidence.
  • Enhanced first creates the complete Standard analysis, then investigates up to two useful unanswered questions using public evidence and refines the recommendations. Additional credits apply.

For All personas, Enhanced work can apply separately to each eligible persona, so the extra credit use can increase with the number of personas.


Enhanced analysis is designed to resolve specific decisions that the monitoring evidence cannot answer on its own. It does not rerun Claims Verification, create content, or treat public research as proof of how an AI platform ranks providers. If no useful research question is available, or optional research cannot be completed, Wheremore retains the Standard-quality report.


Run the analysis

  1. Select one persona or All personas.
  2. Choose the platform, journey stage, provider depth and period.
  3. Select Standard or Enhanced.
  4. Select Analyse persona or Analyse all personas.
  5. Follow the progress in the drawer or close it and continue working elsewhere.

The progress display can include:


  • Preparing selected evidence
  • Analysing persona evidence
  • Looking into useful questions
  • Refining or building recommendations
  • Creating overall comparison
  • Saving reports

Processing continues on the server if you close the drawer or leave the page. Reopening the drawer can reconnect to an active operation even when the browser no longer has its earlier local state. A recovery process also checks unfinished work in the background.


The browser checks for progress for up to 30 minutes. If it stops checking, the analysis remains on the server. Reload the page and reopen Analyse with AI to continue checking it. Do not start another analysis simply because the drawer was closed.


Understand a persona report

A completed persona report has Analysis and Evidence tabs. Analysis opens with a headline and Recommended actions. Evidence shows the measured scope, counts and supporting tables. The measures can include Mention Rate, Share of Voice, Pooled Competitive Win Rate, Average Provider Rank and Weighted Citation Rate for the selected evidence.


Each recommended action can explain:


  • its type and priority;
  • what the evidence shows;
  • the recommended change;
  • why it matters;
  • supporting evidence; and
  • what completion should look like.

Open Supporting analysis for the detailed evidence. Its sections can include:


  • Performance by Platform
  • Competitive Win Rate by Competitor
  • Tracked USP Visibility
  • Claims and Evidence
  • Sentiment distribution, platform and journey-stage findings
  • Platform-Specific Gaps
  • Citation Source Ownership
  • Test Design Assessment
  • Questions Needing Revision
  • Systemic Issues
  • Competitive Insight

Sections appear only when suitable evidence is available.


Understand an all-personas report

Wheremore first produces a report for each eligible persona and then creates a Cross-Persona Strategic Analysis from at least two completed persona reports.


The combined report highlights actions and patterns shared across personas. Supporting analysis can include:


  • Tracked USP Visibility Across Personas
  • Claims and Evidence Across Personas
  • Platform-Specific Gaps
  • Citation Source Ownership
  • Visibility Locks Detected
  • Primary Positioning Recommendation
  • Jobs-To-Be-Done Patterns
  • Adjacency Risks
  • Universal Issues
  • Cross-Persona Competitor Analysis

If fewer than two persona reports complete, Wheremore cannot create the cross-persona report. Review the completed persona report or change the scope so that more personas have enough evidence.


Claims and Evidence

The analysis can use the latest published Claims Verification inventory that matches the selected answer, persona, platform and date scope. It presents two separate questions:


  • Is this claim accurate?
  • Do the pages cited in the AI answers support it?

These are different judgements. A claim can be accurate while a cited page does not support it, or a page can support wording that still needs qualification.


AI Visibility Analysis reads the saved verification evidence. It does not independently verify the claim again and cannot change a verification verdict. Where available, the report can show supporting quotations and links so you can inspect the evidence.


If claims evidence is unavailable, the report can explain why, for example because there is no published Claims inventory or the selected answers and personas are outside the latest inventory. Use AI Claims Verification to create or update the underlying inventory.


Tracked USP Visibility

Tracked unique selling propositions (USPs) are assessed separately from verified claims. The report can show whether monitored answers contain the ideas or differentiators you chose to track.


A tracked USP appearing in an answer does not prove that it is factually correct. Claims Verification supplies the separate accuracy and citation-support evidence.


Interpret citation findings carefully

Weighted Citation Rate and Citation Source Ownership describe the citations attached to qualifying monitored answers. They do not prove that a citation supports a particular claim, endorses the brand, or caused the provider position.


Resolved destinations can be grouped as owned or third-party evidence. An unresolved source is not silently treated as third-party. Use Claims and Evidence when you need the separate claim-support judgement.


Add a recommendation to the Action Board

Select Add to Board on a recommendation to create a task linked to the saved analysis and its evidence scope. The task can retain the action, rationale, priority and a measurable success signal where the available data supports one.


Some aggregate outcomes, such as visibility, citation rate or a platform gap, can be measured again. A detailed recommendation about a particular message, claim or tracked USP may not have an automatic outcome measure. A later metric improvement also does not prove that the task caused the change.


Use a saved report

Open Recent to return to completed reports. It can show up to 100 reports for the project; selecting a persona narrows the list. Wheremore retains the latest five reports for each exact persona scope and the latest five matching cross-persona reports.


A report card identifies the completion time, report type and whether Enhanced analysis was used. Open a report to:


  • review the saved result;
  • discuss it in Chat;
  • Export report as a text file;
  • add recommendations to the Action Board; or
  • delete the saved report.

Deleting a report does not delete Action Board tasks already created from it. Older reports removed by retention are no longer listed, while their billing and existing task records remain.


Saved history preserves the displayed report and its structured result. It is not a store of raw prompts, all source records or supplier receipts. Access follows the viewer's current project and workspace permissions, rather than the permissions of the person who originally ran the report.


If something does not look right

No personas are available

  • Confirm the project has monitoring configurations and processed results.
  • Expand the period or choose broader platform and journey-stage filters.
  • Confirm at least one name is saved under Your Brands.
  • A persona needs at least three eligible conversations in the selected scope.

The analysis is still processing

You can close the drawer or leave the page. Work continues in the background. Reopen the drawer or reload the page to reconnect to the active operation. The browser stops checking after 30 minutes, but that does not delete or cancel the server-side work.


Claims and Evidence is missing

  • Confirm a Claims Verification inventory has been published.
  • Confirm it covers the selected answers, personas, platforms and dates.
  • Check whether the report explains that the latest inventory did not include the selected evidence.

The report differs from a dashboard chart

Compare the project, persona, platform, journey stage, provider depth and exact UTC dates. Remember that Analysis preset periods end yesterday, while dashboard presets can include today. Also check the metric definition: several dashboard modules use different populations and denominators.


A saved report is no longer in Recent

Only the latest five reports for an exact scope are retained. It may also have been deleted or become inaccessible after a project or workspace permission change.


Enhanced produced no extra research findings

Enhanced keeps the Standard report when there is no useful unresolved question or optional public research cannot improve the recommendation. It does not invent a research step merely to make the report longer.


Use AI Visibility Dashboard for current charts and tables, AI Claims Verification to create the claims evidence used by the report, AI Visibility Settings to correct brand and competitor grouping, and Manage AI Visibility Tracking to change audiences, questions, platforms or processing days.

Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.