Visibility
Whether AI includes your organisation when candidates ask about employers without supplying your name.
ChatGPT employer brand tracking
Candidates are asking ChatGPT which employers they should consider, what companies are like to work for and whether an employer matches their priorities.
BrandScore actively monitors what ChatGPT tells candidates about your employer, whether it names you and which narratives and sources appear in its answers. Follow movement alongside five competing or aspirational employers, rather than relying on one conversation.




NEW
NEWActive monitoring
Active monitoring combines repeated measurement with analysis and findings; detailed recommendations depend on the package. It is more than a dashboard displaying changing numbers.
Whether AI includes your organisation when candidates ask about employers without supplying your name.
How AI describes your culture, careers, benefits and employee experience.
Which employers are named in the measured market and how those appearances change.
Returned citations, plus separate readability and support checks in eligible evidence workflows. Not every citation is claim-verified.
What other employers publish across careers sites, job ads and employer content, and which approaches deserve consideration.
Where your team or agency should focus its next action.
Establish the starting position. Follow movement through seven scans across 90 days. Compare five chosen competing or aspirational employers. Examine AI descriptions, sources and evidence differences. Review the patterns and receive prioritised recommendations after the full 90-day period, where included in your package.
Initial findings follow the first scan, with fortnightly updates after each scan. Consolidated recommendations come after the full 90 days (seven scans), not monthly or after every scan. In the 90-Day Programme, Monitor includes scan results; Discovery adds the final recommendations and advisory review.
Why ongoing measurement matters
Google rankings and paid search positions have always changed. Employers could still control the message published through their adverts and destination pages.
AI adds another source of change. Answers include employers and generate a description for the question being asked. Your website can remain unchanged while your visibility, the sources cited and the competitors appearing beside you change.
An audit establishes your position during the assessment. Active monitoring helps you follow what happens afterwards and decide when to respond.
Across 134 repeat comparisons covering 22 employers, BrandScore recorded average visibility movement of 7.6 percentage points between measurements averaging eight days apart. 22% of comparisons moved by at least 10 percentage points.
Observed movement can reflect changes to AI systems, available evidence, competitors or answer variation. These results do not establish what caused each change. This is a historical four-engine repeated-measurement subset, not the full employer sample or a current six-experience study.
Read the BrandScore audit and monitoring researchTwo views
Measures answer presence, descriptions, favourability, competitors and cited sources across six AI experiences. Being named is not the same as being endorsed.
Examines careers sites, job adverts and other employer-controlled content across the selected employers, highlighting approaches worth considering.
A competitor’s published content does not, by itself, establish why AI includes that employer. Examining both views supports better decisions about your team’s next action.
Active monitoring describes the measurement approach. Monitor and Monitor + Reinforce remain distinct packages: Reinforcement is a separate part of the offer, not a promise of improved visibility or proof that your activity caused a change.
Candidate research now begins in conversation. People ask ChatGPT to identify employers worth looking at, compare two organisations, assess what working somewhere may be like, and recommend employers that fit their own priorities.
Asking ChatGPT about your own organisation once is a reasonable starting point, but it is not a measurement. A single answer is one sample drawn under one set of conditions, and those conditions move.
Repeated, structured measurement across a consistent prompt set separates a persistent pattern from ordinary variation. Screenshots from occasional manual tests cannot do that, and no single ChatGPT answer represents an employer's complete AI position.
For the reasoning behind this, read why self-run AI employer brand checks mislead.
BrandScore runs candidate-style questions repeatedly and records what comes back, so the reported position reflects patterns rather than a single answer.
An employer's visibility does not exist on its own. AI may associate other organisations more strongly with the same candidate priority, role, industry, location or talent market, and those organisations appear in the answers your candidates read.
BrandScore shows how an employer compares with the organisations competing for the same talent. The employers surfacing in open-market candidate answers are chosen by the engine, not selected by the employer being measured.
Candidates use several AI platforms, and answer presence, narratives and sources can differ between them. BrandScore monitors employer visibility and narrative across ChatGPT, Gemini, Claude, Perplexity, Google AI Overviews and Google AI Mode. Evidence Readiness is a separate assessment: ChatGPT, Gemini and Perplexity are scored; Claude is memory-only contrast.
Engines are not treated identically: prompts, scoring and measurement differ by engine and are reported accordingly.
Monitor your employer visibility across leading AI platforms.
BrandScore reports what it observes. It does not change or control what an AI platform answers.
The programme builds a fuller view of an employer's AI position over a defined period, rather than a single check. It combines:
Measuring repeatedly, across engines and against competitors, reduces the risks created by relying on one-off self-testing.
You can run an initial check, but one conversation provides a limited view. Prompt wording, previous conversation context, location, language and answer variation can change the result. Structured measurement across repeated questions and samples provides a more reliable picture.
Answers can vary because of prompt wording, conversation history, location, language, model updates and the sources retrieved for the question. This is why BrandScore measures patterns across repeated candidate-style questions rather than relying on one answer.
BrandScore does not control AI answers. It shows employers how they are represented, whether they are named, which narratives and sources appear in the answers, and where stronger evidence may be needed.
AI answers, models, sources and competitor positions can change. Repeated measurement helps employers distinguish a persistent pattern from normal answer variation and identify meaningful changes over time.
Employer visibility and narrative monitoring covers ChatGPT, Gemini, Claude, Perplexity, Google AI Overviews and Google AI Mode. Evidence Readiness separately scores ChatGPT, Gemini and Perplexity, with Claude as a memory-only contrast, not a scored readiness service. Google AI Overviews and AI Mode are excluded from Readiness.
BrandScore measures which employers appear within relevant candidate-style answers, how frequently they appear, how favourably they are represented and which narratives AI associates with them. This shows how an employer compares with organisations competing for the same talent.
Your BrandScore AI assistant
Where included in your package, Brandi is a conversation with your employer intelligence. Brandi helps you understand the findings, explore the evidence and turn your report into a briefing for your team, leadership or agency.
Ask about AI Answer Presence, employer narratives and talent competitors, with explanations grounded in your report.
Create a bespoke briefing for your head of marketing, HR leadership or executive team, with figures checked against your report.
Download report-based Word documents and CSV files for your team to review and share.
90-Day, Monitor and Monitor + Reinforcement subscribers receive a narrated overview with each report. This example uses fictional data.