Total mentions
423Within the selected period.
AI Visibility
What question does it answer?
Frequent mentions do not necessarily mean a favourable brand image. Sentiment analysis highlights positive, neutral, and negative contexts for closer review.
Real project · Intren – SEO
How does AI portray your brand?
The Sentiment dashboard combines an overall picture with individual answer evaluations and recurring phrases. Inspect broad sentiment alongside the actual wording used.
Positive mentions
34581.6% of mentions.
Neutral mentions
7818.4% of mentions.
Negative mentions
00% of mentions.
Sentiment score
90.8%0 means negative, 50 neutral and 100 positive.
The 90.8% sentiment score is different from the proportion of positive mentions, which is 81.6%. It represents aggregate sentiment mapped onto a 0–100 scale.
Fixed project figures from the supplied admin screenshots, not a live report. Mentions and phrases are separate data scopes. The original interface is in Hungarian.
01
Summary, daily trend and sentiment distribution
The own-brand view shows five key metrics, daily changes and the positive, neutral and negative distribution. Filter by period, AI models, tags, sentiment and prompts. Separate tabs cover own-brand and competitor sentiment, phrases and topics.
Open the full-size screenshot
02
Classifications and written evaluations for each answer
For the same question about services, the ChatGPT, Gemini API, Google AI Mode and AI Overview examples are positive. The OpenAI API example is neutral: it lists services without explicitly evaluating them. Each row includes timestamp, model, prompt and written reasoning, with accessible answer details.
Open the full-size screenshot
03
Phrases and topics: recurring wording in answers
Of 2,502 phrases, 1,824 are positive (72.9%), 665 neutral (26.6%) and 13 negative (0.5%). “Data-driven” appears 270 times and “full service” 105 times among positive topics. The risk-related list includes “not necessarily” and other cautious wording; review their meaning in the original answers.
Open the full-size screenshot
A phrase classified as negative does not by itself establish a negative brand image. Topic lists guide investigation; check important conclusions against the saved AI answer and its written evaluation.
What do you see?
- Total, positive, neutral and negative mentions, their proportions and a sentiment score.
- Daily sentiment trends and classification distribution.
- AI answers and evaluations with timestamp, model, prompt, classification, written reasoning and expandable details.
- Own phrases and topics: positive, neutral and negative phrase counts and recurring positive, neutral and risk-related topics.
- Separate own-brand and competitor sentiment and phrase views; period, model, tag, sentiment and prompt filters, plus export.
How do you use it to decide?
- Select the period, models, tags, sentiment and prompt.
- Read the own-brand score alongside the distribution and daily trend.
- Review the written evaluations in the AI answer list, then open relevant answer details.
- Use the phrases and topics view to identify recurring attributes and risk-related wording; switch to competitor views when needed.
Positive overall sentiment can include risk-related phrases
The Intren – SEO example contains 423 mentions: 345 positive, 78 neutral and none negative. Its sentiment score is 90.8%, while positive mentions account for 81.6%. The separate phrase view includes 13 negative phrases among 2,502 phrases: whole-answer sentiment and individual phrase classifications describe different data scopes.
Fixed project figures from the supplied screenshots for September 13–October 13, 2026. Review automated evaluations in the context of the original answers.
What does measurement require?
Saved, deeply processed answers are required. Sentiment is a machine interpretation: review important conclusions alongside the original text.
Frequently asked questions
Does a score of 72 mean 72% positive answers?
No. It is the average sentiment score mapped from the -1 to +1 scale onto a 0–100 scale.
Is this customer satisfaction research?
It describes sentiment about a brand in measured AI answers. It does not replace customer research.
Should I act on one negative mention?
Review text, topic, and frequency first. An isolated uncertain classification differs from a specific claim repeated across platforms.
Can an answer be both complimentary and critical?
Yes. Recommendations can include strengths and weaknesses. Read excerpts and recurring topics alongside the aggregate score.
What does neutral mean?
The evaluated text carries no clear positive or negative attitude. This differs from a missing mention or an unprocessed answer.
How should I investigate an inaccurate AI claim?
Open the saved answer and citations, verify facts and source freshness, and track whether the claim recurs in later measurements.
Related features
AI claims
Discover recurring claims about your brand and inspect their source evidence and verification status.
Explore this feature →: AI claimsPosition and answer analysis
See where AI recommends your brand first, which questions leave it further behind, and how stable its position is across repeated measurements.
Explore this feature →: Position and answer analysisCompetitor benchmarking
See your brand alongside competitors: who receives more mentions, who ranks higher in AI recommendations and which new brands appear in answers.
Explore this feature →: Competitor benchmarkingMake your AI presence measurable
Request a quote for your project
Tell us which brand and website you want to measure and what question you need to answer. We will discuss the subscription and data connections your project requires.