How AI describes you: a census of 122 listings

The GEO practice explicitly hands sentiment work to a different 5W practice, so attribute-level perception is not part of the citation program a buyer signs for; a separate published Reputation Index scores sentiment for named individuals.

Checked 2026-09-10

Sentiment analysis is bundled from the entry plan up and is scored alongside visibility and share of voice in the dashboard.

Checked 2026-09-10

Sentiment is scored per aspect of the brand inside a single answer, so one response can read positive on one attribute and negative on another.

Checked 2026-09-12

Sentiment is one weighted input to the vendor's composite AI visibility score, so tone is measured, though no per-attribute breakdown of what engines associate with the brand is shown.

Checked 2026-09-10

Each simulated answer is classified positive, neutral or negative per engine, and daily community monitoring adds sentiment signals from X, Reddit, Hacker News and Product Hunt.

Checked 2026-09-19

Each simulated run logs sentiment and how the model frames the brand, and sentiment drift is reported per assistant over time.

Checked 2026-09-08

You get the topics AI associates with the brand and the raw responses to read, but the documented metric set is share of voice, mentions, citations and impressions, with no sentiment or attribute score.

Checked 2026-09-19

Sentiment themes show which recurring attributes AI answers attach to the brand, but the dashboard is withheld on Starter, which only sees mock data.

Checked 2026-10-04

Vendor promises a view of how models describe the brand, but publishes no sentiment score or named attribute set, so the depth of that reporting is unstated.

Checked 2026-09-17

A sentiment (BSI) score ships on every tier, but its trend line needs Starter and competitor BSI needs Pro; no attribute-level perception breakdown is documented.

Checked 2026-09-19

Sentiment classification runs over the collected answers and reports how the brand is framed when an engine does mention it, using the buyer's own Gemini key.

Checked 2026-09-16

A Brand Authority pillar judges entity clarity, expertise and trust signals, but the vendor never claims to extract attributes AI associates with the brand.

Checked 2026-09-20

A Sentiment tab classifies each response positive, negative, mixed or neutral by LLM judgement and extracts recurring themes.

Checked 2026-10-13

Perception scores each brand on visibility against narrative strength, surfacing the recurring themes and exact phrases engines attach to it.

Checked 2026-09-12

Sentiment is scored as one of the five visibility inputs and reported as a net figure, but the vendor describes polarity only, not which attributes AI associates with the brand.

Checked 2026-09-20

Entity and language analysis reports the people, organisations, products and topics engines attach to your prompts over time; sentiment is not a built-in score, you obtain it by firing a follow-on question at a model of your choice.

Checked 2026-09-13

Tracks direction-neutral brand attributes and their prevalence over time, for your brand and each tracked competitor separately.

Checked 2026-09-20

A single sentiment scale from negative to positive is reported rather than named attributes, and the card only appears once enough mentions exist to score.

Checked 2026-10-04

A 0-100 tone score is produced from sentences containing the detected mention, which is sentiment only; no attribute or trait association is described anywhere on the site.

Checked 2026-09-21

Sentiment, mention frequency and a direct bias score are reported, so the sheet covers how engines characterise the brand, not just whether they mention it.

Checked 2026-09-09

Sentiment is tracked beside presence, and AI Catalyst surfaces the sentiment gaps and brand-voice inconsistencies it says cost citations.

Checked 2026-09-11

The free Brand AI Scanner scores tone with a keyword classifier plus overlap against a positioning sentence you supply across four lenses; the paid plans list a monthly visibility report but never describe sentiment as a plan feature.

Checked 2026-09-08

Sentiment scoring exists with per-verdict source quotes across three engines, but the vendor labels it experimental, ships it off by default and requires a third-party TypeSafe key.

Checked 2026-10-03

Sentiment is claimed once on the Overview page and rendered as a positive or negative marker on brand cards, but no attribute-level perception analysis is described and the paid platform's docs never mention sentiment at all.

Checked 2026-09-11

Sentiment is broken down per tracked entity as positive, neutral and negative, and is meant to be read against PR activity over time rather than as a one-off score.

Checked 2026-09-11

Every captured answer is classified positive, neutral or negative and charted over time, with a separate LLM-judged context-health signal for wrong-industry or wrong-language descriptions.

Checked 2026-09-13

Sentiment is one of four headline dashboard metrics, but the vendor describes it as overall positivity rather than a breakdown of associated attributes.

Checked 2026-09-21

Sentiment analysis is included on every plan, but the vendor does not say which brand attributes or themes it surfaces.

Checked 2026-09-24

A whole module reports the attributes a model attaches to the brand: tone of voice, core keywords, market position and a dominant trait, with a confidence score.

Checked 2026-10-08

Sentiment analysis is checked on all three plans and maps themes, use cases and pain points AI attaches to the brand.

Checked 2026-09-21

Sentiment is classified positive, negative or neutral and categorised by business driver such as price, service or quality, then traced to the exact statement and source domain.

Checked 2026-10-06

The Narrative Shaping workflow tracks which attributes and differentiators AI associates with the brand and how those descriptions drift, delivering a perception baseline and a deviation list.

Checked 2026-09-15

Sentiment and share of voice sit beside mentions and position, and a free tool reports the strengths and concerns engines repeat about a brand.

Checked 2026-09-13

Sentiment is scored 0 to 100 and trended per brand, but the vendor never claims to break out which attributes AI associates with a brand.

Checked 2026-09-26

Tone analysis exists but is an optional metered AI analyst charged per analysed answer, so brand perception is not included in the base check price.

Checked 2026-09-23

Word Association scores the words models attach to the brand for both frequency and sentiment, and tracks how that language shifts over time against rivals.

Checked 2026-09-12

Sentiment is scored 0-100 across eleven attributes per engine, with a criticism feed that quotes the phrase and names the source.

Checked 2026-09-08

Sentiment is tracked monthly alongside citation share, and kickoff work inventories what each platform believes about the brand across services and reputation.

Checked 2026-10-13

Sentiment is scored as a three-way polarity per mention and trended over time; the vendor never claims an attribute or theme breakdown.

Checked 2026-09-03

Sentiment and brand-name variants are monitored alongside mentions, but the vendor states only that it shows how AI describes you; no attribute or theme breakdown is documented.

Checked 2026-10-03

Sentiment analysis grades the tone and framing of each answer per model and ranks the weakest prompts by impact.

Checked 2026-10-05

An agent extracts the descriptive phrases models use about the brand, so the output is positioning language rather than a bare sentiment score.

Checked 2026-09-24

Sentiment and how engines characterise the brand are included on every plan, though the features page hedges sentiment with 'when available'.

Checked 2026-09-14

Sentiment is supported as raw evidence for an agent or analyst to interpret, not as a scored attribute dashboard; the vendor states no score is computed.

Checked 2026-09-24

Sentiment is reported as positive, neutral or factually wrong per mention rather than as an attribute map, and it arrives through monitoring or the agency's monthly reports.

Checked 2026-09-14

Sentiment analysis splits positive from negative AI mentions across platforms; advanced sentiment reporting is itemised on the paid plans.

Checked 2026-09-25

They report whether an answer classifies the offer correctly as one of four described levels of visibility, but publish no attribute or sentiment scoring.

Checked 2026-10-03

Sentiment tracking is a named reputation deliverable backed by in-house tooling that scores digital visibility by sentiment across search, social and reviews over time.

Checked 2026-09-14

Sentiment is scored positive, neutral or negative and can be segmented by model, geography, persona and topic; the row is checked on every tier.

Checked 2026-09-15

Sentiment is a single polarity score with a trend line; the vendor does not claim per-attribute association breakdowns.

Checked 2026-09-02

Sentiment is scored per engine as positive, neutral or negative with a net score and the topics driving each shift.

Checked 2026-10-07

Brand sentiment and whether the brand is recommended are treated as monitored inputs alongside citations, though the vendor names no specific attribute taxonomy it reports against.

Checked 2026-09-26

Brand sentiment analysis surfaces the tone and the specific positive, neutral or negative words engines use, with sentiment broken down per engine in the API.

Checked 2026-10-09

Sentiment of AI responses is analysed and tracked over time, and brand sentiment alerts are listed on all three plan columns.

Checked 2026-10-10

A dedicated Sentiment page trends perception against competitors by topic and source, and separates neutral from positive and negative.

Checked 2026-09-03

A sentiment score is tracked daily per model and the audit reports how LLMs describe the business, but no named attribute or theme breakdown is published.

Checked 2026-09-03

Sentiment analysis is excluded from the entry Starter plan and first appears on Growth; deeper 7-dimension reputation reports sit higher again, on Scale.

Checked 2026-09-27

Sentiment is scored per engine as a weighting inside the visibility score and shown in the prompt breakdown, but the vendor never claims to surface which attributes AI associates with the brand.

Checked 2026-09-26

A Brand Phrases section lists the attributes and services AI associates with the business and scores sentiment, alongside a Buyer Persuasion Score from -10 to +10.

Checked 2026-10-11

Oracle reports the decision criteria and expectations a brand is qualified on axis by axis, rather than sentiment or attribute scoring.

Checked 2026-10-03

Each stored response carries sentiment, key phrases and mentioned organisations, so the attributes an engine attaches to the brand can be inspected directly.

Checked 2026-09-18

Sentiment is reported as a positive-share percentage and a distribution across prompts, which tells you tone but not which specific attributes the engines associate with the brand.

Checked 2026-09-09

Sentiment is one of four reporting sections in the AI search report, and the report is built around client-chosen attributes as well as topics and prompts.

Checked 2026-09-27

Sentiment is scored positive, neutral or negative with the surrounding context, but no named attribute or theme breakdown is documented.

Checked 2026-09-28

Sentiment Analysis scores named attributes such as pricing, UX, support and integrations, with competitor comparison, and is included on all three plans.

Checked 2026-09-28

They track brand recommendation and mention patterns across prompts, but publish nothing about attribute-level sentiment or which qualities AI associates with a brand.

Checked 2026-09-28

A named Sentiment Analysis feature covering how engines and cited sources characterise the brand by topic, with competitor comparison; the free brand report adds how AI classifies you and what you seem known for.

Checked 2026-09-03

They report how the brand is described and represented as an entity, and flag misrepresentation, but no attribute or sentiment scoring of brand perception is described.

Checked 2026-09-28

The benchmark captures the positioning language models use about the brand, and the vendor treats competitor-framed descriptions as a problem it exists to fix.

Checked 2026-09-03

Scores sentiment per engine and names the attributes attached to the brand, separating top positive themes from the objections driving negatives.

Checked 2026-09-29

A sentiment analysis grades how the engines speak about the brand, and sentiment is returned per prompt and per brand rank through the API.

Checked 2026-09-29

How the brand is described is recorded next to whether it was named, and inconsistent descriptions are surfaced as a fixable cause, but the vendor says sentiment is not a headline metric here.

Checked 2026-10-07

A dedicated Brand Perception module extracts the attributes models attach to your brand, scores them 0 to 100, and ranks you against tracked competitors on each one.

Checked 2026-09-03

Sentiment in AI-generated answers is analysed and reported monthly alongside mention counts, and improving sentiment is stated as an outcome of the digital PR workstream.

Checked 2026-09-29

Sentiment reporting extracts the themes and narratives AI attaches to the brand rather than only counting mentions.

Checked 2026-09-09

Sentiment is named twice in prose as part of the platform, but no dashboard module, plan row or feature page describes attribute-level perception, so the depth is unpublished.

Checked 2026-09-03

Sentiment is scored per response on a 100-point scale with a trend, a distribution and a breakdown by prompt type, and you can drill into the individual answers dragging the score down.

Checked 2026-09-30

Sentiment and Perception rows are included from the entry Starter plan, scoring per provider and comparing AI descriptions against the brand's stated positioning.

Checked 2026-10-10

A sentiment breakdown scores mentions positive, mixed or negative and surfaces recurring attribute phrases such as the traits AI associates with the brand.

Checked 2026-09-30

Sentiment and how AI describes the brand are tracked and folded into the Search Performance Index.

Sentiment is one of the four scoring factors and is broken out per platform, alongside a positioning matrix that maps which attributes each engine attaches to the brand against rivals.

Checked 2026-10-11

It measures how precisely engines associate and match the brand - Brand Match, Brand Target, category association - but publishes no sentiment or attribute analysis of what AI says about the brand.

Checked 2026-10-11

The Brand Audit scores accuracy, completeness and sentiment, and reports which attributes each engine attaches to the brand.

Checked 2026-10-12

Sentiment analysis by brand, topic and model exists, but the plan table marks it absent on the $20 Essentials tier, so it starts at Pro.

Checked 2026-09-30

A dedicated Brand Perception module groups positive and negative themes per AI model and lets you drill into the underlying responses, on top of a numeric sentiment score from -10 to +10.

Checked 2026-09-30

Sentiment is scored 0 to 100 per response and per competitor, but the tool reports tone rather than named brand attributes.

Checked 2026-10-13

Sentiment classified positive, mixed or negative, filterable by platform, persona, funnel stage and country.

Checked 2026-09-02

Mentions are scored for sentiment and context, so you see how answers frame the brand, not only whether it is named.

Checked 2026-10-13

Brand Performance analyses sentiment, share of voice and the narrative drivers behind mentions, then produces AI-generated positioning recommendations, refreshed weekly rather than daily.

Checked 2026-09-18

Every mention is scored positive, neutral or negative and the engine's own sentence is retained, up to 600 characters per engine, for weekly trending.

Checked 2026-10-03

Sentiment and narrative tracking records how engines portray the brand, but the comparison table shows AI Sentiment Analysis starting on Starter, with full sentiment tracking reserved for Pro.

Checked 2026-09-15

Sentiment is recorded alongside each captured answer, but the vendor never describes attribute-level or theme-level perception reporting beyond that single sentiment signal.

Checked 2026-10-08

Sentiment ships today and is returned per answer through the API, but the deeper sentiment product with source attribution behind each shift is badged coming soon on its own page.

Checked 2026-09-17

Sentiment Analysis classifies each mention positive, neutral or negative and scores it by topic against rivals.

Checked 2026-10-08

Per-mention sentiment grading is real but ships only on AI Search Analytics, Pro and Peace of Mind; the comparison table marks it absent on Standard.

Checked 2026-10-09

Branded and comparison prompts diagnose how AI frames the brand, but there is no attribute or sentiment scoring, and they never enter the score.

Checked 2026-10-10

Answers are scored across twenty perception dimensions with the quotes behind them, and the same twenty scores are exposed as metrics in the Looker connector.

Checked 2026-09-16

Each answer is scored positive, neutral or negative per engine, which is tone rather than attributes: no extraction of the qualities or themes a model associates with the brand is described.

Checked 2026-09-17

Sentiment is reported alongside mentions and share of voice across the sampled AI responses.

Checked 2026-09-02

Sentiment framing of brand mentions is sold from the Pro subscription upward; the free and one-off Starter audits do not list it.

Checked 2026-09-18

An Evaluation metric reports what AI presents as strengths and weaknesses; mentions are split into recommended, neutral and negative.

Checked 2026-09-02

Sentiment of each mention is classified and themes are shown on the dashboard, so you see how engines characterise the brand.

Checked 2026-09-18

Attributes such as trusted, premium or expensive are classified from each stored answer and rolled up, alongside a per-answer tone read.

Checked 2026-10-12

Sentiment scores, keywords and themes exist but start at the Growth plan.

Checked 2026-09-02

Sentiment polarity is tracked and the product page lists Brands' Sentiment, but no attribute or theme extraction is described.