Sentiment Audit | Does AI think positively or negatively about your brand?
Sentiment Audit
Does AI think positively or negatively about your brand?
- Find out whether AI likes your brand or quietly sends buyers to competitors instead
- Understand which topics and buying criteria AI associates positively with your brand and which ones it doesn't
- See how AI sentiment shifts persona by persona so you know exactly which buyers AI is winning for you and which it's losing
What the Sentiment Audit shows you
Understand whether AI likes you and track favorability, opinions, and positive and negative traits AI associates with your brand and competitors
Overall and Competitive Sentiment
Gives your brand one sentiment score from -1 to 1, built from every scored answer, and ranks it against the competitors you track. Each monthly run adds a point to the trend line, so you can see whether AI is warming to you or cooling off.
What Shapes Your Sentiment
Ranks every persona and every topic by sentiment, so you can see which buyers AI favors you for and which buying criteria pull your score down. In this audit, AI rates Porsche at +0.81 on driving performance and below zero on ownership costs. Click any row to read the conversations behind it.
Recurring Themes and Buyer Stages
Groups what AI says about your brand into recurring themes, ordered by how many answers raise each one. Buyer stages split your score between buyers who are still learning the category and buyers who are comparing vendors, which shows whether AI turns against you as the decision gets closer.
Sentiment Against Visibility
Plots each topic by how often AI mentions you (horizontal) and how positive it is when it does (vertical), with bubble size showing how often the topic comes up. Fix the topics in the lower right first: AI already brings you up there, but what it says is not helping you.
Sentiment Breakdown
Crosses every topic with each persona, AI model, or buyer stage on one heatmap, so you can tell whether a weak score comes from one model or from all of them. Each cell shows its score and the number of scored observations behind it.
Every score links back to the answer that produced it
The Conversations view lists every question your personas asked and how each AI model scored on each topic. Filter by topic, persona, model, buyer stage, theme, or brand, then click a model to read its full answer.
Why a Visibility Audit isn't enough
A Visibility Audit tells you if AI mentions you. It won't tell you what AI thinks.
Visibility prompts are open-ended, and the answers skew positive. A standard audit can make everything look fine even when AI is lukewarm on your brand or quietly steering buyers to competitors. The Sentiment Audit is built to pull out the real opinion, using prompt types a Visibility Audit never runs.
Direct brand-mention prompts
We ask models point-blank what they think of your brand, so the real tone, positive or negative, has nowhere to hide.
Head-to-head prompts
We put you directly against two or more competitors so you see who AI actually prefers, and the reasons it gives.
How it works
How the Sentiment Audit scores AI answers
- Your personas ask buying questions Each persona asks questions at two buyer stages: while they learn the category, and while they compare vendors. Most questions name your brand alongside two or more competitors.
- Every AI model answers The same questions go to each AI model selected for your project, so you can compare how ChatGPT, Gemini, Claude, and Perplexity each describe you.
- A judge scores each answer eight times An AI judge scores sentiment toward your brand and each competitor the answer discusses, topic by topic, from -1 to 1. It scores every answer eight times with competitors in a different order each time, and the audit averages all eight.
- Themes are pulled out and grouped A separate pass pulls the pros, cons, and caveats each answer gives for your brand and your competitors, then groups them into the recurring themes on your dashboard.
Which plans include the Sentiment Audit
- Free sample: Not Included
- Build your own: Add for $100/mo
- Monthly: runs with the Visibility Audit
- Playbook: Monthly, runs with the Visibility Audit
Frequently Asked Questions
What is a Sentiment Audit?
A Sentiment Audit measures how positively or negatively AI models talk about your brand versus competitors. Visibility tells you whether you appear in AI answers. Sentiment tells you whether those appearances help or hurt you.
Why does AI sentiment matter?
Buyers increasingly ask AI for recommendations and comparisons. If models consistently attach caveats to your brand, or describe a competitor as the safer choice, that framing shapes deals before you ever see them. You can't correct a narrative you can't see.
Where does negative sentiment come from?
Usually from the sources AI models rely on: old reviews, outdated pricing complaints, a rough patch in your history that remains well-documented. The Sentiment Audit pairs well with the Citation Audit, which shows which sources are feeding the narrative.
How is the sentiment score calculated?
An AI judge reads each answer and scores sentiment toward your brand and each competitor it discusses, topic by topic, on a scale from -1 (negative) to 1 (positive). Every answer is judged eight times and the scores are averaged. Those averages roll up into your overall score and into the persona, topic, model, and buyer stage views.
Can I see the answers behind a score?
Yes. The Conversations view lists every question your personas asked, with each AI model's score on each topic. Filter by topic, persona, model, buyer stage, theme, or brand, then click a model to read its full answer. Every row on the dashboard links straight to its conversations.
Which AI models does it cover?
The Sentiment Audit asks the AI models selected for your project, such as ChatGPT, Gemini, Claude, Perplexity, Grok, and Google AI Overviews.
Which plan includes it?
The Sentiment Audit is included in the Playbook ($299/mo), or you can add it to Build your own for $100/mo. Either way it runs monthly alongside your Visibility Audit.