Skip to main content
Sentiment in Prefer is a read on how AI engines describe your brand in their answers, not a read on social media or reviews. Every brand mention Prefer extracts is scored as positive, neutral, or negative, then rolled up into a mix, themes, and per-engine breakdowns so you can see both the shape of the conversation and what to change.
Sentiment reflects the language AI engines use in their answers. It is not a measurement of user opinion, review scores, or social posts. Treat it as “how the model summarizes you”, not “what customers feel”.

​
What the Sentiment tile shows

  • Mix strip: percentage share of positive, neutral, and negative mentions in scope.
  • Positive trend: how positive share is moving run-over-run.
  • Drivers: theme quotes pulled verbatim from engine answers, grouped by topic.
  • Competitor benchmark: your sentiment mix next to every tracked rival’s.
  • Per-engine mix: sentiment split by engine, so you can spot models that describe you differently.
  • Negative-source drilldown: which cited URLs are driving negative mentions.

​
Investigate a negative theme

1

Open the Sentiment view

Go to Monitor > Sentiment or scroll to the Sentiment tile on the Overview.
2

Click a negative driver

Themes are pulled straight from answers. Click one to see the full engine response.
3

Open the source drilldown

Use the negative-source panel to see which cited URLs are shaping that theme.
4

Send to the Action Center

Create an action to address the source directly: a correction, an outreach, an updated owned page, or a support case.
A rising negative mix on a single engine while others stay flat almost always traces back to one or two cited sources. Start with the negative-source drilldown before rewriting content.

Sources & Citations

Find and address the sources shaping negative themes.

Action Center

Turn sentiment drivers into ranked work items.