Audience mood
Social media sentiment analysis for your comments
OutReply scores the mood of the comments people leave on your Facebook, Instagram, YouTube and X posts, then shows the trend: what lifts the mood, what drags it down and who is behind it.
Sentiment also shapes the reply. Comments in the review queue show the commenter's mood and likely business impact, and a workflow can route complaints to a person.
- Positive, neutral or negative on every scored comment
- Mood over time, pain points and emerging topics
- No credits used for scoring
Which comments are scored
Comments on your Facebook, Instagram and YouTube posts are collected in the background, and X comments are added as you open posts in OutReply. Mentions found by Reddit monitoring carry their own sentiment label. TikTok comments aren't included, because OutReply doesn't read them.
Scoring runs in batches every ten minutes on OutReply's own servers, so it doesn't use credits and no comment is sent to a third-party AI just to be labelled.
Two sentiment engines
VADER on Free
A rules-and-lexicon model built for social media. It understands emoji, ALL CAPS, repeated exclamation marks, negation such as "not bad" and "but" contrasts.
Neural model on paid plans
Advanced and Elite use a multilingual transformer tuned on social media posts, which reads comments in many languages. After upgrading, you can re-score comments that were labelled by VADER.
What you see in the dashboard
- Mood split: how your audience divides between positive, neutral and negative, on every plan.
- Audience pulse: mood over a window you choose, recent pain points, what's resonating and topics picking up fastest compared with the previous period.
- Mood spikes: days when the share of negative or positive comments jumps well above the previous week are flagged. Quiet days with a single angry comment are ignored.
- People, not just percentages: each commenter's mood over time lives in their profile; see audience intelligence.
Sentiment where you reply
In the review queue, OutReply can read an unanswered comment and label the commenter's mood (from frustrated or confused to happy or excited), its likely impact on your business and a suggested next step. In DMs, the AI sidekick reads the conversation and reports the customer's mood, key topics and a recommended approach for one credit.
Instagram comment
Third time my order has arrived late. Not impressed.
What OutReply shows
Mood: frustrated. Impact: negative. Suggested action: apologise and move the conversation to DMs.
Facebook comment
Obsessed with the new menu!! 🔥🔥
What OutReply shows
Mood: excited. Impact: very positive. Suggested action: thank them and invite them back.
To act on it automatically, add an AI condition such as "Is this comment a complaint?" to a workflow, or hide abusive comments with comment moderation.
Sentiment analysis FAQ
How does OutReply calculate sentiment?
Each comment gets a positive, neutral or negative label and a score between -1 and 1. Free brands are scored with VADER, a lexicon model built for social media; Advanced and Elite use a multilingual neural model. Both run on OutReply's servers.
Which platforms are included?
Facebook, Instagram and YouTube comments are collected automatically, X comments as you view them in OutReply, and Reddit mentions through Reddit monitoring. TikTok comments aren't included because OutReply doesn't read them.
Does sentiment analysis use credits?
No. Comments are scored in batches every ten minutes at no credit cost. The AI analysis of a DM conversation costs one credit per run.
What is included on the Free plan?
Free shows the overall mood split of your audience and a capped list of commenters. Mood over time, pain points, topic insights and per-person timelines are part of Advanced and Elite.
Related features
Know how your audience feels, today
Connect your accounts on the Free plan and see the mood split of your comments.