Sentiment and insights
Every mention scored positive, neutral or negative across seventeen channels, tracked against your own baseline rather than a fixed one, and grouped into the themes actually driving it. The chart arrives with its explanation attached.
Free for 7 days. No credit card. The last 30 days give you a trend on day one.
Scored across
+47−6 pts
net sentiment, against your 90-day normal
A category where people complain loudly sits at a different baseline from one where they do not. We compare you against your own 90 days, so a normal week reads as normal.
Negative mentions are grouped into what they are about, ranked by weight. You get 'refunds' and 'delivery times', not a number that dropped for reasons unknown.
Sarcasm, in-jokes and industry shorthand are where every sentiment model struggles. Low-confidence mentions are flagged rather than quietly counted, and you can correct them.
What it is
Every tool in this category can tell you sentiment fell four points. Almost none can tell you which conversation did it, on which channel, starting when, and whether the same thing happened last quarter. That gap is the entire difference between a chart and a decision.
Read across 17 named channels, from Facebook and Reddit to news, reviews and podcasts.
See the full listWhat it answers
Sentiment work goes wrong in the same place every time: someone presents a movement and cannot say what caused it. These are the four follow-ups.
How it works
The first two are routine. The third is the part that usually falls to whoever has an afternoon free.
Third time this month. Is there a delivery app that actually refunds on time? Ready to switch today.
BrandRobin · brief
Mentions up 18.2% to 12,480. One refund thread drove most of the negative share; creator coverage offset it.
Delivery
Negative spike · refunds
22:14 · +11 pts · 402 mentions
New buyer signal · 91
LinkedIn · Head of marketing, SaaS
Creator mention · 41.7K reach
YouTube · Tech with Marcus
Slack, Teams, email or a webhook when sentiment moves beyond your own normal, with the theme already named.
Net sentiment, what moved and why, at 09:00. A quiet week gets a short digest rather than an invented one.
A scheduled PDF carrying the chart, the themes underneath it and the mentions behind each theme.
Compared with the rest of the category
Who it is for
Marketing leadership
Perception is in your objectives, and the only evidence you have is a line with no explanation under it.
Net sentiment against your own baseline, the themes moving it, and a before-and-after on every campaign window.
YouTubePositivecredit where it's due, they actually fixed the thing people complained about
Brand and insight teams
You can see that something changed, but finding out what took two days of reading.
Themes ranked by weight, each opening onto its mentions, so the reading is the last step and not the first.
RedditNegativeThe recipe change is the problem, not the price.
Marketing and PR agencies
The client wants to know the campaign shifted something, and volume alone never convinces them.
Sentiment and volume either side of the campaign date, in a white-label report with the mentions attached.
InstagramPositivewhoever runs this account deserves a raise
Honest note: machine-scored sentiment is a good instrument, not a perfect one. Sarcasm, in-jokes and industry shorthand are genuinely hard, so low-confidence mentions are flagged rather than quietly counted, and your corrections are kept. Check the mentions before a number goes in front of a board. Replying is not in the first release.
FAQ
Every mention is read and scored positive, neutral or negative based on what it says about you specifically, not the general mood of the post. A furious review of somebody else that mentions you in passing is not a negative mention of you. Scoring runs on every mention on every channel, not on a sample.
Good on plain statements, weaker on sarcasm, in-jokes and industry shorthand, which is true of every tool in this category including the expensive ones. What we do differently is flag low-confidence calls instead of counting them quietly, and keep your corrections so the same mistake stops repeating.
Yes. Change the score on any mention, and the correction is kept and applied to similar mentions in your workspace. Corrections are yours; they do not leak into anyone else's data.
The share of positive mentions minus the share of negative ones, over a period you choose. It is a single number for a slide, and it always sits next to the volume it was calculated from, because a +90 from eleven mentions is not the same fact as a +47 from twelve thousand.
Because categories differ. People post about a bank differently from how they post about a snack brand, and a fixed threshold fires constantly in one and never in the other. We use your own 90 days, per channel, so a normal week reads as normal.
Themes are what the mentions are actually about: refunds, delivery times, packaging, a recipe change. Positive and negative mentions are each grouped and ranked by how much of the movement they account for, and every theme opens onto the mentions that created it.
You can mark a campaign window and see volume, net sentiment and themes on either side of it, per channel. That is evidence of a shift in public conversation around your campaign. It is not a causal claim, and we will not dress it up as one.
Sentiment is on every plan from Starter at $49 a month. Themes, campaign windows and competitor comparison start on Growth at $129, and Scale at $299 adds white-label reports and 15-minute refresh. All start with a 7-day free trial and no credit card.
7 days free. No credit card. The last 30 days give you a trend on day one.