Audience intelligence platform
You know your follower count. You do not know who those followers are, which of them other people actually listen to, or which segment is quietly forming in your replies. Sence builds that picture out of the comments your own channels already collect.
AudienceWhen comments arrive
Audience leaderboardComments
Trusted by teams using Sence
How Sence builds the picture
Connect the channels you already run
Sence connects to your Facebook, Instagram, YouTube and TikTok accounts. Facebook and Instagram connect through the Facebook Graph API and YouTube through the YouTube Data API, so analysis happens where the comments actually live. There is no manual export and no batch processing.
Every comment is classified on arrival
Each comment is read as it lands and classified for sentiment, theme, brand mention and any custom tag you have defined. No keyword list, no Boolean query, no manual tagging.
The audience assembles itself
Those classifications roll up into segments, superfans, heat maps of where engagement concentrates, and the key voices, influencers, media figures and stakeholders significantly shaping opinion. Nobody has to nominate them first.
What you can do with it
Themes26,558 comments
Controversial selections of players3,187 (12%)
Sence identifies new audience segments from real-time conversation rather than from the demographics on file, so a group forming around one product, one issue or one fixture shows up while it is still forming.
The people who comment most, defend you unprompted and answer other people’s questions are worth knowing individually. Sence identifies them instantly, so your community team can nurture loyalty where it already exists.
Sence dynamically detects the key voices, influencers, media figures and stakeholders significantly impacting opinion, which is rarely the account posting most often.
Slang, dialect and generational shorthand are read in context rather than scored against a dictionary, which is where tools trained mostly on US and UK English quietly misread an Australian or New Zealand comment section.
Conversation heat maps pinpoint the posts, formats and moments your audience turns up for, so the next content plan is built on where they are rather than where you hoped they would be.
Sence retains historical data, so you can track audiences over time, understand the impact of a tactic and put this campaign beside the last one instead of beside a memory of it.
Your owned audience, read in full
Sence works from your owned audience: every comment left on the channels you run, read as it lands. There are no Boolean filters and no keyword lists to maintain, because a query only ever finds the theme you already knew to look for, and the theme worth having is the one nobody thought to search for. Publishers use the same signal to show advertisers proof of audience quality and engagement depth, in a form a sales team can actually use. See News & Media.
- Comments on Facebook, Instagram, YouTube and TikTok
- Facebook and Instagram via the Facebook Graph API, YouTube via the YouTube Data API
- Up to 10,000 comments a minute at peak
- Sentiment classified in over 45 languages, with no manual tagging
- Historical data retained, so trends run over weeks, months or years
- Insights in the dashboard, or via API and dashboard widgets
- No Boolean filters, and no third-party or open-web monitoring
Works with
Understand
The pillar this sits in: sentiment, themes, mentions and trends, all off one comment stream.
Learn moreSentiment & Themes
The number and the reason for it. Brand-calibrated sentiment with the named themes that explain the movement.
Learn moreBrand & Competitive Intelligence
Which brands your audience names alongside yours, and how your sentiment benchmarks against theirs.
Learn moreFrequently asked questions
No. Social listening monitors third-party platforms for mentions of your name: other people's forums, other people's feeds, other people's news sites. Audience intelligence works from your owned audience, which is the comments left on channels you run. They answer different questions, and only one of them is a signal you can moderate and reply to as well as analyse.
No. Sence does not run Boolean filters. A query only returns the theme you already knew to search for, which is exactly how keyword-driven tools miss the thing that was about to matter. Every comment is classified on arrival instead, and the segments and themes assemble from what is genuinely there.
That is the point of a classifier built from your own decisions. Slang, dialect and generational shorthand are read in context, and sentiment is classified in over 45 languages. Tools trained mostly on US and UK English are the ones that read an Australian or New Zealand comment section wrong.
Sence's infrastructure handles up to 10,000 comments a minute at peak, so a launch, a final or an announcement day is read in full rather than sampled. That throughput is what makes "every comment" a description rather than a figure of speech.
In the Sence dashboard, and via API or dashboard widgets if you want them inside something your team already has open. Beyond that, Sence integrates with major digital and social media platforms as well as custom enterprise marketing tools, and custom integrations tailored to your organisation are available.