Your newsroom cannot switch the comments off
Turning comments off is the easy answer, and it costs you the audience that came to argue. Sence reads every comment on arrival across Facebook, Instagram, YouTube and TikTok, hides what breaches your policy, and leaves the rest of the thread standing. Your classifier is built from your own moderation decisions, not from somebody else's idea of harmful.
Moderation — The Morning WireThe Morning WireJudge releases sentencing decision in high-profile case3,058 comments · 118 flagged
@courtwatcher_889:21 pm
Great reporting, this took real digging
Positive
@late_edition_reader9:26 pmHidden
NegativeVictim Blaming
@grimjab_779:31 pmHidden
NegativeHarassment of JudgesPersonal Hate
Decision log
Hidden · Victim Blaming, Negative
Rule and reasoning logged
Trusted by newsrooms and broadcasters
What a newsroom is up against
Three things make a newsroom comment section harder than anyone else's, and none of them is solved by moderating faster.
- 01
The story that needs comments most is the one you cannot moderate
Health, crime and politics draw the highest volume and the worst of it, on the days your desk is already stretched. That is where the complaints come from too, and the harm your readers meet lands most often on Facebook and Instagram — the two channels your masthead posts to hardest.
- 02
You are answerable for what readers write under your posts
The Voller case made Australian publishers legally responsible for reader comments on their Facebook posts. Plenty of newsrooms responded by switching comments off. That is not moderation, it is surrender, and it hands your audience to somebody else's platform.
- 03
A comment section going bad is a revenue problem
For publishers who depend on engagement metrics to justify advertising rates or to unlock Meta's monetisation programmes, a deteriorating comment section shows up in the numbers. Advertisers want proof of audience quality and engagement depth, and that proof has to come from somewhere.
Keep comments open on the stories that draw the worst
Judge releases sentencing decision in high-profile case
@courtwatcher_889:21 pm
Great reporting, this took real digging
Positive
@late_edition_reader9:26 pmHidden
NegativeVictim Blaming
@grimjab_779:31 pmHidden
NegativeHarassment of JudgesPersonal Hate
HideANYPersonal HateORHarassment of Judgescomments
HiddenVictim Blaming, Negative· rule + reasoning logged
A big story arrives with volume attached, and it arrives all at once. Sence reads every comment on arrival across Facebook, Instagram, YouTube and TikTok and hides anything that breaches your policy before your readers see it, at up to 10,000 comments a minute at peak. Nothing lapses overnight, which matters when 72% of comments are posted outside regular working hours.
- Hate, harassment, spam and scams hidden in seconds, 24/7
- Comments locked down on any post in one move
- Every decision logged with its classification and the reasoning
Read the room on every story you publish
Every comment is a reader telling you something about the piece. Sence turns the reaction under a story into brand-calibrated sentiment and named themes, so your editors can see what actually landed instead of counting likes. Themes are detected in real time and tracked over any period you choose, so this story can be compared with the last one like it.
- Sentiment on every story, section and masthead post
- Themes surfaced from the noise, not from a keyword list
- Benchmark conversation shifts over weeks, months or even years
Give your commercial team something to sell
Advertisers want proof of audience quality and engagement depth. Sence surfaces audience sentiment, themes and share of voice signal from the same comment stream your moderators already work in, in a form a sales team can use. It is your own audience rather than a sample of somebody else's.
- Sentiment, themes and share of voice from your own readers
- Moderation and insight in a single workflow, not two tools
- Historical conversation analysis, so a section can be compared like for like
Judge releases sentencing decision in high-profile case
@courtwatcher_889:21 pm
Great reporting, this took real digging
Positive
@late_edition_reader9:26 pmHidden
NegativeVictim Blaming
@grimjab_779:31 pmHidden
NegativeHarassment of JudgesPersonal Hate
HideANYPersonal HateORHarassment of Judgescomments
HiddenVictim Blaming, Negative· rule + reasoning logged
The parts of Sence a newsroom leans on
Not the whole platform. These three do most of the work in a newsroom, and the first two run before anyone on your desk has opened a tab.
Thanks for having me London!
Comment & DM Moderation
Every comment and message read on arrival across your channels, with anything that breaches your policy hidden before readers see it.
Learn moreHideANYPersonal HateORHarm WishANDRaceORReligionANDNegativecomments
Applies to: 2 channels · News section
@terracecal9:12 pm
We were robbed, that ref needs glasses 🤓
NegativeOpponent Hostility
@grimjab_779:14 pmHidden
NegativePersonal HateRace
Matched rule: Identity Based Abuse
Custom Classification
Sence does not ship a default classifier. Every classifier is built from your own moderation decisions and your own policy taxonomy. A health desk and a sports desk can set different categories and different thresholds.
Learn moreAudience Intelligence
Sentiment, themes and share of voice signal from your own readers, in a form your commercial team can put in front of an advertiser.
Learn moreWhat a newsroom should check before it buys
Sence works with national news publishers and broadcasters across New Zealand, Australia and the United States, with the first Australian publisher customer signed in late 2025. Moderation runs at up to 10,000 comments a minute at peak, which is the number that matters on a breaking story rather than on an average Tuesday.
Every moderation decision is logged with its classification, the action taken and the reasoning, so your team can see what was hidden and why. Sence supports comment liability obligations including frameworks such as Voller in Australia and is GDPR compliant. The mechanics of that sit on Comment & DM Moderation.
83 to 410 hours
saved each month by Sence clients, that would otherwise be spent manually moderating comment sections.
Frequently asked questions
Yes. The classification taxonomy is customer defined, so a desk covering sensitive health topics can set different thresholds and categories from one covering sport or entertainment. Sence trains on your own content policy decisions rather than applying a generic toxicity classifier, so the model learns what harmful means to your newsroom rather than to the internet.
Every moderation decision is fully auditable, with the classification, the action and the reasoning all visible to your team. Sence supports comment liability obligations including frameworks such as Voller in Australia and is GDPR compliant, and it is a New Zealand company operating under the Privacy Act 2020.
It is a real integration. Sence connects directly to Facebook and Instagram via the Facebook Graph API, and to YouTube via the YouTube Data API, so moderation happens where the comments actually live. There is no manual export and no batch processing.
They read different data. Social listening monitors third party platforms and tells you what the world is saying. Sence works from the comments on your own channels, which is what your readers said under your stories. For a publisher the second one is the owned asset, and it is the one an advertiser will pay for.