Every client's comment section lands on your desk
The brand posts, the audience replies, and the mess arrives with your account team. Sence reads every comment on arrival, hides what breaches that brand's rules, and logs the decision with its reasoning so you have something to hand back at the review. What is left is the conversation worth reporting on.
Moderation — Zentra Mobile@grimjab_779:47 pmHidden
NegativePersonal Hate
Hidden · 9:47 pm · Rule: Hide any Personal Hate or Group Hate comments
@mia_moves9:03 pm
the new range is actually so good, wore it all weekend
PositivePraise
@dealhound_dana8:12 pmNot visible
cheap followers here → bit.ly/…
NeutralSpamWebsite Link
Trusted by the teams behind these brands
What an account team is actually managing
None of this is a moderation problem on paper. It becomes one at about 8 pm on a Thursday.
- 01
You are accountable for a comment section you do not own
The brand owns the channel. The platform owns the rules. Your team owns the consequences, and "we did not see it" has never once been an answer that landed well in a client meeting.
- 02
It happens after your team has gone home
72% of comments are made outside of work hours, peaking between 6 pm and 10 pm. That is the window nobody is rostered for, and it is when a thread turns. Since Meta reduced its platform moderation in early 2025, Sence has seen a 30% increase in the number of harmful comments on Facebook.
- 03
Every brand draws the line in a different place
One client wants the argument left up, because that argument is their audience. Another wants it gone before anyone screenshots it. A single house filter cannot serve both, and a generic toxicity score serves neither: it was trained on somebody else's idea of harmful, not on your client's.
Cover the hours nobody is rostered for
The new range — weekend drop
Most of what you need to catch arrives in the evening, on a post nobody is watching. Sence reads every comment and message on arrival and hides anything that breaches the rules that brand has set, in seconds, 24/7, at up to 10,000 comments a minute at peak. Your team picks it up in the morning with a record of what happened rather than a backlog.
- Evening and weekend threads moderated without a night shift
- One move to lock a post down while a pile-on runs
- Every hidden comment on record, with the reasoning attached
One place to work, instead of six platform tabs
Your team should not be moderating in four native interfaces and answering messages in a fifth. Comments from Facebook, Instagram, YouTube and TikTok arrive in one place, along with direct messages from Facebook, Instagram and TikTok, with spam quarantined before anyone has to read it. Moderation and insight run in a single workflow, so nobody is switching between tools.
- Comments and direct messages from every connected channel in one queue
- Spam quarantined before it reaches the person replying
- See, moderate and respond across every channel from one place
Turn the month's comments into the slide you have to present
Somebody has to explain what the audience actually said, and screenshots are not evidence. Sence detects themes, brand mentions and custom tags in real time and quantifies them, so the monthly report is built from what people wrote. Historical data stays available, so this campaign can be set against the last one.
- Sentiment and themes quantified per post, campaign or period
- Ready-to-share insights instead of hours compiling reports
- Historical conversation analysis over weeks, months or even years
The new range — weekend drop
The parts of Sence an account team runs on
The order matters. Cover the hours first, put the channels in one queue second, and the reporting stops being a job somebody does on a Friday afternoon.
Comment & DM Moderation
Comments and direct messages read as they arrive, hidden when they breach the rules that brand set, and logged either way.
Learn moreUnified Inbox
Comments and direct messages from every connected channel in one queue, so your team is not living in platform tabs.
Learn moreAudience Intelligence
Sentiment, themes and share of voice from the audience that actually commented, in a form that goes into a client report.
Learn moreThe numbers behind the pitch
These are Sence's own measurements, published on this site. Sence has analysed millions of social media comments in the last six months and moderates at up to 10,000 comments a minute at peak.
What that buys an account team is coverage of the hours nobody is rostered for, and a decision trail to take into the review. The mechanics are on Comment & DM Moderation.
37%
of the harmful comments Sence has analysed are spam and scams, ahead of harassment at 27% and hate at 26%.
8
categories of harmful content, defined and tuned per brand rather than inherited from a default model.
Frequently asked questions
Talk to us about the roster. Sence is deployed and quoted per customer, with a classifier built from that customer's own moderation decisions, so what one brand calls harmful is never imposed on another. How that maps onto an agency roster depends on who holds the channels and who signs, so it is a conversation with sales rather than something to assume from a web page.
That is the design. Sence does not apply one shared toxicity score to everybody. Each classifier is trained on that brand's own content policy decisions and its own taxonomy, so one client can keep the argument up and another can have it hidden, out of the same system.
Every moderation decision is logged with its classification, the action taken and the reasoning, so your team can see what was hidden and why. That turns the monthly review into a conversation about policy rather than an argument about what happened.
As the brand. Sence is designed to understand and mimic a brand's authentic voice, learning from that brand's existing conversations to create consistent, human-feeling responses at scale. Your team still decides what goes out.