Hear the posts that never tag you. Get an alert when they spike.
Listeners collect keyword matches across X, Threads, Instagram hashtags, TikTok, and YouTube, including spoken words and on-screen text in video. AI keeps what matters and alerts your team when volume jumps.
Volume on “outage” tripled in 40 minutes. 3 posts are from accounts with over 100k followers.
What it does
Collect
Keywords per platform across X, Threads, Instagram hashtags, TikTok, and YouTube.
Video, not just captions
On TikTok and YouTube, matches include spoken audio and on-screen text.
Keep and skip rules
Plain rules decide what stays. Every archived post shows the rule that archived it.
Spike alerts
Slack alerts after AI analysis, plus automatic alerts within the hour when volume jumps.
Promote to the inbox
One click turns a post into a conversation for your team. Nothing is deleted.
Chart it
Explore listening in Analytics or ask SiftGPT to chart it.
One Listener for each thing you follow
Most people talk about a brand without tagging it. A Listener sweeps X, Threads, Instagram hashtags, TikTok, and YouTube for your keywords, then decides what your team should see.
- Brand names, products, campaigns, or competitors
- Keywords set per platform
- Posts from the last 90 days, kept for up to a year
Hear what people say in videos
On TikTok and YouTube, Sift matches what people say out loud and the text on screen, not just captions. A short AI summary tells you what each video is about before you press play.
- Spoken words and on-screen text
- An AI summary of every video
- Skip small accounts with a follower minimum
The creator says a hotel had no record of their Wanderly booking and asks if it happened to anyone else.
Simple rules, and a reason for every call
Keep and skip rules are plain chips, like “mentions the app” or “skip giveaway”. Three or four skip words remove most of the noise. Every archived post says which rule archived it, in a sentence.
- Keep, skip, follower count, source, and channel rules
- Every archived post explains why
- Nothing is deleted. Bring any post back in one click
Ask SiftGPT “why was that post archived?” It reads the same rules you see.
Know before it trends
Watch conditions run after AI analysis, so you can get an alert for things like a negative post from a big account. Spikes alert on their own: if a Listener pulls far more than usual, your team hears about it within the hour.
- Slack alerts on the conditions you choose
- Automatic spike alerts within the hour
- Each alert links to the posts behind it
Most of the spike comes from one video about late refunds. Three posts are from accounts with over 100k followers.
See the whole picture
The Listening tab in Analytics charts mentions over time, what was kept and archived, volume by platform, and the sentiment of what you kept. Ask SiftGPT to chart it in chat, or export any view to CSV.
- Pulled, kept, archived, and actions at a glance
- Keep rate for each Listener
- CSV export of the current view
Set up with our team
Sift’s team sets up Listeners with you, so collection stays focused on what matters. Check setup tests every rule and previews the volume to expect. Going live is always a person’s call.
- Check setup previews expected volume
- Managers view, send to inbox, and export
- You decide when a Listener goes live
Where agents act, and where you decide.
Agents do the sweeping and sorting. Your team decides what to answer.
- AgentCollect
Sweeps your keywords across five platforms, including speech and text in video.
- AgentKeep or archive
Keeps the real conversations, archives the noise, and says why.
- AgentWatch
Reads sentiment and alerts Slack on spikes and the conditions you set.
- PersonAct
Sends a post to the inbox, answers it, or shares the trend with the team.
Works across your channels
All integrations →Questions
Which platforms does listening cover?
X, Threads, Instagram hashtags, TikTok, and YouTube. On TikTok and YouTube it also matches spoken words and on-screen text.
What happens to posts that get skipped?
They go to an archive you can review, with the rule that skipped them. Nothing is deleted, and you can send any post to the inbox in one click.
Can my team answer a post Sift finds?
Send it to the inbox and your team picks it up like any other conversation.
How do alerts work?
Alerts go to Slack. You choose watch conditions, like negative sentiment, and Sift alerts on volume spikes within the hour on its own.
Who sets up listeners?
The Sift team sets up rules, sources, and alerts with you. Your managers view the feed, send posts to the inbox, and export.
How far back does it go?
Your Listen feed keeps posts for up to 365 days. Sift only collects posts published in the last 90 days, so the feed stays current.
More of the platform
Find the posts that never tagged you.
We’ll set up a listener for your brand and show what it catches.