← Help centre

Brand listening and sentiment

Where: Inbox > Mentions (what people said) and Analytics > Inbox (the summary). Inferno only.

Three things at once

SAVED SEARCHES. You tell Ignix what to watch for and it checks on its own: your brand, a competitor, a product or a topic. Two sources: X and the NEWS (press and blogs, from Google News). Reddit is paused while Reddit's API is closed to third-party apps. Go to Inbox > Mentions > "Manage searches" and add one: a name, the source, and what to look for. On X you can also tick "no retweets" and set a language without learning the syntax; if you know X operators ("exact phrase", OR, -is:retweet) you can type them straight into the query. Only posts that actually contain the words are kept: on X a search also returns posts published WITH a tool of that name, and those are dropped. In the news, your own website is left out.

Brand listening is an INFERNO feature: saved searches, mentions, sentiment, themes and reputation alerts all live there. Inferno includes 50 saved searches. X and the news need no connected account and spend no X credits. If you downgrade, your searches are kept but paused, so nothing is lost if you come back up.

Nothing runs until you create a search, and sentiment and alerts only work once you switch them on: if you do not use it, it costs nothing.

How often they run: X every six hours and the news once a day.

What you do with a mention: read it, reply to it from Ignix, open it on the network, or mark it read. Replies go out through your connected account.

SENTIMENT. Every comment, direct message, review and found mention is labelled positive, neutral or negative automatically. Analytics > Inbox shows the split for the period, a summary of what people are talking about, the recurring themes, and the most recent negative messages so you can act instead of just looking at a bar. Message text used for this is deleted after 90 days. Sentiment analysis starts switched off: turn it on from Analytics or in Sparky AI > Replies & alerts. It measures from that moment on; what arrived before is not labelled. The negative-spike alert only works while it is on.

ALERT. When negative messages in a day rise clearly above the usual average, Ignix warns you the same day. Configure it in Sparky AI > Replies & alerts: turn it on or off, set the minimum number of negative messages worth warning about (default 3), and choose which TEAM receives it. That last part matters in a company: whoever opened the account is rarely the person who handles reputation. With no team chosen, it goes to owners and admins.

OPPORTUNITIES. A search on X can also look for people asking for what you sell: tick "Look for opportunities" when you create it. When a post is a real opportunity, Sparky drafts a reply that says openly it comes from you, and the post shows up in Mentions marked "Opportunity". You read it, edit it if you like, and approve it to publish (Claude can do the same through Ignix). Nothing is ever posted on its own — on X, unsolicited automatic replies break the network's rules and get accounts restricted. Posts that are not opportunities are marked read so they do not clutter your inbox.

What Ignix does not watch: Instagram, TikTok, YouTube, Facebook and LinkedIn — people rarely talk about other brands there, and what matters to you on those networks is what reaches your own accounts: comments, DMs and reviews, which are labelled and counted just the same.

All articles

Brand listening and sentiment — Ignix Help