Comms & PR
Catch a crisis while it is still one post
A bad thread does not announce itself. It looks like an ordinary complaint until the hour it does not. SoTrace watches the rate as closely as the words, and pages someone while the story is still one post long.
The problem
The first hour is the only hour you control
A crisis almost never starts as a crisis. It starts as one annoyed customer with an audience, on a platform nobody on your team was reading at the time. The gap between that post and the moment someone senior hears about it is where a manageable complaint turns into a story with its own momentum, and by then the narrative has been written by everyone except you.
You find out from a screenshot
Someone forwards the thread hours later, usually after it has been quoted somewhere with a bigger audience. The team's first move is reconstructing what happened instead of responding to it.
Keyword alerts cried wolf until you muted them
A tool that pings on every mention of your name trains a team to ignore it. The one alert that mattered arrives in the same tone as the four hundred that did not.
Nobody agrees on whether it is serious
Without reach and sentiment attached, the first hour is spent debating whether the post is worth escalating rather than deciding what to say.
The solution
How SoTrace solves it
- 01
Watch the baseline, not the keyword
Every keyword carries its own normal hour. A rule fires on the deviation, so a quiet brand and a loud one both get a useful threshold without anyone tuning numbers by hand.
- 02
Page one channel, once
The first matching mention opens the escalation and the rest attach to it, so the team reads one thread instead of forty notifications of the same story.
- 03
Arrive with the context
Reach, sentiment, the platform and the summary travel with the alert, enough to decide whether to reply, brief the CEO, or wait it out.
What changes
Before and after, for the team doing the work
Without SoTrace
- The story is found by chance, usually after it has spread
- Alerts fire on every mention, so the team stops reading them
- The first response is written under pressure, with no context
With SoTrace
- Minutes between the first angry post and someone reading it
- One escalation thread per story instead of one per mention
- A record of when you knew, for the review afterwards
Questions
What teams ask before they switch
- How fast does an alert actually arrive?
- Keywords are polled continuously and matched mentions are classified as they land, so an escalation typically reaches Slack or email within minutes of the post appearing, not on the next scheduled report.
- How is this different from a keyword alert?
- A keyword alert fires on a word. A spike rule fires on a change in rate: it compares the current hour against what is normal for that keyword, so an ordinary complaint stays quiet and an unusual run of them does not.
- Which platforms are watched?
- X, Reddit, YouTube, TikTok, Instagram, Threads, LinkedIn and GitHub, all in one feed, with the same filters and the same alert rules across every one of them. Reddit, YouTube and LinkedIn are on the Pro plan and above.
- Can alerts be routed to the team that owns them?
- Yes. Rules route by severity and by keyword, so a reputational spike can page the comms channel while routine mentions go to a daily digest, and every rule can be tested before it goes live.