The velocity clock
A real baseline per brand, so one threshold doesn't have to fit every client you run.
PlatformCrisis Detection
A velocity clock measured against your baseline, not a smoke alarm.
Spikes measured against your brand's real history, severity that escalates only when a second signal agrees, and a brief that states plainly what is known and what isn't yet.
ContinuousChecked around the clock
Your baselineNot a generic threshold
Evidence-ledEscalation needs a second signal
Volume crosses 3.1× your baseline while your team is asleep.
The question isn't whether something spiked. It's whether the spike deserved waking anyone, and most tools can't tell.
1 / 6
- 23:40
The clock notices
38 posts in the last three hours against a normal 12 an hour for your brand. Not a generic threshold, your number, learned from your history.
- +1 min
It opens at Attention
Social volume alone doesn't earn a critical. Severity starts where the evidence starts, so the 2am phone call is reserved for something real.
- +2 min
The brief writes itself
What's known: one product batch, started on X, two regional pickups. What isn't: whether the batch issue is confirmed, whether it's regional or national. Kept visibly apart.
- 00:20
Press agrees, independently
Press volume crosses its own baseline on the same subject. Two independent signals now agree, and that's what escalation is supposed to mean.
- 00:21
Escalated to High
Severity moves on evidence, not on volume shouting louder. The on-call escalation fires, with the brief attached rather than a bare alert.
- 00:24
A statement, already drafted
Content Studio has a holding statement built from the actual facts in the brief, in the language the story broke in, waiting for a human to approve.
Everything feeding this layer.
Crisis doesn't come from one channel, and a threshold that works for a loud brand is useless for a quiet one. Both problems are handled by design.
Signals feeding the clock
Social velocityReal post-level volume against your hourly baseline
Instagram clustersHashtag co-occurrence forming around an issue
Press spikesStory volume moving past your normal coverage rate
Sentiment collapseA negative cluster forming faster than volume alone
YouTube wavesSeveral videos making the same argument
AI answer shiftsAn engine suddenly repeating a damaging claim
The screen your team actually opens.
Illustrative numbers. Your workspace runs on your brand's real data from day one.
Crisis
1 open · own brand
Velocityvs your baseline
3.1×Attention
Time to alertMedian, spike to notification
4 min
Volume against baselineLast 12 hours
Open alert · what we know
- Complaints centre on one product batchKnown
- Two regional outlets have picked it upKnown
- Whether the batch issue is confirmedOpen
- Whether it is regional or nationalOpen
From raw signal to something you can act on.
- 01
A baseline that is actually yours
Volume is measured against your brand's real history, so the threshold means the same thing for a quiet B2B brand and a national FMCG name.
- 02
Severity that escalates on evidence
A spike alone opens low. It escalates when a second signal agrees: sentiment collapsing, or press moving independently on the same subject.
- 03
A brief, not a guess
Confirmed facts and open questions are stated separately, drawn from the real posts and stories behind the alert, so your team acts on what's known.
Crisis Detection, in full.
Severity laddering
Escalation requires agreement from a second channel, which is what keeps alerts worth reading.
The crisis brief
Known facts and open questions, kept apart, drawn from the real rows behind the alert.
A drafted response
The same brief feeds Content Studio, so a holding statement starts from facts rather than a blank page.
On-call escalation
Out-of-hours escalation to a human, with a response SLA, available as a retainer.
Resolution tracking
Every alert carries its own timeline, so the post-mortem has the real sequence rather than a reconstruction.
The same layer, three different jobs.
Be the one who calls first
Per-client baselines mean a spike on a small brand isn't drowned out by a loud one, and the call comes from you.
Brief leadership in minutes
A known-versus-unknown brief is exactly the structure a CEO needs, and it's ready before the first meeting.
Cut false alarms
Evidence-led escalation means the alerts that reach you are the ones worth reading, which is what makes people trust the system.
With and without Brillaince.
Without BrillainceWith Brillaince
ThresholdOne generic number for every brandYour own baseline, learned from your history
EscalationEvery spike reads as criticalEscalates only when a second signal agrees
What arrivesAn alert and a numberA brief: what's known, what isn't, and a draft
Out of hoursNobody is watchingContinuous checks, with on-call escalation available
After the eventA reconstruction from memoryA recorded timeline for the post-mortem
What this deliberately doesn't do
Every monitoring tool's page is wall-to-wall claims. Here is where the honest line sits.
- We won't call something critical on one signalA single spike opens at a lower severity by design. Tools that escalate everything to critical get ignored within a month, and then they're worse than nothing.
- Volume is not amplificationWe measure how many real posts and stories exist, not how far they travelled, because platforms don't expose that for third-party content. The alert says so.
- A competitor's crisis is not yoursVelocity tracking is scoped to your own brand and products. A rival's bad week shows up as intelligence, not as an alarm at 2am.
Works with the rest of the platform.
Crisis Detection isn't a standalone tool. It feeds, and is fed by, every other layer Brillaince tracks.
About Crisis Detection.
What actually triggers an alert?
Volume moving meaningfully past your brand's own baseline, on press or social. A social-only spike opens at a lower severity and escalates only if sentiment or press confirms it independently.
How is this different from a generic spike alert?
Generic alerts use a flat threshold, which means constant false alarms on a busy brand and silence on a quiet one. Ours is calibrated per brand, and escalation needs corroboration.
Does it draft the response too?
Yes. The crisis brief feeds Content Studio directly, so a holding statement or media response starts from the real facts behind the alert.
Can we get a human on call?
Yes, as a 24/7 crisis retainer add-on with a response SLA. The platform detects continuously; the retainer puts a person behind the alert outside business hours.
How quickly do alerts arrive?
Social is checked on a 30-minute cycle that tightens during an open crisis, press every 2 hours, and the velocity check runs continuously against both.
See crisis detection on your own brand's data.
A live walkthrough with our team. No slideware.
Book a Demo