Workflow
The Crisis-Response Workflow You Build Before the Crisis
A rapid-response framework whose real work happens in calm: pre-built severity tiers, a pre-cleared approval chain, and reply templates, so a live crisis becomes activation instead of improvisation.
By Bell Chen, founder of Superdirector — how the analysis works.
A measurement rule from the social-operator world doubles as a crisis rule: pick the two or three numbers that would actually change what you do next. In a crisis the two numbers are time-to-detect and time-to-response-live, and both are decided long before the crisis, by how much you prepared. A response window measured in hours is not enough time to design a severity scale, negotiate an approval chain, and find the brand voice from scratch.
In my experience running brand channels at Backlinker AI and through a consumer launch window in February 2026, crisis response is really two workflows. One runs in calm: build the severity tiers, pre-clear the approver and legal, write the pause protocol and the reply templates.
The other runs live: detect, pause, draft, approve, publish, monitor, and review. The live workflow only moves at the speed the calm one made possible. The sequence below documents both, with named operators, a worked example for a clearly fictional brand, and the failure modes that turn a manageable incident into a self-inflicted second story.
Why the first move is defensive, not the statement
The instinct in a crisis is to rush out a statement. The faster, higher-leverage first move is to pause everything already scheduled. Distribution is algorithmic now, so a queued promotional post does not just sit quietly until you get to it. It gets pushed. Platforms now distribute content aggressively beyond a brand’s own followers. A tone-deaf scheduled post during a crisis can travel to exactly the people forming an opinion.
So the order matters. Detect, then pause, then draft. The pause is what stops a single problem from becoming two, and it costs minutes if the queue is visible in one place. The statement is the second move, and it lands better when the team is not simultaneously apologizing for an automated post that went out mid-crisis. Defense first, because the cheapest crisis to handle is the second one you did not cause.
How operators stay fast without losing the brand
Rachel Karten, formerly social media lead at Bon Appétit, now Link in Bio
100,000 subscribers, in-house and agency social media managers
A clear measurement discipline is what keeps the retrospective honest. The principle is to pick the two or three numbers that would actually change what you do next. For crisis those are time-to-detect and time-to-response-live. If detection lagged, the fix is monitoring and alerts; if approval lagged, the fix is the pre-cleared chain. The numbers tell you which half of the playbook to harden before the next incident.Kendall Hope Tucker, Ramp
Maintains a documented, recognizable brand voice.
Tucker leads creative experimentation at Ramp, whose voice stays recognizable even on reactive content. The brand rule is to borrow a recognizable frame but always carry a brand-specific twist. The crisis equivalent is the pre-approved reply template: it encodes the brand voice in advance, so the person answering comments under pressure sounds like the brand instead of like a panicked intern.The calm-time playbook (where the speed comes from)
Everything that makes a live response fast is built when nothing is on fire. The severity tiers (Tier 1 isolated complaints, Tier 2 a trending topic or press coverage, Tier 3 viral backlash) with a response-time target attached to each, so the moment you classify an incident the clock and the path are already set. The pre-cleared approval chain, agreed with legal and an executive sponsor during calm: ideally one decision-maker plus legal, with the timeframes written down.
Then the two artifacts that save the most time live: the pause protocol (one place to see and freeze the entire scheduled queue) and the reply-template library (pre-approved language for acknowledging, updating, and redirecting, so community management does not improvise tone). Add the living-document pattern for evolving situations: a pinned statement you update as facts confirm, rather than a new post per rumor. None of this is dramatic, which is the point. The drama is what happens when it is missing.
A worked example (fictional brand)
Take a fictional DTC kitchenware brand, Coppertop, with a two-person social team. A product-quality complaint starts trending and a small trade outlet picks it up: Tier 2 on the pre-built scale, which sets a two-hour response target. The first move is not the statement. The team opens the scheduled queue and pauses a cheery promo set to publish that afternoon, the post that would have become the second story.
Then they draft against a reference: a short statement, a talking-head script from the founder, and a reply template for comments. The pre-cleared chain (founder plus a legal contact) approves in 90 minutes against the two-hour target. They publish on the most visible platform first, pin it, and work comments from the template. The retrospective shows detection lagged three hours because nobody owned monitoring that morning, so they add an alert and a morning-owner rotation. The brand is fictional; the sequence is the one I would run.
The failure modes that cost you the window
Improvising the approval chain mid-crisis. Negotiating who signs off while the clock runs is how a two-hour response becomes a two-day one. The chain has to be agreed in calm; the crisis is for activating it, not designing it.
Leaving scheduled posts live. The most avoidable self-inflicted wound. An auto-published promotion during a crisis travels on algorithmic reach to exactly the wrong audience. Pause the queue before you draft a word.
Responding everywhere at once with copy-paste. Simultaneous identical posts read as automated and ignore platform-specific tone. Lead on the most visible channel, then cascade with light tailoring.
Rebutting every rumor point by point. It extends the story and signals defensiveness. Pin a living statement of confirmed facts and actions, and redirect to it.
Skipping the retrospective. A crisis you do not review is a crisis you will handle just as slowly next time. The two numbers from the retro are what make the playbook faster.
What to track to make the next crisis faster
Time-to-detect
Gap between the incident starting to spread and the team noticing. A long gap points at monitoring and alerting, not at the response itself.
Time-to-response-live versus tier SLA
Minutes from classification to the approved response publishing, measured against the tier target. Creeping past it usually means the approval chain, not the drafting.
Sentiment delta
Direction of mention sentiment in the 24 to 48 hours after the response. The signal you want is a clear inflection, not a number to optimize.
Scheduled-post pause completeness
Whether the entire queue was frozen before the response went out. Anything that slipped through is a process gap in the pause protocol.
Alex Hormozi’s rule is the whole case for preparation: “Focus beats intelligence. Consistency beats intensity. Action beats intention.” The boring work, the tiers, the pre-cleared chain, the pause protocol, the templates, is exactly what lets you respond well under pressure. Heroic improvisation is what you do when the boring work was skipped.
Where a planning-first tool fits
The severity tiers, approval chain, and pause protocol live in a doc and your scheduling tool. The place a planning tool earns its slot is the reference and template work: analyzing how comparable brands framed similar incidents so your drafts start from a studied baseline rather than a blank page, and building the reply-template library that keeps voice intact under pressure.
A tool that turns reference content and a brand profile into draft framings and templates is one option, alongside a saved swipe file and a shared doc. Superdirector is the planning-first tool I built around this kind of reference-to-draft procedure.