
Product recalls
Run the recall language past the customers, the regulator, and counsel before it's published — and find the phrase that reads as blame-shifting.
A recall, an outage, a breach, a layoff, a story that's about to break. Presaga Crisis shows how customers, employees, the media and regulators are likely to react to what you're about to say — and drafts, tests and refines the statement with you — before you say it.
You tell it what happened, in your own words. It builds an AI version of everyone who'll hear about it and lets them react. Then it argues over what that means, and hands you a statement you can actually send — with the reasons attached.
By the time the statement is public, every angle should already have been tested. Most statements are tested on the people who wrote them, the general counsel who has to sign them, and the clock. Presaga Crisis runs it past the people who will actually receive it — the customers, the staff, the journalist on the story, the regulator who has to be seen to act — and tells you what breaks first.
And if you haven't written anything yet — which in a crisis is normal — it drafts the statement from the facts you gave it, tests that, and revises it on what the test found.
An illustrative example, written to the shape of a current Crisis Decision Report — the situation as a customer would type it, and the call as the report delivers it.
A batch of our chilled dip tested positive and has to come off shelves. About 9,000 tubs shipped to grocery stores in three provinces between the 8th and the 14th. Our largest retailer knows and is clearing it overnight; the inspection agency has been notified. We have had no illness reports. We have no statement and we need one tonight.
Publish tonight, and lead with the batch codes. The audience that matters in the first twelve hours is the person standing in their kitchen holding the tub, not the trade press. Open with the batch codes and best-before dates, then the symptom to watch for and how long it can take to appear, then what to do with the product and how the refund works. Say what you know, and name what you do not know yet. Name the inspection agency you notified, and give a direct line with a person's name on it. Cut the sentence beginning “we take food safety seriously” — every segment read it as the company talking about itself while they were working out whether their child will be ill.
Written for the person who has to decide in the next hour — the call first, then everything it rests on.
Six seats read the reactions, and each exists because without it a specific way of being wrong would go unnoticed. Where they disagree, you see the disagreement — never an average.
Well-prepared organizations keep a crisis firm on retainer and run a tabletop exercise once a year: weeks of scripting, a half-day in a boardroom, and a cast made up of their own staff. Everyone else sends the version written at 9 p.m., by someone who has now read it forty times.
Describe what happened. PRISM builds the audience — the customers, the staff, the reporter, the regulator, the real public bodies it looks up — drafts the statement if you don't have one, tests it, and revises it on what the test found. You read the briefing, change a word, and send.
Industry estimates from published sources; see Compare for every figure and where it came from.

A few sentences in your own words, a draft if you have one, a link if the conversation is already out there. A short guided conversation fills in the rest — mostly by clicking.

A simulated population built from your stakeholders — customers, employees, residents, journalists, regulators — including the real public bodies your situation pulls in.

The audience reacts to the actual material. A named panel of AI reviewers argues about what the reactions mean, and a reviewer whose job is to break the draft attacks the conclusion.

The call, the evidence, the disagreements, what would change the call, and an honest confidence reading. Read it, share it from your own portal, download it, or listen to it.

Run the recall language past the customers, the regulator, and counsel before it's published — and find the phrase that reads as blame-shifting.

Say what you know, what you don't, and what you're doing — tested against the people whose data or service it was.

See how staff, families, the community and the press are likely to receive hard news before you deliver it — and what they'll need to hear next.

Surface the sentences that read as deflection or hollow contrition before they end up on the evening news.

Mandated notices — a boil-water advisory, a privacy breach, a safety finding — tested against the public and the body that will be asked about it.

A reporter has called. Test the holding statement, the full statement, and silence — and know which one the room supports.
Every finding in the report names the real-world event that would reverse it. After you act, those become simple follow-up questions — did that happen? — and your answers build a track record of how the briefing compared with reality.
When the situation moves, run it again: the new report opens with Since your prior run — what got resolved, what's new, and how confidence moved and why. And one click rewrites your statement with the report's refinements and re-tests it.
Read it in the app, publish it to your own branded portal, share it by secure link, download the PDF, or listen to the audio briefing on the way to the meeting.
Every name, description and price below is read from the same catalog the app bills from, at the moment you load this page. Change a product in the back office and this page changes with it.