Eight questions. These are the ones we are asked.

Eight competencies are measured: crisis composure, decision quality, prioritization, data literacy, stakeholder management, written communication, ethical reasoning and self-awareness. Each one gets a whole number from 1 to 5, and the overall figure is their average. Every score comes out of what the candidate did in the simulation: which email they picked up first, which way they went in the crisis, what they wrote in the press statement. No part of it reads a CV, a university or a previous job title.

Every score sits next to the thing that produced it: a sentence the candidate wrote, or a call they made. So the report does not merely say "weak written communication" — it shows you the line that judgement rests on. In the press module the quote is additionally checked word for word against what the candidate actually wrote; if it is not there, no report is produced at all. An invented quote is worse than no report, because it survives being challenged when it should not.

The candidate gives explicit consent before the assessment starts, and the consent text names what is processed: their name and contact details, their answers and decisions in the simulation, and the transcript, scores and evidence quotes that come out of them. The data is processed only to run and report the assessment, and no candidate answer is ever used to train an AI model. How long it is kept is set in your contract with us. A candidate can withdraw consent at any time and ask for their data to be deleted. The detail is on the security and privacy page; our legal texts are currently drafts awaiting legal review.

You build an assessment in the panel wizard: the role, the scenario, the modules. The assessment is born with its own invitation link, a tokened address beginning /a/. You then add the candidate to it by email address, copy the link and send it to them yourself — no email goes out from here. When the candidate opens the link they have to enter the address you listed; any other address and the assessment will not start. So even if the link gets forwarded, the only person who can get in is the one you invited.

Three counters. Candidate invitations per month: how many candidates you have invited within the calendar month, which falls back to zero on the first of the month on its own. Open assessments: every assessment your workspace holds — there is no archived state, so at the ceiling you delete one to open another. Seats: people with access to the workspace, counted per workspace, so the same person in two of them costs a seat in each. The ceilings come with your package and are set in your contract; a workspace that is not on a package has no ceiling at all.

Two ways. You can run the candidate's side of it without creating an account: the demo is a shortened version of the real assessment, takes about 15 minutes — the real one runs 75 — and never leaves your browser. It produces no score, and your answers go nowhere. If you would rather talk through the panel, the report and the process against a role you are actually hiring for, fill in the demo form and we will set up a forty-five minute call.

It will not. No email goes out from here, so there is no such thing as a reset link. Your password is reset by the owner of your workspace or by our support team: you are given a temporary password, shown once, which you sign in with and then replace with your own. A reset also ends every open session, so you are once again the only person who knows the password. Leave your request on the password recovery page or in the form here.

No. The simulation puts an inbox, a dashboard and a running clock in front of the candidate at the same time, and those cannot sit side by side on a phone without one of them getting less room than it deserves. If a candidate opens the link on a phone the assessment does not start; they get a screen asking them to come back on a computer. A current browser on a desktop or laptop is all that is needed: nothing to install, no extension, no camera.