Being antifragile means going beyond resilient

Below 50

Without knowing what cannot stop, how long it can stay down, and how much data can be lost, every continuity plan is a guess. Start with a single critical process: list what it depends on and what each day of downtime costs. The first version fits on one page and already guides decisions.

50 to 79

There is a sense of impact, but part of it lives in people's heads rather than in a formal agreement with the business. Turn it into numbers: maximum tolerable downtime and acceptable data loss per process, validated by whoever answers for the operation. That agreement is what keeps recovery investment from being too much or too little.

80 or above

Impact is mapped and agreed, one of the strongest areas in your answers. The next gain is keeping the map alive: revisit priorities and dependencies when the business changes, because an impact analysis ages along with the company.

Below 50

If no plan exists today, the good news is that the wrong first move would be writing a long document. Start with the essentials: for the most likely scenarios, who does what, in what order, with which contacts. A two-page playbook reachable outside the environment beats a manual nobody opens.

50 to 79

Plans exist, but with the typical gaps: generic scenarios, a single copy stored inside the very environment they protect, content that no longer matches the real systems. Check three things: the plan answers ransomware, survives the network being down, and describes today's environment. It is a one-day check that prevents the worst surprise.

80 or above

Concrete, reachable, current plans. The next refinement is integration: continuity, incident response, and crisis communication telling the same story, so nobody discovers a contradiction between documents mid-crisis.

Below 50

If an attacker with admin access can delete every copy, your backup protects against disk failure, not against the scenario that takes companies down today. An immutable or offline copy is the priority, followed by one end-to-end restoration rehearsal. Everything else can wait; this cannot.

50 to 79

Backups exist and pass isolated tests, but restoring a file is not the same as standing a process back up: restoration order, dependencies, and total time only show themselves in a full rehearsal. Also define how the operation runs in minimal mode while restoration is underway. That combination is what turns backup into recovery.

80 or above

Recovery looks mature in your answers: protected copies, rehearsed restoration, and a fallback way to operate. Keep the full rehearsal on a cadence and time it against the deadlines agreed with the business, because environments change and restoration times change with them.

Below 50

A plan without exercises is a hypothesis. The first test does not need to be big: two hours of tabletop, with the people who decide in the room, on a realistic scenario. What it reveals about decisions, communication, and gaps is usually worth more than months of plan writing.

50 to 79

Exercises happen, but under friendly conditions: known date, systems up, everyone available. The leap is getting closer to reality: less notice, one key person out, one system genuinely unavailable. And record the outcome in numbers, so the next exercise has something to be compared against.

80 or above

Regular, measured exercises, the heart of antifragility. The next step is widening the reach: scenarios with a critical supplier down, multi-day crises, and participation from areas that have not yet sat at the table.

Below 50

In a crisis without defined roles, the first hours are lost figuring out who decides, and those are the most expensive hours. Define three things on one page: who declares the crisis and on what criteria, who talks to customers and authorities, who has authority to shut systems down. Name backups in the same act.

50 to 79

Roles exist but are concentrated in a few people, and availability is never guaranteed in a crisis. Name a backup for every role and secure a communication channel that works with the environment down. The test is simple: if your two most important people are unreachable, does the plan still move?

80 or above

Clear roles, with backups and an alternate channel. The refinement is training the judgment: rehearse the hard calls, like shutting down billing as a precaution or communicating before you are certain, so the first time is not live.

Below 50

Without a review after an incident or exercise, the company pays the price of the failure and throws away its product: the lesson. Start with a short meeting and one rule: hunt for causes, not culprits. The question that organizes everything: what do we know now that we wish we had known before?

50 to 79

Reviews happen, but the actions that come out of them compete with routine and lose. An owner, a deadline, and a check that it happened is what separates a lesson learned from a lesson noted. One cheap warning sign: the same failure appearing in two consecutive exercises.

80 or above

The learning cycle works, and it is what makes the company come out of each round stronger. The next gain is making the curve visible: show leadership the evolution between cycles, because readiness leadership can see is readiness that keeps its budget.