THE SECURITY BRUTALIST

Building the Consequence Map

A consequence map traces one question through a system, asking what the business loses if that system fails or gets compromised and whether the loss can be recovered. Most security diagrams describe a system at a single point in time, showing its components, its boundaries, and the known risks attached to each part, and they tend to go stale soon after the system goes live. A consequence map follows a chain of cause and effect. It starts with a single compromise and traces it step by step until it reaches something the business cares about, such as downtime, data loss, or legal exposure, so it asks what happens next once something breaks and where that chain ends.

Building the map produces a ranked list of systems ordered by what failure costs. Systems at the top produce outcomes the organization cannot recover from, while systems at the bottom hurt but stay contained. This list drives every later prioritization decision, and it relies on direct answers for each system, without probability estimates or maturity scores.

Security facilitates the process, and the answers come from the people who own each system. Each system needs a technical owner, a business owner, and a finance or legal stakeholder when it touches revenue, regulated data, or contractual obligations. Before classifying anything, ask an executive to define what “existential” means for this organization.

Start with a draft list of every system whose compromise would be significant, including authentication, customer data storage, payment processing, production deployment, financial reporting, source code, customer-facing services, and anything holding credentials, regulated data, or intellectual property. A first pass usually runs 15 to 25 systems, and each needs an assigned owner before scheduling. Give each system 30 to 45 minutes, keep a shared document open so everyone can see the answers as they are recorded, and confirm the classification before the session ends.

Owners consistently underestimate what happens when their own systems get compromised, so the facilitator’s main job is to counter that optimism. When an answer sounds too low, ask for specifics, such as what an attacker could do with the credentials of the most privileged account on that system; that question produces sharper answers than asking about consequence in the abstract. Tying the existential question to the specific answers already collected in the same session keeps the discussion concrete and makes the optimism harder to sustain.

Ask these questions in order, since each one builds context for the next.

  1. What does this system do? One sentence, business function, not architecture.
  2. What does it connect to, and what depends on it?
  3. What data does it hold or process? The categories that change the classification: credentials or access tokens, customer personal data, customer financial data, regulated data (health, legal, government), intellectual property, internal operational data.
  4. What happens if it is unavailable for one hour, one day, one week?
  5. What happens if its data is corrupted or destroyed, not just the system being down?
  6. What happens if its data is exfiltrated without any disruption? Who could be impersonated with it? What regulatory or contractual obligations are triggered?
  7. What does recovery look like today? Has the system ever been restored from backup? Has anyone timed a full restore?
  8. Is there any realistic scenario where compromise of this system produces an outcome the organization does not recover from?

Ask question 8 last, because people answer it more directly after they have walked through specific consequences than when it comes cold at the start.

A system is “Existential” if its realistic worst case is unrecoverable, which covers regulatory action beyond what the organization can absorb, permanent loss of data it cannot reconstruct, financial loss or liability beyond its capacity to survive, or damage that permanently destroys customer trust. Calibrate this to the organization’s actual financial and reputational capacity. “High-recoverable” systems are costly but survivable, with revenue loss measured in days or weeks, fines the organization can absorb, customer attrition that can be rebuilt over time, or operational disruption that needs real but limited recovery effort. “Low-impact” systems keep their consequences contained, hold no sensitive data, and recover quickly. When a system sits between existential and high-recoverable, classify it as existential, then rank within each category. The top of the existential list is where hardening and survivability work starts.

The output is a table with one row per system, covering system, function, data held, worst realistic scenario, classification, recovery status, and owner. Recovery status reads “tested with a date” when a restore has been run, “untested” when a procedure exists but nobody has run it, and “unknown” when no one can say.

A sample row reads as follows:

The customer auth service stores credentials and session tokens, and the worst case is credentials extracted, full customer impersonation, and a pivot to internal systems through service accounts. That system is existential, its recovery is untested, and the platform team owns it.

Some participants will push to add likelihood scores to the classification. Redirect that right away, since likelihood is easy to manipulate and hard to verify, and the map runs on consequence alone. A map built only by the security team is fictional, so the business and technical owners need to be in the room.

If you are just starting, run sessions for five systems only, namely authentication, customer data storage, payment processing if it applies, the production deployment pipeline, and source code. A clear map of five systems is more useful than an optimistic map of twenty, and coverage can expand each quarter.

As you go through the exercise, adapt it to your business. In regulated environments, the regulatory consequence of a breach dominates the existential classification, so systems touching regulated data, along with the systems controlling access to them, consistently land at the top. For SaaS companies, authentication, multi-tenant data isolation, and the production deployment pipeline usually classify as existential, since compromising any of them affects every customer at once.

Review the map quarterly for new systems, significant changes, or shifts in business context. Do a full refresh annually, and again after any major organizational change, such as an acquisition, entry into a new regulated market, a major architectural shift, or a new product line with a meaningfully different data or access profile.