Building Your Business Continuity Plan (BCP)
In this episode, we are building a business continuity plan. Continuity focuses on keeping prioritized business activities operating at an acceptable level during disruption, even when normal systems or facilities are unavailable. We will connect workarounds, people, suppliers, communications, dependencies, and technical recovery into a coordinated plan.
Imagine a fictional pharmacy network called Summit Pharmacy Network experiencing an outage in its identity provider. Core applications remain technically available, but staff cannot authenticate normally. The business continuity plan identifies which patient-support activities must continue, which functions can use controlled manual procedures, how supervisors approve temporary access, how records are reconciled later, and when the situation must escalate. The disaster recovery team works on restoring the identity service, while business leaders manage staffing, communications, suppliers, and alternate work methods. This fictional scenario demonstrates the boundary between continuity and technical recovery. Continuity is concerned with sustaining prioritized operations at an acceptable level. A strong plan therefore includes people, facilities, information, suppliers, communication, authority, and workarounds, while staying connected to the recovery activities that will return the organization to normal service.
Business continuity and disaster recovery support one another, but they solve different problems. Business Continuity (BC) is the discipline of keeping critical business services operating at an acceptable level during disruption, even when conditions are degraded. It is concerned with what the organization continues to do, for whom, and at what minimum service level, not merely where systems run. BC focuses on people, process, facilities, third parties, and communications as much as technology, because customers experience outcomes rather than server uptime. In exam terms, BC is about sustaining mission-essential operations across a range of events, including regional outages, labor shortages, supply chain interruptions, and cyber incidents.
Disaster Recovery (DR) is a subset of BC that concentrates on restoring technology services after an event causes a significant loss of capability. DR is narrower in scope because it primarily addresses information technology assets, such as applications, data, infrastructure, networks, and identity services. The DR question is usually “how do we recover the platform and data,” while the BC question is “how do we keep delivering the service while recovery is underway.” The boundary matters because a strong DR plan can still fail the organization if business operations cannot function during the recovery window.
A common exam trap is treating BC as a synonym for backup and restore, as though continuity is a storage problem with a schedule. Backups, snapshots, and replication help DR, but BC depends on operational workarounds and decision rights that exist outside the data center. If the call center cannot authenticate agents, if the warehouse cannot print labels, or if procurement cannot approve emergency spend, continuity fails even when backups are perfect. The best answer in a boundary question typically recognizes that DR supports BC, but BC sets the business requirements that DR must meet.
The boundary becomes clearer when you separate “service” from “system” and force yourself to describe what the business actually provides. A payment service, for example, might be delivered through multiple systems, but continuity can be achieved through alternate channels, manual fallback, or temporary throttling to a safe operating mode. DR, by contrast, might define how the payment gateway is rebuilt, how encryption keys are restored, and how transaction logs are reconciled after failover. On an exam item, if the scenario asks about maintaining customer-facing capability during disruption, you are usually in BC territory even if technology is involved.
Business Impact Analysis (BIA) is the bridge that turns the boundary into measurable requirements rather than organizational opinion. It identifies critical business processes, dependencies, and impact over time. As an earlier lesson explains, MTD is the business disruption limit, RTO is the recovery target that should leave time for necessary business resumption, and RPO specifies the required recovery point or acceptable data-loss interval. The business owns the impact assessment and approves the requirements; recovery teams design and test the technical methods to achieve them.
Consider an exam-style scenario: ransomware takes down the enterprise resource planning platform and file shares, but the organization still needs to ship orders and collect payments. A BC response might include invoking predefined manual order processing, switching to alternate fulfillment processes, using preprinted pick lists, and authorizing emergency procurement for temporary logistics support. A DR response would focus on isolating affected networks, restoring clean systems from immutable backups, rebuilding identity services, and validating data integrity before reconnecting production. The boundary is visible in the outcomes: BC preserves essential operations, while DR restores the technology ecosystem that supports full operations.
The boundary also clarifies roles and ownership, which is a frequent “best answer” discriminator. BC typically sits with business leadership and risk governance because it requires prioritization decisions, acceptance of degraded service, and customer or regulatory communications. DR commonly sits with technology leadership and security operations because it involves architecture, runbooks, restoration sequencing, and technical validation. When a question asks who should approve continuity priorities or define acceptable downtime, the most defensible answer is a business owner with governance authority, not an engineer with a recovery script.
A final boundary that shows up in exam questions is the difference between “alternate processing” and “alternate recovery.” Alternate processing is a continuity concept where the business continues the service in an alternate manner, such as shifting work to another site, using a third-party service, or operating in a reduced-capability mode. Alternate recovery is a DR concept where the technology stack is reconstituted at a warm site, hot site, or cloud environment according to recovery design. If the scenario is about sustaining operations despite missing systems, you are usually being tested on BC; if the scenario is about restoring systems and data after a major outage, you are usually being tested on DR.
The exam also distinguishes boundaries by looking at triggers, time horizons, and validation criteria. BC triggers can be based on business impact thresholds and can start before systems are fully down, because early action reduces operational loss. DR triggers are often tied to declared disasters or confirmed loss of key technology capabilities, and success is measured by restored services meeting defined performance, integrity, and security controls. If a question asks what you validate after restoration (such as data consistency, access control, logging, and configuration baselines) you are being pulled toward DR; if it asks what you keep operating under constrained conditions, you are being pulled toward BC.
Continuity and recovery commitments must align with criticality and the organization’s risk appetite. Alignment begins when risk appetite is translated into specific tolerance statements that responders can act on under pressure. Many organizations say they have a “low appetite for downtime,” but that is not a usable requirement until it is tied to business impacts and time horizons. Exam questions often reward recognizing that risk appetite is set by senior leadership and expressed through governance decisions, not by individual project teams. When continuity and recovery plans reflect those decisions, they become consistent across the portfolio instead of being negotiated service by service in the middle of a crisis. The best answer usually links resilience requirements to business objectives and accepted risk, not to the preferences of the loudest stakeholder.
Criticality is the sorting mechanism that prevents the organization from trying to recover everything first, which is a common exam scenario. Critical services are those whose loss rapidly produces unacceptable harm, such as safety impacts, legal or regulatory exposure, major revenue loss, or severe customer trust damage. Less critical services may be painful to lose but remain tolerable for longer periods, especially if workarounds exist. The exam often tests whether you can distinguish “important” from “critical” using time-based impact rather than emotional urgency. In best-answer logic, criticality drives sequence, and sequence drives realistic recovery design.
A Business Impact Analysis (BIA) provides the structure to connect risk appetite to criticality in a way that produces measurable recovery requirements. The BIA identifies business processes, dependencies, and impact over time, and it produces targets like Recovery Time Objective (RTO) and Recovery Point Objective (RPO) that are grounded in business tolerance. MTD states the business disruption limit. RTO states the agreed recovery target within that limit, allowing for any required business resumption time, while RPO states the required data recovery point. Those two targets are not purely technical; they are risk decisions expressed in operational terms. Exam questions frequently hinge on remembering that the business defines acceptable loss, and technology implements the methods to meet it.
Risk appetite alignment also means accepting that not every service can be engineered to “always on” without disproportionate cost and complexity. For high-criticality services, appetite may demand costly strategies like multi-region architectures, automated failover, and continuously validated backups with strong isolation. For lower-criticality services, appetite may allow manual restoration, longer outages, or limited functionality modes, provided those limits are explicitly accepted and communicated. Exam items often present a budget constraint and ask for the best response, and the strongest answer is usually to align targets with business priorities rather than to pretend priorities can be met without resources. This is where risk acceptance, not wishful thinking, becomes a governance act.
A subtle alignment issue is that risk appetite must account for different types of disruption, not just “a data center outage.” Cyber incidents can create conditions where rapid restoration is unsafe because the environment is compromised, data integrity is uncertain, or credentials are untrustworthy. In those cases, risk appetite may prioritize containment and integrity over speed, accepting longer downtime to prevent reinfection or fraud. Exam questions sometimes imply this by describing ransomware, destructive malware, or insider manipulation, and the best answer often emphasizes verified clean restoration and controlled reintroduction of services. Aligning appetite to scenario type prevents the organization from optimizing for the wrong success metric.
Criticality alignment also requires defining minimum viable service levels, because continuity is not binary. A customer portal might be down, but orders might still be accepted through a call center; a trading platform might be degraded, but risk limits and settlement must remain intact. These degraded modes must be pre-approved, because they often involve temporary process changes, policy exceptions, or altered customer communications. The exam often rewards recognizing that continuity strategies should prioritize essential outcomes rather than identical user experience. When plans specify what “good enough” looks like for a period of time, responders can act decisively without inventing policy during the incident.
Another exam-relevant alignment point is that recovery priorities should be consistent with risk ownership and accountability. If a business unit insists its application is “critical,” governance expects evidence through BIA results, dependency analysis, and impact justification, not just internal urgency. Recovery sequencing should be approved through a governance process so that, during a crisis, responders are not forced into political negotiation while the clock is running. Exam scenarios that include conflict between departments often point to missing governance decisions about criticality and order of restoration. The best answer is usually to reference pre-established prioritization based on business impact and risk tolerance.
Risk appetite alignment also extends to third-party dependencies, because criticality often concentrates in services the organization does not directly control. If identity, cloud hosting, payment processing, or customer communications are vendor-delivered, the organization’s resilience is bounded by vendor capabilities and contractual commitments. Appetite may require stronger assurances, such as redundancy across regions, additional vendor due diligence, or negotiated service-level commitments for support and recovery. Exam questions often test whether you recognize inherited risk and the need to manage it through architecture, contracts, and monitoring. Plans that assume vendors will always be available on demand are misaligned with real-world constraints and are weaker exam answers.
Finally, testing compares demonstrated capability with the requirements approved by the business. A missed RTO or RPO exposes a capability gap; it does not redefine risk appetite or make the business disruption limit longer. Escalate evidence and options to the appropriate risk owner. Changing objectives requires a justified reassessment of business impact and applicable obligations, not simply an easier target for the next test. Exam questions frequently ask what to do when recovery objectives are unrealistic, and the best answer is usually to escalate to risk owners with evidence and options. Alignment is therefore a living governance loop: define tolerance, design to meet it, test to verify it, and correct the mismatch through investment or acceptance. When that loop exists, continuity and recovery plans reflect real risk decisions rather than aspirational statements.
Restoration requires disciplined communication so stakeholders understand progress, risk, and remaining uncertainty. Restoration is the phase where the technical work becomes most visible, because systems are changing rapidly and business teams feel the effects in real time. Stakeholder communications during restoration is not a public relations exercise; it is operational control. When communications are vague, inconsistent, or overly optimistic, leaders make poor decisions, users improvise workarounds, and the incident expands through self-inflicted risk. The purpose of restoration communications is to align expectations, coordinate actions, and support risk decisions with clear, bounded information.
A communications plan during restoration starts by recognizing that different stakeholders need different truths at different levels of detail. Executives need business impact, decision points, and risk statements expressed in plain language, not a stream of indicators and tools. Technology leadership needs dependency-aware timelines, resource constraints, and clear definitions of what “restored” means for each service. Legal and compliance teams need facts, not speculation, and they need to understand what is confirmed versus what is suspected so they can manage reporting obligations. Frontline staff need practical guidance on what to do now, what not to do, and when the next update will arrive.
A scenario shows how messaging can either reduce or amplify risk: during recovery from a suspected credential compromise, an organization restores access to email and collaboration tools but tells users to “resume normal operations.” Within hours, users log in from unmanaged devices, create ad hoc file shares, and forward sensitive documents to personal accounts to keep work moving, because “normal” means “make it work.” The incident team later discovers that conditional access policies were temporarily relaxed to accelerate restoration, increasing exposure precisely when the organization needed stricter controls. The communications failure was not the lack of technical action; it was failing to define the temporary rules of engagement. In exam terms, the best response communicates constraints and compensating controls, not just availability.
Effective restoration updates are structured around decisions and dependencies rather than around activity. Stakeholders rarely benefit from hearing that responders “rebuilt servers” or “rotated keys” unless those actions map to a specific risk reduction or service milestone. Leaders do benefit from understanding whether restoration is gated by eradication confidence, whether monitoring is fully restored, and whether a return-to-service decision includes residual risk. Communications should clearly separate what is complete, what is in progress, and what is blocked, using consistent definitions across updates. Consistency prevents the classic situation where one team says “systems are back” while another says “systems are unstable,” and both are technically correct from their vantage point.
Precision language matters because restoration often includes uncertainty, and uncertainty can be communicated responsibly without creating panic. Words like “clean,” “secure,” and “all clear” imply a level of assurance that is rarely justified in the middle of incident response. Better language distinguishes functional status from security status, such as “service is available with restricted access,” or “service is restored while enhanced monitoring remains in effect.” Communicators should also label what is confirmed versus what is suspected to avoid accidental misrepresentation that later creates credibility and legal problems. The exam-reality wink here is simple: overconfidence is expensive, and it is rarely rewarded.
Restoration communications must also include behavioral guidance to prevent well-meaning users and administrators from undermining recovery controls. If passwords are being reset or tokens revoked, users need to know what authentication prompts are expected and what patterns should be treated as suspicious. If certain tools are temporarily disabled, users need sanctioned alternatives, not a blank space that invites shadow IT. If administrative access is restricted, operations teams need a clear path for break-glass procedures, approvals, and logging expectations. These instructions should be operationally realistic, because guidance that users cannot follow becomes guidance they ignore.
There is a governance component to restoration updates that supports accountability and auditability. Key decisions should be recorded with the rationale, including who accepted residual risk, what compensating controls were implemented, and what criteria will trigger changes in posture. Communications should reinforce that emergency changes are still changes, meaning they must be tracked, reversible where feasible, and reviewed afterward. This is especially important when restoration includes temporary network rules, elevated privileges, or exceptions to standard identity controls. Clear documentation prevents the organization from carrying incident-era shortcuts into steady state unnoticed.
External communications often run in parallel with internal restoration updates, and coordination is essential even when the organization is not ready to disclose details. Customer-facing teams need accurate service status and constraints so they do not create false expectations or contradict internal guidance. Vendors and partners may need targeted notifications if integrations are affected, credentials are rotated, or traffic patterns change in ways that could be misinterpreted as abuse. Regulators and oversight bodies may require timely notice under certain circumstances, and premature or speculative statements can create compliance and reputational harm. The best approach keeps external statements aligned to confirmed facts and clearly separates incident investigation from service restoration.
Finally, restoration communications should be timeboxed and predictable, because predictability reduces anxiety and reduces disruptive “status chase” requests that consume responder time. A cadence such as scheduled updates, clear ownership for messaging, and defined escalation paths helps stakeholders plan their work and reduces rumors. Predictability also forces discipline in measuring progress against return-to-service criteria, rather than narrating effort. When stakeholders know when they will hear next and what the update will contain, they stop inventing their own story. In a high-pressure restoration window, that alone can be a meaningful risk control.
Now apply the idea to the management decision at hand. A staffing model is a risk control as much as it is an operating decision, because it determines whether security outcomes are repeatable or accidental. When an organization relies on heroics, it may still pass an audit or survive an incident, but it cannot reliably demonstrate that the same result will happen next quarter. For exam purposes, the key idea is that staffing should be driven by required capabilities and accountability, not by headcount targets or org charts. A mature program starts by translating objectives such as vulnerability reduction, incident containment time, and access governance into roles that can be assigned, measured, and sustained.
The most common staffing mistake is confusing “having people” with “having coverage.” Coverage means the right functions exist across prevention, detection, response, recovery, and governance, with continuity across business hours, time zones, and planned absences. A small team can achieve strong coverage if roles are defined and reinforced with procedures, but an oversized team can still fail if duties overlap randomly and nothing is clearly owned. Leaders should look for single points of failure in people, not just systems, because a sole specialist with undocumented knowledge is operational fragility. A well-designed staffing model makes the organization resilient to turnover by design rather than by luck.
A staffing model also needs to address the difference between strategic work and interrupt-driven work, because interruptions consume capacity in ways that are easy to underestimate. If the same engineers expected to deliver long-term improvements are also the default escalation point for alerts, audits, and executive questions, then strategy work becomes “whenever there is time,” which means it does not happen. Mature programs protect time for design, automation, and control improvement while still maintaining operational response coverage. This may involve rotations, dedicated operational roles, or a service model where routine requests are handled through standardized intake and triage. The goal is to avoid the hidden tax where a few high-skill individuals spend their week answering the same questions repeatedly.
Use the design in an earlier program-development lesson to make capability gaps concrete. The retailer may already operate an identity service but lack capacity to integrate the acquired application, support supplier access, and test recovery. Compare required work with available skills, coverage, and workload before choosing recruitment, training, automation, or a partner. Name an owner and a completion measure for each gap. A proposed design becomes an executable program only when those commitments are realistic.
Now apply the idea to the management decision at hand. Escalation is a control when it is the deliberate mechanism that prevents a risky decision from being made at the wrong level, with the wrong context, or without the right accountability. In risk governance, many failures do not happen because people are careless; they happen because decision authority is misaligned with business impact. A project manager may be empowered to approve schedule tradeoffs, but not to accept regulatory exposure or systemic operational risk. Escalation exists to correct that mismatch by moving the decision to an owner whose role includes accepting that class of risk. On the exam, the “best answer” often emphasizes proper governance and decision rights rather than trying to solve a governance problem with a technical workaround.
Escalation is most defensible when it is criteria-based and pre-defined, not ad hoc or personality-driven. If escalation only happens when someone feels uncomfortable, the organization ends up with inconsistent decisions and political friction. Criteria should be tied to business impact, such as potential outage duration, exposure of sensitive data, compromise of privileged access, breach notification implications, or violation of contractual obligations. The decision to escalate should be as procedural as the decision to patch, deploy, or onboard a vendor, because the control objective is repeatability and evidence. A mature organization treats escalation thresholds like guardrails that keep routine teams within their authority while ensuring material risks reach senior ownership.
A common exam-realistic scenario is a high-urgency change request that would weaken segmentation to restore service during an outage. The operations team argues that the business cannot wait, and the security engineer sees a clear exposure that could enable lateral movement or unauthorized access. In this situation, escalation is the control because it forces the organization to weigh immediate availability impact against increased compromise risk, using the risk appetite and business priorities that senior leadership owns. The engineer’s job is to present the risk clearly, propose safer alternatives if available, and ensure the decision is made by an authorized risk owner. The “best answer” typically is not “deny the change” by default, but “escalate and document a risk-based decision” with time bounds and compensating controls.
Escalation also functions as a control when it interrupts the normalization of deviance, where repeated exceptions slowly become the standard operating model. Small “temporary” deviations (skipping testing, disabling alerts, delaying patching, granting broad access for convenience) accumulate until the environment’s real risk posture is far higher than the documented one. Escalation triggers create a forcing function that says, in effect, “we are outside the guardrails; someone with the mandate to accept this must own it.” Without escalation, exceptions become silent risk transfers to future teams, future budgets, and future incident responders. Exam questions that mention recurring exceptions, chronic delays, or “temporary” workarounds often point toward escalation as the control that stops drift.
The most important feature of escalation is not the meeting it creates, but the accountability it produces. Proper escalation assigns a named risk owner, defines what is being accepted or deferred, and records the rationale in business terms. It also attaches conditions, such as compensating controls, reduced scope, monitoring requirements, and a deadline for remediation or reevaluation. When these elements are missing, escalation becomes theater: everyone “knows” there is risk, but no one owns the consequences. For exam purposes, the presence of documented ownership and time-bounded treatment is a strong indicator of a correct governance action.
Escalation is also a control for conflicting priorities, especially when security, operations, and delivery teams each optimize for different outcomes. Security tends to optimize for reduced likelihood and impact of compromise, operations optimizes for uptime and stability, and delivery optimizes for feature velocity and deadlines. None of these priorities are wrong, but unmanaged conflict leads to unsafe compromises made under pressure. Escalation provides a structured way to resolve conflicts by tying decisions back to enterprise risk appetite and business objectives. On the exam, when you see a scenario where teams disagree and the impact is material, escalation to the appropriate governance body or risk owner is often the best answer.
Third-party incidents are another place where escalation becomes the control because the organization may have incomplete information and limited leverage. If a critical vendor reports a security event with uncertain scope, the enterprise must quickly decide whether to suspend integrations, rotate credentials, trigger contingency plans, or notify stakeholders. Those decisions can have significant business impact, and they often require executive involvement to balance continuity, legal exposure, and reputational considerations. Escalation ensures that the response is not solely a technical reaction but a managed business decision with coordinated communications and accountability. Exam questions that mention vendor events, contractual obligations, or ambiguous impact often test whether you recognize escalation as a governance control rather than a sign of failure.
Escalation should also be paired with clear documentation and communication paths, because a control that cannot move information quickly is not a control under real conditions. This includes knowing who can authorize emergency risk acceptance, who must be informed for regulatory or contractual reasons, and how decisions will be recorded for later review. The quality of escalation is measured by clarity and speed: the right people receive the right information in time to influence outcomes. When escalation is ambiguous, teams stall or act independently, and risk increases through delay and inconsistency. For exam logic, “escalate through defined channels with documented risk acceptance” generally outranks “handle it within the team” when impacts extend beyond the team’s authority.
Finally, escalation becomes a control only when it feeds improvement rather than recurring crisis. Post-escalation reviews should identify why escalation was necessary and whether upstream processes can prevent similar situations, such as clearer requirements, better change categorization, stronger vendor onboarding controls, or more realistic project gating. The objective is not to punish teams for raising risk, but to reduce the frequency of high-pressure exceptions that force executive decisions. Organizations that treat escalation as a normal part of risk governance tend to make fewer catastrophic mistakes, because they surface uncomfortable tradeoffs early enough to manage them. On the exam, this reflects governance maturity: escalation as a designed mechanism for aligning authority, accountability, and enterprise risk outcomes.
Return to the fictional scenario from the beginning of this lesson on building your business continuity plan (bcp). The strongest management response is not a slogan or an isolated tool purchase. It is a documented decision that connects business impact, authority, evidence, operating responsibility, and follow-through. Ask what must be true before the organization can proceed, who is accountable for making or accepting the decision, and how the result will be verified over time. That reasoning pattern helps on the CISM exam because it prevents a technically plausible action from being mistaken for the best management action.
For practice, choose a critical activity and describe how it would continue if its primary system, facility, supplier, or workforce became unavailable. Define the minimum acceptable service, manual or alternate methods, decision authority, communications, resources, reconciliation, and transition back to normal operations. Then connect the continuity steps to the technical recovery plan so the two efforts support rather than contradict one another.
