Business Continuity Gap Before Call Centre Consolidation
Calder Wealth Management is consolidating two call centres into a new single site, moving 120 client order-taking staff in six weeks. The business impact analysis sets a maximum tolerable disruption of four hours and an RTO of four hours for client order taking; client contracts promise 99.5% availability and 'best efforts' business continuity. The new site has one telecoms provider with no diverse routing; the generator has eight hours of fuel; only 40 of the 120 staff have remote system licences. A full recovery test took 11 hours because staff could not log in remotely and manual order logging required dual entry into two systems. The proposed manual workaround is for staff to email orders to a shared inbox, where two supervisors re-key them; the inbox is unencrypted. The go-live has a remaining budget of USD 300,000. The business says the test was 'only a test' and the manual workaround is 'temporary'. Insurance covers property damage but not business interruption. The board risk committee meets in two weeks.
1. Should the consolidation proceed in six weeks? What decision and conditions would you recommend?
Read the worked answer
Do not proceed unconditionally. The business impact analysis says the maximum tolerable disruption is four hours, but the tested recovery took 11 hours. That is not a test failure to be waived; it is evidence that the proposed operating model cannot meet the stated impact tolerance. Client contracts promise 99.5% availability and business continuity on a best-efforts basis, and order-taking is a regulated activity; an 11-hour outage could cause client harm, complaints and regulatory breach. The manual workaround is not a recovery solution because it is untested at volume and insecure. The decision should be made by the accountable executive with board risk committee visibility. Options: delay go-live until remote licences, telecoms diversity and a second test are in place; or proceed with a formal, time-bound risk acceptance that is within appetite, with conditions. A seven-hour gap is unlikely to be acceptable without a credible compensating plan. The business's 'only a test' argument confuses a test with a rehearsal; the test is the evidence.
2. Is the proposed email workaround acceptable as a temporary control? What would you require instead?
Read the worked answer
The email workaround should not be approved as proposed. Orders sent to an unencrypted shared inbox create data protection, fraud and conduct risk: no sender authentication, no integrity check, no audit trail, and a high chance of mis-keying. A shared inbox also lacks dual control. If a manual fallback is needed, use a secure portal or encrypted email, callback verification to a known client number, dual entry with maker-checker, a sequential log, and a clear cut-off. Test it at peak volume, not with a small sample. Limit it to low-risk orders and set a maximum duration. The workaround should be documented as a temporary control with a risk acceptance and an expiry date. Without these, the fallback may create a bigger loss than the outage it is designed to cover.
3. What actions, reporting and go/no-go criteria would you set?
Read the worked answer
Actions: procure 80 additional remote licences or a secure remote access solution; contract a diverse telecoms route and test failover; extend generator fuel to at least 24 hours or contract refuelling; run a second full test with 120 staff at peak volume and a target of four hours; update the business impact analysis and client contracts if the RTO cannot be met; report the gap and the go-live decision to the board risk committee; set go/no-go criteria and a rollback plan. If the second test fails, delay go-live. Track the remediation as a change risk with an owner. Notify clients only if required by contract or regulation, but prepare a communication plan. Insurance should be reviewed because business interruption is not covered. The committee should also decide whether to accept a temporary deviation in writing and for how long.