Imagine a 300-seat centre in a mid-sized Indian city, handling a retailer’s customer calls. It is 11 am on a Monday, the busiest hour. A road crew digging nearby cuts the fibre line. Screens freeze, calls drop, the dialler shows everyone logged out. The site head’s phone starts ringing before he has stood up. This is a made-up site, but every part of what happens next is real, and it makes a good thread for explaining business continuity.

The first hour, with and without a plan

Let us put a number on the damage. Suppose 300 agents are on calls, occupancy is 80%, and average handle time is 6 minutes. Each agent is on calls for 48 minutes of every hour, which is 8 calls an hour. Across 300 agents that is 2,400 calls an hour. Two hours of outage means 4,800 calls not answered by this site. Those customers hear a busy tone or an apology, or wait. Meanwhile the service level for the day is sinking, the client’s own customers are getting angry, and agents are sitting idle, still on payroll.

In a centre without a plan, the next hour is improvisation. Someone phones the telecom provider. Someone else tries to inform the client but is not sure who the right contact is. Team leaders gossip with agents in the corridor. Nobody knows whether to send people home, hold them, or relocate them.

In a centre with a plan, the incident has a named owner, the telephony team switches to the backup link in minutes, and the client gets a short message within a defined time saying what happened and what is being done. That difference is what a Business Continuity Plan (BCP) buys.

Starting with what matters most

A BCP does not begin with backups. It begins with a business impact analysis: for each process, what happens if it stops, and how fast does the damage grow? A collections process may survive a day’s pause. A payments helpline for a bank may not survive an hour. Each process gets two targets, borrowed from IT disaster recovery:

  • Recovery Time Objective (RTO): how long the process can be down before the damage is unacceptable.
  • Recovery Point Objective (RPO): how much data you can afford to lose, measured in time. An RPO of 15 minutes means backups or replication must be fresh enough that at most 15 minutes of work disappears.

The targets are set by business need, then tested against cost. A one-hour RTO needs more redundancy, and costs more, than a one-day RTO. Clients often write these into the contract, so the BCP is not only your own idea of safe.

Finding the single points of failure

Go back to the fibre cut. The weak spot was one line into the building. A single point of failure is any component whose failure stops everything. Typical ones in a centre: one internet provider, one entry route for cabling, one power source, one server room, one telephony vendor, one person who knows how to reroute calls, one site handling a client’s entire volume. The BCP work is to list them honestly and decide, for each, whether to fix it, accept it, or cover it with a manual workaround.

Some standard answers. Two network providers, entering the building by different physical routes, so a single digging accident does not cut both. UPS and diesel generator for power, with the generator tested under load and fuel actually in the tank. Call routing that can send traffic to another site. Cloud-based telephony, which lets agents log in from elsewhere. Spare laptops and VPN capacity for work-from-home fallback, together with the security controls that go with it (clean desk, screen privacy, restricted data), which must be agreed with the client in advance, not improvised on the day.

Different disasters, different answers

A fibre cut is a technical failure. A flood, a cyclone or a transport strike is a people-access failure: the building may be fine but nobody can reach it. A cyberattack takes the systems away while the building and people are fine. A pandemic is a long, slow version of the access problem. Plans therefore are usually written around scenarios: loss of the building, loss of systems, loss of people, loss of a key supplier. Coastal and flood-prone cities need to think about weather warnings and moving shifts early; Indian monsoon and cyclone seasons are predictable enough to plan for. Cities with frequent power cuts or transport disruption have their own pattern. The point is not to imagine every event, only to cover the four kinds of loss.

People and communication

Most failures of continuity are not technical. They are about who tells whom. A good plan has a call tree with current phone numbers, a named incident commander, a rule for when to notify the client, a script for what agents say to customers if the outage is widespread, and an agreed place to meet if the building is evacuated. Staff safety comes before service levels, always. Also decide in advance how the team will be paid, rostered and transported during disruption, because those answers affect whether people turn up at all on the second day.

A plan that was never tested is a rumour

Documents get stale. Phone numbers change, the backup provider’s contract lapses, the generator’s battery dies. Plans need rehearsal: a desktop walk-through where managers talk through a scenario, a technical failover test where the backup link is really used, and occasionally a full simulation that moves agents and traffic. After every test and every real incident, write down what failed and fix it. Many audits and client reviews, including those under ISO 22301, the international standard for business continuity management systems, and ISO 27001 for information security, look for exactly this evidence: that you tested, recorded and corrected.

Back at our imaginary centre, the best outcome of the fibre cut is not that the line was restored in 40 minutes. It is that, by lunch, someone filed a note titled “what we change before the next time”. That habit is what continuity really means.

Figures and names in the examples are illustrative.

Leave a Reply

Trending