Customer-facing systems - often the most important in the business - are vulnerable to failure. Tim Stanley of Computer Configurations Holdings subsidiary Business Recovery Services says it is mandatory that companies implement disaster recovery options for the call centre.
"Call centres are not only particularly vulnerable to failure, but when disasters occur, they place an immediate increased load in the number of inbound calls from customers who need reassurance that their interests have not been affected," warns Stanley. "This places an additional onus on the company to provide adequate backup as panic can lead to severe knock-on effects."
The convergence of IT and telephony brings with it attendant weaknesses. In South Africa we have the ubiquitous theft of copper telephone lines. Other risks are flooding and fire. Hardware can crash, software can fail and accidents and natural disasters do happen. Even inadequate planning can precipitate disaster.
"Typically, call centres grow organically with the business they service," observes Stanley. "A mature call centre can consume an investment of millions in leading-edge technology, personnel training, and constant upgrading and refurbishment. It makes sense, therefore, to plan for continuity while the call centre is still in its early development phase."
When developing a business continuity plan, business processes are analysed, the points of contact with technology identified and the extent to which a business is reliant on that technology infrastructure assessed. The risk is then quantified financially by calculating the cost associated with a breakdown.
However, lost transactions are not the only cost consideration. For example, where a call centre is used for ordering, there will be a shortfall in stock, meaning that orders aren`t fulfilled. The knock-on effect will be to lose customers to more reliable sources of supply. Relationships are soured, negative sentiment can spread through the market and potential new clients can be put off.
With a service-oriented call centre with no direct revenue generation, the risk is difficult to quantify. For example, cellphone customers whose complaints and queries are not attended to may move to a competitor. Risk analysis, therefore, must attempt to quantify that potential loss.
"The next step," advises Stanley, "involves identifying those critical resources and contact points which demand minimal downtime. Those identified as priority will not necessarily include all in/outbound functionality, such as IVR (interactive voice response) and predictive dialling. Each point of interface needs to be analysed and quantified."
Some very simple contingencies can be built into the call centre itself to provide elements of continuity:
. Integrated database systems with a manual data retrieval function will allow for an alternative method of retrieving required information should the CTI (computer telephony integration) link become unavailable. This CTI link would normally be responsible for the automated lookup of caller information. Integrated systems which completely exclude a manual retrieval option can result in the company`s having to find a disaster recovery site with an identical switch or adopting a less than satisfactory disaster recovery plan.
. The "virtual" call centre links sites in different locations via ISDN lines. The inbound caller remains ignorant of the existence of more than one centre as he is switched between locations. The costs, however, become prohibitive as the telephony infrastructure has to be replicated. One should also consider the impact of additional overflow calls being addressed by an already busy alternative call centre environment.
. Consider an ISDN line. In addition to the advanced digital services ISDN can offer, in the event of a disaster the re-direction of the ISDN number range can be completed in 30 minutes, while an 0800 or 0860 number will take up to 24 hours to redirect.
"Disaster recovery should be part of an organisation`s ongoing risk management strategy. Technology and business processes are in a constant state of flux. Any change should include a check for the possible need for updating the disaster recovery plan," concludes Stanley.
Share