Barclays improves call centre service and cost-efficiency with Twinn Witness

The challenge

Determining the best way to deploy staff and structure shift schedules to deliver the right service levels

BMS’ call centre supports retailers with PDQ equipment issues and financial queries. It also answers new members’ procedural questions and handles responses to merchant recruitment drives. 

To provide the right level of service, BMS needs to know when and how to deploy employees during core volume shifts.

• What’s the best way to account for call patterns and peak hours?
• How do they ensure phones aren’t over- or under-manned – while meeting specified service levels and managing costs?
• How long should shifts be?

To answer these questions and test different theories, BMS wanted to simulate call centre operations. They looked at various solutions, prioritising functionality together with support and training for the staff who would use the product. “After seeing 3 demos of different products, we chose Twinn Witness (formerly under the Lanner brand),” said Mike Tubb, Continuous Improvement Team Manager at BMS.

Witness has been used by all major UK banking groups, including Midland, Lloyds, NatWest Life Assurance, Bank of Scotland, and a number of building societies and insurance companies, such as Prudential Assurance.

“We chose Witness because of its capabilities and reputation for support,” Mike added.

The solution

Using predictive simulation to test different shift patterns and understand the implications for service levels and cost

Using Twinn Witness predictive simulation software, you can create models that replicate real-life situations and experiment with multiple what-if scenarios to find the best solution. Accounting for everything from call volume variations and customer service to resource utilisation and cost, Witness helped BMS understand the implications of different process changes in a risk-free way.  

BMS started with a pilot to validate the Witness model’s reliability. This involved predicting the helpline’s service requirements for the following month. The model’s forecasted figures were remarkably close to the actual service provided, Mike confirmed. “The trial was essential because we needed robust proof of its reliability as a re-engineering tool.”

Initially, the model accounted for changing call volumes throughout the week. Following the successful pilot, BMS fleshed it out to include assumptions about agent skills and effectiveness as well as the availability of systems that support agents. Then, they tested out different theories and scenarios for manning the call centre. 

The impact

High service levels with fewer man-hours

Thanks to the Witness simulation, BMS discovered they could achieve specified service levels with fewer man-hours than they had predicted. 

“We were able to reduce overtime costs by between £80,000 and £100,000 a year while still fulfilling the service promise to our customers,” said Mike. “Witness has become a vital tool in our customer service and budgeting processes – identifying resource requirements for varying scenarios.”

He continued: “Our ultimate aim is to model the whole business. We picked the merchant support centre first because correct resourcing of our frontline units for excellent customer service is our number one priority. Now that we can measure the impact of change in an environment that mirrors the true one, we can avoid experimenting with the risks and costs of people in real-time. And where we can calculate an ideal shift pattern but may not be able to implement it immediately, we can use Witness to measure the implications.”

According to a Business Consultant at Twinn, BMS has been continually surprised at Witness’ flexibility. “Witness models scenarios that BMS couldn’t model any other way – and answers questions they couldn’t answer before. Because the model is so flexible, we can throw any imaginable idea into the works to stress test operations. For example, if it was announced that VAT would double and panic buying followed in the shops, how many agents would be needed? What proportion of temps would be optimal, and where would they all sit? When would the call centre operation finally fall over? Witness can model all kinds of disasters and simulate the overflow.”

Mike summed up the value of Witness in one word: confidence. “What the software offers is confidence. You don’t need to worry because you’ve already modelled all the possibilities and know what will happen. It’s not just savings and improved service, it’s a good night's sleep!”

As a result of this success in the support centre, Barclays Merchant Services selected a second area to model in Witness – the authorisation call centre. Barclays Cardholder Services began using Witness as well.