When to use Virtual Hold Technology in Call

When to use Virtual Hold Technology in Call Centre Operations Dave Worthington, Chris Kirkbride Department of Management Science, Lancaster University, UK. d. worthington@lancaster. ac. uk c. kirkbride@lancaster. ac. uk (Thanks to Zubin Sethuraman, MSc student, Lancaster University). OR 54 Conference, Edinburgh, Sept 2012.

2 Outline o o o Energy provider call centre context Previous work on balking queues Do our models fit? Model predictions Implications for call centre management. OR 54 Conference, Edinburgh, Sept 2012.

US Energy Provider: Daily call volumes to call centre (Winter) OR 54 Conference, Edinburgh, Sept 2012.

Hourly: call volumes (mins) & ASAs OR 54 Conference, Edinburgh, Sept 2012.

Hourly: ASAs (mins) & abandonment %s OR 54 Conference, Edinburgh, Sept 2012.

The problem posed o Virtual hold technology (VHT) gives callers opportunity to be called back without losing their place in the queue (accept /stay in queue/ balk) o When should they use VHT? o Should they use it differently for different queues? OR 54 Conference, Edinburgh, Sept 2012.

7 Reminder OR 54 Conference, Edinburgh, Sept 2012.

Time-dependent queues with balking o Balking can be used to represent balking and reneging o We investigate M(ln(t))/G/S and assume geometric balking, i. e. OR 54 Conference, Edinburgh, Sept 2012.

Results: Easy-to-use approximation o Queue length behaviour of M(ln(t))/G/S can be approximated (well) by a Normal distribution with: OR 54 Conference, Edinburgh, Sept 2012.

Self Validation Property o Approximation works well when Prob (n<S) is small, which we can check using standard Normal tables, e. g. Prob(n<S) < 0. 05 if mean is >=1. 645 SDs above S, i. e. : OR 54 Conference, Edinburgh, Sept 2012.

Results: Easy-to-use approximation o Anticipated Queueing Time (AQT): o Abandonment behaviour: OR 54 Conference, Edinburgh, Sept 2012.

‘Sub-optimal’ behaviour is frequent Impatience is a virtue! OR 54 Conference, Edinburgh, Sept 2012.

CARE: non-balking % V queue length/server 100% b*(off) =0. 8205 Non-balking % b*(on) = 0. 8578 50% 0. 0 Queue length/server 6. 0 OR 54 Conference, Edinburgh, Sept 2012.

ACQU’: non-balking % V queue length/server 100% b*(off) =0. 6773 Non-balking % b*(on) = 0. 7361 50% 0. 0 Queue length/server 2. 5 OR 54 Conference, Edinburgh, Sept 2012.

PAYG: non-balking % V queue length/server 100% b*(off) =0. 7807 Non-balking % b*(on) = 0. 8091 20% 0. 0 Queue length/server 6. 0 OR 54 Conference, Edinburgh, Sept 2012.

Performance prediction tool: … based on: ASA OR 54 Conference, Edinburgh, Sept 2012.

ASA analysis: CARE calls OR 54 Conference, Edinburgh, Sept 2012.

ASA analysis: CARE calls OR 54 Conference, Edinburgh, Sept 2012.

ASA analysis: CARE calls OR 54 Conference, Edinburgh, Sept 2012.

ASA analysis: PAYG calls OR 54 Conference, Edinburgh, Sept 2012.

ASA analysis: ACQUISITION calls OR 54 Conference, Edinburgh, Sept 2012.

ASA analysis: What if ……. . 1? OR 54 Conference, Edinburgh, Sept 2012.

ASA analysis: What if ……. . 2? OR 54 Conference, Edinburgh, Sept 2012.

ASA analysis: What if ……. . 3? OR 54 Conference, Edinburgh, Sept 2012.

Management implications o VHT is not necessary for Acquisition calls o For other call types VHT reduces abandonments and therefore increases ASAs! o So do current results suggest that VHT is a bad idea? n Depends how many customers accept VHT n Maybe depends on call centre agent allocation software o And …………………? OR 54 Conference, Edinburgh, Sept 2012.