Call Centre Model
"We suggest that WINZ must reconsider its human resource management strategy on student loans and allowances processing. Rather than throwing large numbers of ill-trained staff at the problem, in a regimented, production-line model, WINZ should consider a different approach. The customer service officer model adopted by most banks might be preferable: a smaller number of staff with a sound training in all the systems, who can take responsibility for effective processing and advice. The institution helpdesk staff we dealt with provide such a model."
"It was inevitable that bottlenecks would occur and that the call centre would come under immense pressure as students sought information on the status of their application. WINZ failed to prepare for this predictable eventuality."
Application and other special forms, such as the "exemption" for limited full time status, and "extension" forms, had to be requested by phone. Often students received conflicting advice about which form to use. Sometimes it was not clear from the title of the form which one to use (eg "extension" or "summer extension"). They were not available on campus or always available at local DWI offices. Once these supplementary forms had been returned to DWI, some students were still being told their loan had been declined, as either operators answering queries could not deal with such complexities, or the system was not being updated quickly enough.
Students were discouraged from applying face to face, unless using the combination application form. As at 11 May 2000, 56% of applications were received by phone, 23% by post, 17% via Outreach centres, and 4% by internet. Students either had to phone in with their application or mail it. Faxed applications were not acceptable. Some students were able to apply by internet, although there were some problems with this service. One institution commented that file transfer by internet, such as is done by banks, would be more secure than by PKZip which is currently done by DWI. The institution also saw the internet as a move to the future, providing greater self service (primarily for institutions). There is no intention to review technology for transfer of data at present.
Assumptions About Work Flows
"WINZ's approach was "If they [students] did what they were supposed to do, it would have been OK.""
"WINZ once again failed to prepare for, and cope with, predictable peak loads (both phone traffic and processing bottlenecks)"
"The Department's planning should be based primarily on the likely behaviour of students, not the behaviour it desires from students."
Two key assumptions were made about expected work flows. Firstly, that students' past behaviour in applying late (on enrolment) could be modified. Secondly, that processing would be relatively straightforward with little "rework" required (from incomplete or incorrect applications or information). Both of these assumptions turned out to be incorrect.
The system design attempted to change the past patterns of behaviour by students by encouraging students to apply early. Those students that did apply early did not appear to have their applications processed any faster (and were more likely to have their documents go missing) than those that left it to the last minute. The pattern of "loan accounts established" for 2000 had a similar profile to past years (see Graph in Appendix 2), since enrolment patterns are essentially similar, which means a significant part of the work (ie VOS) cannot be done until enrolment. Although students should be encouraged to apply early to smooth work flows as much as possible, planning work flows and resource requirements needs to be realistically based on enrolment patterns.
Recommendation 11
DWI plan work flows and resource requirements based on actual enrolment patterns (while still encouraging students to apply early). |
Anomalies and Error Rates
The system was designed on the assumption of a relatively simple work flow of applications coming in which were automatically processed, resulting in a contract being signed and an account being established. The systems design did not allow for many anomalies, such as interpretations of full time study status, or frequent course changes (now estimated by DWI to be 80% of students) after initial enrolment (which could affect full time study status and/or fees payable). Nor was the high error rate in student applications and other information supplied anticipated.
These combined factors meant that the system could not cope as well with the actual processing once it went live. Many students still applied for loans close to or at enrolment. Errors or missing information on applications meant more correspondence with students than anticipated. Many anomalies required manual override which was difficult to do in a system that had been designed to be automated, resulting in delays. Anomalies were put in the "pending" pile (which one visitor to Palmerston North was told was "on top of that filing cabinet"), for which proper processes did not appear to be designed. Processing delays resulted in more student queries to the call centre, overloading the system. All these factors had a snowballing effect.
Recommendation 12
DWI plan resources for work flows based on actual numbers of anomalies and error rates, taking into account expected improvements. |
Staff and Skills
Staff Specialisation and Training
"Staff were trained to take loan applications over the phone. They were not initially expected to have to provide detailed information on the loan scheme. However, increased volumes meant these staff had to handle more complex enquiries without all of the training needed to effectively do so. This has impacted on the quality of some of the information provided on the 0800 service."
WINZ Student Allowances and Loans Newsletter, March 2000
The national standardised system meant that call centre staff were primarily trained to receive applications by phone or do the data entry for mailed applications. There were insufficiently trained staff to answer student queries regarding eligibility and entitlement, or the status of their loan applications. Nor could they easily correct or update anomalies or errors since the automated system was not designed to require significant human intervention. For various reasons, few staff were recruited from the institutions, leading to little transfer of knowledge of the loan scheme.
Most call centre staff did not understand how to access information on loan status, and this coupled with the pressure they were under, resulted in many students being told their application was awaiting the VOS response from their institution when in fact this was not always the case (discussed further below under Issues and Solutions - Interface with Institutions).
Under the original call centre model, most staff were expected to carry out all functions: taking applications as well as answering queries on eligibility, entitlement and status of loan application. Early in the process, the staff were divided into processing and call centre staff, but with limited specialist expertise for answering queries.
This also meant that a small core of permanent staff at Palmerston North were supplemented by a large number of temporary staff, which ranged from about 10-15 to 100 at peaks. Temporary staff were expected just to do data entry, but had to answer phone queries. They were insufficiently trained to deal with other than straight forward queries, a factor in their giving incorrect information to students. Some submissions have suggested that perhaps a fewer number of more highly trained permanent staff would be more effective. This needs to be considered alongside the other recommendations which would reduce the work flow through the call centre.
DWI has recognised the deficiencies in call centre staff knowledge of loans policy and process and is developing an extensive staff training exercise. However, the ratio of permanent to temporary staff (including local staff as well as call centre staff) and their functions need to be considered as well.
Recommendation 13
DWI review the ratio of permanent to temporary staff, their functions, numbers, location, and training requirements. |
Training material and manuals for staff training may need to be reviewed, in particular the Student Loan Scheme Manual for Tertiary Administrators manual would benefit from being revised and reformatted as an easy reference guide for DWI staff advising students and institutions.
Recommendations 14, 15 and 16
DWI review and revise, for completeness and ease of use, the manual for DWI staff advising students and institutions.
DWI ensure staff who answer queries on loan status are able to use and correctly interpret the relevant information on the computer system.
DWI ensure sufficient specialised staff, whether located centrally or locally, are adequately trained to deal efficiently with both the anomalies and with students with complex circumstances, for example, TIA or postgraduate students. |
