Information technology was crucial to Centrelink from the start. DSS computers and IT staff were transferred to Centrelink to enable the payment of income support to continue. In contrast, the department responsible for employment chose to provide access to Centrelink to its computer network but to retain ownership of it within the department.
The size and scope of Centrelink’s IT were significant. By 2004, it was considered to be the fourth-largest IT operation in Australia, with more than 35 000 personal computers in its network processing 14 million transactions a day. Centrelink’s provision of income support was regulated by more than 80 000 rules. Every small policy change involved rule changes and associated coding changes. Each major new policy created and changed rules about eligibility and payment and generated new and revised coding. Each change had to be made, tested and documented, with new or revised forms and letters generated, and the changes explained to the staff on the front desk responsible for implementing the policy with customers. Bringing together six programs into one Youth Allowance program created delivery problems for Centrelink in meeting the government’s policy start-up deadlines.
Centrelink extended the call centres inherited from the DSS to become an increasingly important part of its business with customers. Each year, the demand for services grew and, by the end of 2002, there were 4500 people working in Centrelink’s 27 call centres. By integrating all its electronic and telephonic systems, Centrelink was able to offer a seamless service to its staff and customers throughout Australia. It had connected its mainframe systems to software able to receive and interpret incoming inquiries from any device, such as a personal computer, kiosk, telephone or palm pilot, as well as remote-access laptops around Australia, selecting the appropriate database to conduct the transaction and respond to the inputting device (Vardon 2002c:14). By using this switching capability through its call centres from Queensland to Western Australia, it was able to use Australia’s time zones to extend its hours of operation to eastern state callers up to 10pm on weekdays without requiring extensive shift work.
This was a significant improvement on what Centrelink’s customers rated poor access in its early years. The ease of accessing Centrelink’s services was rated ‘good to very good’ by only 40 per cent of respondents in 1998 (CSS November 1998). A year later, this rating had risen to 58 per cent, and stabilised at 60 per cent in 2003.
Views about access in 2001 were influenced primarily by perceptions of telephone accessibility (35 per cent), limited ways of dealing with Centrelink (13 per cent), lack of staff at Centrelink (12 per cent), location of offices (10 per cent) and other issues, such as transport and phone-line operating hours. While the morning was the time preferred by most customers to contact Centrelink, the 2000 survey report noted the ‘growing demand…to transact business after hours using non-office channels (such as the internet and telephone)’ (Millward Brown 2000:20). One in five customers said they would prefer to contact Centrelink after hours (mainly Austudy, families and Youth Allowance customers, and those who were studying, employed full-time or living in rural areas).
Nevertheless, most people continued to prefer to visit a Centrelink office (a reversal from the apparent trend towards preferring the telephone before the call centre difficulties of mid-1998). The reasons given for going to a Centrelink office in 2001 were, first, the need to talk to a person face-to-face (87 per cent) and second, the ability of office staff to deal with the problem straight away (35 per cent). Those who preferred the telephone did so because it was easier (54 per cent), quicker (38 per cent) or more convenient—a particular preference of rural customers. The few who preferred the Internet gave similar responses.[5]
After the delays of 1998, a substantial expansion of call centres had significantly reduced the call waiting times of call centre customers. A review of Centrelink cost efficiency showed that call centre workload almost doubled from 1997–98 to 2001–02. The number of calls had risen from 19 million to more than 25 million in 2002–03, with the duration of calls about 8.8 minutes, a continuing trend upwards from the 5.4 minutes of 1997–98 (BCG 2002:21, 23; Centrelink 2003a:115). Call waiting time, a primary driver of customer satisfaction, dropped from a high of 325 seconds in July 1998 to 144 seconds in 2002–03, well within Centrelink’s target of 210 seconds and identified customer expectations. As already noted, customer satisfaction with these services rated consistently between 85 and 90 per cent. The call centres had become the largest single-purpose call centre operation in Australia (Centrelink 2003a:114–15).
As Centrelink moved its service model away from the initial separation of programs, it was necessary to integrate all program information and customer records into one record. The one-to-one officers, assisting their customers with problems that went across several programs, needed one complete, coherent customer record. Centrelink aimed to establish a ‘customer account’ for each of its customers, providing a clearer display of customer data. This reduced training for the staff required to update the information and allowed customers to see their own information and, eventually, to be able to update their own records.
The growth of the Internet required a further sustained response. Like other federal government agencies, Centrelink was required to respond to the government’s whole-of-government initiative to have all appropriate services available through the Internet by 2001 (Commonwealth of Australia 1997). Centrelink had already developed a web site, signalling an extension of its operational boundaries, and now undertook a new range of initiatives to guarantee service delivery online to designated special groups, and, by 2001, to be accessible 24 hours a day, seven days a week. Rural and regional communities were being targeted as well as special groups.
After this, and particularly after the life-events service model was initiated, the web site was extensively redeveloped and enhanced. The passing of the Electronic Transactions Act and the Social Security Administration Act removed legislative barriers to Centrelink conducting its business with its customers electronically (Bashford 2000). Nevertheless, while customer access to the Internet rose, largely in line with that of the general community, this did not translate into greatly increased use of the Internet for Centrelink business. In 2001, while 44 per cent of customers had access to the Internet, only 5 per cent had visited the Centrelink web site in the previous three months—most (76 per cent) to get information or access job search facilities (24 per cent). Part of the problem was that many of the transactions customers wished to do on the web site, such as checking eligibility or lodging a form, required a signature or other identification that was not yet achievable through the web site. Nevertheless, the proportion of customers who claimed they would be interested in dealing with Centrelink via the Internet in the future rose from 1997 to 2003. Acceptance of voice-recognition and keypad technologies also rose (CCS November 2003:80).
By early 2004, 80 000 customers a week were using Centrelink’s self-help services, although this represented only 2 per cent of all Centrelink transactions—a figure that Centrelink wanted to see grow quickly (Senate CALC 2004:52). There was potential for Centrelink to expand communications and provide customers with convenient services while lightening the load of call centres and CSCs.
Despite this limited acceptance and use of technology, Centrelink customers were experiencing improved convenience. In the November 2003 survey, 86 per cent agreed that their Centrelink office was conveniently located and 83 per cent agreed that they could usually do business with Centrelink at a time that suited them (CCS November 2003).
As Centrelink broadened its boundaries beyond social income payments to new business and greater connectivity with the community, it developed new options for its customers. CentrePay, for example, allowed social welfare recipients to have deducted directly from their welfare payments essential bills such as rent and electricity to state or local government agencies.
Beyond these approaches, Centrelink was trying to harness leading-edge technologies to enhance the quality, improve the accuracy and/or reduce the cost of their service delivery. Not all were successful. Perhaps most innovatively, but least successful, were attempts to develop a decision support, or ‘expert’, system, to help customer service staff by guiding them through the technical rules and assisting them to determine questions of eligibility (for example, an expert system supporting the FAO payments was installed at several sites but by 2004 Centrelink had withdrawn from further development). Another initiative that failed in the short term was the web post office for Youth Allowance customers. The idea was to deliver email letters through the Internet. It failed because recipients wanted to reply through the same means, but problems of authentication made web replies unacceptable to Centrelink.
Other initiatives were more successful. In 2003, a multimillion-dollar speech-recognition project was trialled in which students could report key information on lifestyle changes relevant to their benefits; a 30 per cent take-up of this option led to the extension of this service to other welfare recipients. A trial in 2002 of SMS text messaging to students to convey information reflected the fact that 78 per cent of Youth Allowance customers had a mobile phone and they were the group of Centrelink customers who most used SMS in their daily lives. Using SMS messages was a more cost-effective method than the use of either paper or other phone messages (Senate CALC 2002:16).
In spite of these developments, the BCG report noted that Centrelink’s spending on IT had been declining at a time when that of ‘comparable’ organisations (that is, banks as large networked organisations) was increasing, and that a large proportion of the spending was ‘geared to keeping the lights on’, by which it meant the operation of the IT system. As Hickey noted at a Senate Estimates hearing, a large proportion of total IT spending was ‘targeted at maintaining…existing service delivery networks…the ongoing effective operation of the IT systems is absolutely critical to making the 6½ million payments each fortnight’ (Senate CALC 2003:28).
The allocation in the 2003/04 budget of $312 million over five years for Centrelink to enhance its IT capability would enable Centrelink’s IT to be upgraded. Called ‘IT Refresh’, it was to be used to ensure continuing IT reliability, allow Centrelink to deliver online self-services to customers, enter partnerships with community organisations, increase payment accuracy through automated data matching and information exchange with business, government agencies, banks and others (DFaCS 2003:255).
For some time, Centrelink’s customers had been using each of Centrelink’s four service channels: visiting a Centrelink office or agent or using a kiosk, phoning a call centre, using the Internet site, or writing to Centrelink (colloquially called on-site, on-call, online and on-paper). By 2004, 72 per cent were using the telephone, 76 per cent using face-to-face contact, with 20 per cent using other means (and many using more than one channel) (Centrelink 2004c:23).
There were significant variations in the cost to Centrelink of the use of each channel. As Vardon (2002b:59) noted, ‘[O]ne-to-one service is at the most expensive end of the range and we want to develop it for people who need intensive help.’ Others, such as students, ‘would prefer to do business electronically and for the government, cheaply’. Vardon added, however, that ‘we have learnt that as tempting as it is to open an interactive internet channel, the resources required to do so are prohibitive. We have to go self-service’ (Vardon 2002c:13). There were also challenges for many Centrelink customers in gaining computer access and developing confident usage.
The cost of different channels and different activities was reviewed by BCG, which argued that Centrelink should seek to reduce the frequency with which it contacted customers and customers had to contact Centrelink. It estimated that ‘notifiable events’ (the requirement by the government that customers who were seeking work must report to a Centrelink office every two weeks) accounted for 40 per cent of Centrelink’s costs, and recommended that Centrelink campaign to reduce the frequency and cost of these ‘events’ by using risk profiling, shifting the processing to lower-cost channels, such as call centres or self-service channels, and changing its channel economics to drive down the cost of processing notifiable events (BCG 2002:34–7).
Some of this activity took Centrelink into the realm of policy change and negotiation with its client departments, illustrating how policy decisions influence costs and service delivery. Customer satisfaction surveys showed that Centrelink customers were using more resource-intensive channels, such as face-to-face, for simple transactions such as receiving information and checking payment details. The 2001 customer satisfaction survey found that 53 per cent of respondents had visited a CSC in the previous three months. In terms of Centrelink costs, it was preferable to shift these service transactions to the Internet channel. After these surveys and the BCG report, Centrelink began to manage its multiple channels more actively—for example, it placed a new emphasis on gaining as much information as possible from customers during their initial interview to reduce repeat visits or follow-up calls.
In examining the different costs of each channel and working out which channels were best suited to respond to the range of tasks, Centrelink used a ‘service response framework’ that identified three main approaches from its customers, each associated with a channel.[6]
In balancing its customer needs and its own costs, Centrelink had to integrate its multiple channels to ensure that customers could use the channels most convenient for them, move smoothly from one channel to another through a common interface, experience reliable data recording from one channel to another and be confident in terms of security and privacy. Centrelink underwent a significant program of business process re-engineering. In association with this channel-management project was the project looking at where work was done in Centrelink and arguing for greater consolidation and specialisation—the creation of the CSSCs. The aim was to drive down costs and lead to ‘better quality decision making and more timely and more efficient service delivery’ (Hickey 2004).
The management of channels required greater flexibility within the organisation in the use of organisational resources, development of broader staff skills and understanding how customers viewed particular channels, in terms of trust and privacy, and whether these views could be changed (Moore and Flynn 2004:5). It also involved Centrelink in seeking customer advice and feedback on the design of the services they used.
The issue of channel management was closely related to customer satisfaction levels. A Service Integration Survey in 2003 on call centre and CSC service concluded that customer satisfaction was lowest when customers had to deal with multiple staff and/or channels (DBM Consultants Pty Ltd 2003c). It was also shown that those who had repeatedly used call centres for one piece of business had a more negative view of Centrelink, while those who made repeated visits to CSCs were among the most positive of those surveyed. For more complex matters, all categories of respondents (except those who had used only call centres) expressed a preference for dealing with Centrelink in person at a Centrelink office (DBM Consultants Pty Ltd 2003c:41, 44).
The research also showed that customers had more confidence in the CSCs than they did in the call centres. At the same time, another survey showed Centrelink staff believed that most payment problems were the result of office staff, not call centre staff, who were considered to be better trained and more up-to-date with legislative changes (DBM Consultants Pty Ltd 2003c:43, 48).[7] This was not good news for Centrelink in terms of reducing costs, as visits to CSCs remained the most costly of all service channels.
The IT Refresh initiative to increase the stability and capacity of Centrelink’s IT systems and to make self-service more easy and interactive was regarded as crucial to Centrelink shifting from high to lower-cost services. The $312 million program over five years was designed to improve service delivery and to provide customers with greater choice, particularly through technology that supported self-service options (Centrelink 2004c).
As Centrelink increasingly linked its services with those of other organisations—business, educational and community—its customers often received services from a group of organisations working through the Centrelink gateway. The capacity of the IT systems to capture data and make them available to other channels in an accurate and consistent way was regarded as critical to customer satisfaction and seamless service.
[5] The details were: easier, 52 per cent; quicker, 38 per cent; convenient (that is, don’t have to leave home or work), 27 per cent; and can be used at any time, 21 per cent).
[6] The three key tasks customers sought were, first, those requiring a transactional response—‘just let me do it’; second, those requiring a problem be fixed—‘help me’; and third, those requiring personal assistance from Centrelink staff who understood the context of the problem—‘relate to me’. The first approach could usually be handled through technology—online or on-call; the second might need a person but not necessarily one face-to face, and could be handled most effectively by the call centres; the third was more likely to require personal, face-to-face interaction. Paper information supported all three channels.
[7] Respondents appreciated the greater convenience of the call centres, but only 29 per cent agreed that ‘I have more trust that things will be done right if I ring up the call centre rather than go into the office’ (DBM Consultants Pty Ltd 2003c:43).