The Novopay Project: The Dilemmas in Global Software Outsourcing
One of the first of what will undoubtedly be many independent academic studies into aspects of Novopay was published recently at the International Conference on Global Software Engineering in Bari, Italy. In this week's Newsline, the paper's authors provide a brief summary of their findings.
The widely publicised 'Novopay' project, a New Zealand-wide transition from an onshore to a near-shore service provider (Datacom to Talent2), was intended to update the payroll processing system of the Ministry of Education and so implement a new nationwide payroll system responsible for the payment of about 110,000 teachers and education sector staff.
After various changes of direction and delays from its initiation in 2005, the project eventually went live in a 'big bang' cutover towards the end of 2012. The cutover occurred with known outstanding issues, and the result was that the bulk of the schools had to deal with underpayments, overpayments, or non-payments and a series of compounding errors.
Subsequently two public inquiries have been undertaken, and a Minister of the Crown has been allocated responsibility for injecting funding and resolving issues relating to the system. While some stabilisation has been achieved, issues have continued to the present, with One News reporting on October 3rd a system communication advising thousands of staff that they are on leave until next May. The project has been canvassed previously in Newsline, for instance John Rusk's piece of 20 February 2013 on the flaws inherent in price based tendering.
Here we briefly outline the results of a study by researchers from the Software Engineering Research Lab (SERL) at Auckland University of Technology, which was recently presented at the International Conference on Global Software Engineering in Bari Italy, and subsequently at last week's CITRENZ 2013 conference in Hamilton.
The study adopted a critical evaluative stance. Such critical studies, rather than simply reporting 'facts', aim to critique the status quo, through the exposure of what are believed to be deep-seated, structural contradictions within social systems. They critique taken-for-granted assumptions, and bring to light the historical and ideological forces which have led to contradictory social practices within the study context.
Given the mixed views of parties and interests in any large-scale project, questions arise of: who is a stakeholder, and what influence does each have on the outcome? Whose concerns are most likely to be taken into account in the implementation of a new system, and at what stage do they become salient?
The study analysed publicly available data sources, applying a technique known as 'dilemma analysis'. The corpus analysed comprised some 65 data sources, which included briefings to ministers; project group meeting records; an official project chronology; press and expert commentaries; and included the Official Information Act 'data dump' response on the Ministry site.
Through this technique, conducted from differing stakeholder perspectives, data was analysed in terms of: dilemmas, tensions or contradictions (categorized as ambiguities, judgements and problems). The outcome was a set of dilemma clusters depicting groups of dilemmas associated with different stakeholders or pairs (e.g. customer/vendor) which they had experienced during the project.
A sample of a customer/vendor dilemma cluster uncovered is depicted below, with a brief explanatory narrative. A fuller analysis with references can be found in the original paper:
Figure 1: Global Software Development - dilemma cluster
For the first dilemma How to better understand customer context, two perspectives can be contrasted. At one stage the Talent2 CEO John Rawlinson had cried foul about press reports of the $30 million paid to the company, by noting the role of New Zealand partners Fronde and Asparona in the infrastructure provision and software customizations. Contrasting this view was an anonymous blog entry, highlighting the cultural sensitivities between a large neighbour and a small country:
"NZ is littered with examples of an Aussie centric organisation of blow hards who relocate / centralise support services across the ditch and find that having to then contract local expertise is far more costly than the local staff they had laid off or other such shining examples of management talent".
How the other dilemmas become apparent can be seen in the paragraphs below:
Paul Matthews (IITP CEO) noted how the Government's open tendering procurement strategy can disadvantage local firms and hamper development of a local IT industry. The Education Ministry's briefing to the Minister noted that Talent2 had no senior management bandwidth in New Zealand… impeding their ability to make decisions about resourcing, priorities and future improvements to the service.
The business case for the BPO project noted that the Ministry would work closely with the existing service provider to support them in achieving 'a profitable end of contract'. But when the Ministry was evaluating contingency options, Leanne Gibson, Chair of the Payroll Reference Group, noted that both the incumbent New Zealand provider and the Australian firm had independently stated "that they would not work with each other as business partners" to provide a shared service.
Here we see, even in a near-shore setting, the classic Global Software Development challenges arising from a combination of cultural and geographic distance.
Given the topicality of the Novopay project, it provides a fascinating context for our research. Our main goal is to understand the challenges in software related transitions and benefit from the insights that these types of large multi-stakeholder public sector projects provide. It is hoped that techniques arising from the study may be able to be refined for use by practitioners and auditors to prevent such failures in future.
You can view the full paper here.
This research is being conducted within the SERL lab by PhD student Bilal Raza under the supervision of Associate Professor Tony Clear, from Auckland University of Technology and Professor Stephen MacDonell, University of Otago.
You must be logged in in order to post comments. Log In