Drama surrounding the Australian Customs Service's Integrated Cargo System offers valuable lessons for those at the coalface of managing outsourcers.
When the Australian Customs Service decided to rebuild the technology of its core business process technology seven years ago under the name of Cargo Management Reengineering (CMR) nobody was under any illusions it would be a monumental job.
After the existing green-screen electronic data interchange system's more than 10 years of faithful service to industry, it was becoming clear the service that processes almost every import and export declaration in the country would eventually have to be replaced.
The department commissioned scoping studies, industry stakeholders were invited to express their fears and aspirations and Customs officials scoured the globe examining the experience of other countries. All the right intentions to build a robust, cost-effective, world-class system that would last for decades were there.
Most importantly, outsourcing was sold by those holding the reins of government as something that would deliver great efficiency, flexibility and renewal of what was perceived as a stale, slow, ineffective and stubborn public service.
In the case of Customs, this meant carving up an established code shop that held the critical knowledge of how the department's core business processes interoperated and the various technologies that made it happen and replacing it with outsourcer EDS. Heartache has followed ever since.
Lesson 1: If they say "yes", make sure they mean it.
When EDS became Customs' outsourcer in 1997, it inherited development responsibility for the CMR project. For five long years, CMR and its core transactional hub, the Integrated Cargo System (ICS), sat under the outsourcing deal with no real development. When the project failed to meet deadline in 2001, the cause was not that EDS had done a bad job in so far as it had not done anything at all. For nearly five years the project lived in a form of vendor limbo until EDS and Customs agreed that the project would be better served if it was re-tendered to the market. Eventually, a consortium led by Computer Associates and including IBM and Teradata won the tender in February 2002.
While EDS was prepared to provide all of Customs business systems needs, it was clearly not well suited to developing a highly complex, bespoke transactional system which involved considerably more resources and risk.
Nor did it want to, possibly a prudent move considering it was the arrival of EDS that saw so many specialised IT staff leave Customs' employ.
Lesson 2: Know what you are replacing before you replace it.
A flawed assumption was made that the business process knowledge at Customs would transfer from the department to the outsourcer. This did not happen. Decommissioned public servants went to work where they were valued the most: the clients of Customs to whom they had previously been providing a service.
Customs is now in the unenviable position of having some of its customers know more about how elements of its business process works than it or its outsourcer does. Because Customs legally compels clients to interface with it, the commercial element of choice is removed. Clients are not free to shop elsewhere and Customs cannot choose its clients, thus creating a frequently adversarial relationship.
The transference of knowledge occurred because Customs failed to effectively assess the real market value of the knowledge it was showing the door, or the impact it would have on both its organisation and customers. Many value judgements were made along ideological rather than economic grounds, thus miscalculating the impact to the business.
Lesson 3: Know the limits of your suppliers
Having missed numerous successive deadlines and blown successive budgets Customs is now under extraordinary political pressure to deliver ICS code to industry so that clients can interface with the system. Two successive deliveries of largely inoperable code have provoked a severe backlash from stakeholders out of pure fear that the system would be allowed to go-live in July next year.
Grave concerns arose from developers that because broken code was being released to industry to test as operational, a broken system would go live resulting in the Australian economy being crippled by virtue of the fact Customs would not be able to clear import or export cargo. In an attempt to speed up outcomes, senior Customs management sacrificed quality control on the project outputs because suppliers had most likely underestimated what they took on.
The result has been a grave erosion of confidence for both clients and suppliers. It appears likely that by attempting to threaten punitive sanctions against the vendors struggling to build the code, a bad situation has been made worse. While vendors can ultimately exit the project, albeit with penalties, clients of the Customs system cannot.
In a classic "David Murray" moment (where IT over-promises and under-delivers), Customs management greatly overestimated the capacity of their suppliers to deliver and underestimated the negative impact that delivery failures would have on their customers. Yet, despite the clear failure of vendors to deliver on time, Customs insiders say that the department only heard what it wanted to, regardless of the reality.
Lesson 4: Know your exit strategy
A key driver of many of the problems the CMR project has faced originated because of a legislated deadline by which the ICS system must go live. This has been moved twice, a year at a time. To add insult to injury, ICS users now face a user-pays levy to cover the cost of budget blow outs - which have escalated from early estimates of $20 million to the $145 million revealed before then Senate Estimates this month. The introduction of a new tax to bail out what is seen as government incompetence will be profoundly unpopular and has the potential to cause splits in the government.
The legislative deadline effectively rules out any ability to compromise and has forced clients to adhere to unrealistic deployment timeframes. By committing early to 'cast iron' deadlines and forcing stakeholders to do the same, Customs paints itself into a very messy corner every time the deadline is missed. Realistic rescheduling or options for flexibility currently exist and need to be introduced.
If they are not introduced by those in charge of the project relatively soon, the stakeholders (which include supermarkets and large retailers, manufacturers and the transport sector) will almost certainly remove those in current management the only way left open to them - brute political force.
The alternative is the Australian economy grinding to a halt... an unlikely choice for any government.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.