Lessons from HBL's blowout
What's the story behind the story? Click the NBR Radio box for on-demand special feature audio.
A software upgrade at Crown-owned Health Benefits Ltd (HBL) has joined the roll-call of government IT projects that have gone wrong, and way over budget.
A report just released by Auditor General Lyn Provost (and instigated after prodding from Labour’s Annette King) notes delays and cost blowouts with a major upgrade to an Oracle software platform.
Boosters of agile software development will likely have a poke at Oracle. Like most multinational software companies, it’s not the first time Larry Ellison’s giant has been involved in a project that’s blown deadlines and budgets.
And there might be some technology issues. But Ms Provost's report finds human folly is at the root of HBL’s problems. It's a familiar theme.
HBL was set up by former then-Health Minister Tony Ryall in 2010 with the aim of saving District Health Boards $750 million over six years by establishing shared contracts for key hospital services.
Initially it had $68.3 million in funding, drawn from DHBs.
Its core project was the development of a shared financial management and procurement system called the Finance, Procurement and Supply Chain (FPSC) programme. It was to be finished by November 2014.
Initially, HBL hoped to create the FPSC using its existing Oracle installation. That proved too optimistic. Upgrades and modifications ensued. By March 2015 it was still incomplete and $80 million of a revised budget of NZ$92.1 million had been spent.
"At that time, the programme was modified with a revised budget of NZ$120 million," Ms Provost finds.
She adds: "Because the programme is still developing, it is too early to assess the level of benefits it will deliver."
The forecast completion date is now some time in 2018, despite scaled-down features (notably the removal of centralising receivables and payables).
Ms Provost found: "Actual savings achieved during the first three years of the five-year period were NZ$247.7 million, compared with estimated savings for the first three years of $220 million."
However, only $71 million of those savings were directly attributable to HBL, the Auditor General says. That’s less than the $79 million cost of the FPSC upgrade to date.
"The savings came from several sources, including DHB's individual and collaborative procurement initiatives; MBIE's all-of-government initiatives; and healthAlliance procurement, as well as savings from HBL-led procurement initiatives," Ms Provost says.
Jonathan Coleman wound up HBL on June 30, just a few months after becoming Health Minister. It was replaced by a new vehicle, NZ Health Partnerships Limited from July 1, with a governance overhaul during the transition.
"New Zealand Health Partnerships has already begun to address many of the issues identified in the Auditor General's report," Mr Coleman says.
So what went wrong with HBL’s software upgrade?
All the usual suspects are present, including poor planning, a wildly optimistic timeline, poor communications, poor governance, poor change-management and staff training and a solution that couldn’t scale (that is, be used by different sizes of organisation).
“The programme’s goals were ambitious, requiring creating a single system that could replace 20 systems and different ways of operating. It appears that HBL underestimated the health sector’s fragmentation. This made achieving the programme’s objectives in the time allotted particularly challenging,” Ms Provost notes.
And having a second stab at things was costly. The Auditor-General says the cost of “re-planning” was $6 million.
Lessons identified by Ms Provost include a clear decision-making process and two-way decision making.
“Communication from HBL was one way. DHBs felt that HBL used the forums it attended to disseminate information about the FPSC programme and did not take the opportunity to hear DHBs’ views,” Ms Provost finds.
That meant HBL did have all the information it needed to navigate a complex upgrade, and no buy-in from DHBs who cold-shouldered it when things went wrong.
In short, poor communications and poor planning were the primary culprits – and both were symptoms of poor management. No amount of technology will help you with that.
There is a ray of light amid the various bungled projects. IRD's $1 billion-plus transformation project is running 15% under-budget so far. Either lessons have been been learnt – either about tighter management, or setting a sky-high initial budget.
Use MyNBR Tags to track people and companies - and receive key-word email alerts. Find out how here.