It’s clear that 2015 wasn’t the year that bucked decades of reports and lawsuits over California IT projects that have struggled or failed. It certainly wasn’t the long-sought year when things went smoothly, where there were no bad headlines and everything was positive.
There were legislative hearings and a February audit slamming the BreEZe project, an effort to streamline business for the Department of Consumer Affairs that’s ballooned from $29 million to $98 million even while shrinking in scope. And bidding was recently further delayed for the Board of Equalization’s Centralized Revenue Opportunity System (CROS), which has been in procurement for approaching four years.
“Unfortunately it’s a constant drip, drip, drip of bad news about these big projects that have continued to go awry,” said John Thomas Flynn, the state’s first CIO and now host of TechLeader.TV. “In essence they’ve become political battles, allowed to go on and on because nobody wants to pull the plug.”
It’s not likely 2016 will be a banner year either, state officials acknowledge, with a new procurement process aimed at righting some of those persistent wrongs a couple years from full implementation. But once that Project Approval Lifecycle (PAL) process is in place, leaders have high hopes for some positive change.
“I think what you’re going to see this year and really into next year is transition years,” said Andrea Wallin-Rohmann, chief deputy director of policy for the Department of Technology. “I really think in 2017, publicly you’ll start to see things differently than what you’ve historically seen.”
That’s good news, since what we’ve historically seen are big projects with staggering losses.
In a March report, State Auditor Elaine Howle discussed California’s history of failed IT projects, citing nearly $1 billion spent on seven projects that were terminated or suspended from 1994 to 2013. That includes a $50 million DMV upgrade canceled in 1994, a $500 million Court Case Management System dumped in 2012 and the $373 million MyCalPAYS payroll management system canceled in 2013.
The State Controller’s Office sued the MyCalPAYS vendor, SAP, in November 2013, hoping to recover some portion of the $50 million it paid plus damages. SAP is countersuing. Court records show the case is scheduled for a jury trial starting May 23, 2016.
Still, there are projects underway today that are raising some of the same red flags, such as BreEZe and FI$Cal, the Financial Information System for California. That raises concerns over why California can’t seem to learn from its mistakes.
“Who ever got fired because of a failed IT project? You just don’t see that,” Flynn said. “Until that happens, I think you’re just going to see this continue and continue.”
Eric Steen was well aware of those high-profile failures when he left the private sector to become project director for CROS in 2011. “I took the job because that cloud existed. I was frustrated with projects that went sideways,” he said. “I really wanted to see the state become successful.”
Steen, who resigned Sept. 23, knew the process of finding a vendor for the $309 million CROS project, which will replace the Board of Equalization’s legacy systems, would not be speedy. So he decided to use that time to try to mitigate risks that have triggered problems before.
State agencies typically make assumptions about their technology systems and what it will take to upgrade in terms of interface, business rules and functional requirements, Steen said. Then, when the vendor starts implementation, everyone is surprised when things like data conversion trigger problems for projects like MyCalPAYS.
“Those landmines that you’re unaware of can cause massive overruns and massive delays,” Steen said. “Our approach actually managed risks by addressing those issues up front.”
Steen’s process was like what the state is attempting to implement for all IT projects with its PAL system.
Previous projects and ones that are well underway were launched using the cumbersome Feasibility Study Report method the state has used for years, Wallin-Rohmann explained. There were inherent challenges in that process that triggered past problems, she said, including how funding approvals were done, lack of a detailed master schedule and no honest assessment of the departments’ needs and resources.
In response, the state created PAL by adapting the federal government’s “stage gate” model. PAL breaks up the process into phases, requiring much more detail along the way. And before a project can move to the next phase, the Department of Technology will collect data and sign off that it’s on track to be successful.
“It’s going to force us to have the right level of discussions at the right time to get to the right information so that we’re not doing that in implementation,” Wallin-Rohmann said. “We shouldn’t be seeing the cost overruns [and] the excessive change orders that we’ve seen in the past. We shouldn’t see schedule slippages because we’re having more intentional discussions up front.”
The Department of Technology implemented the first two phases of PAL on July 1. It expects to go live with the third stage, which is pre-procurement, in early 2016. And the final procurement stage should launch in June.
With similar steps taken for CROS, Steen remains optimistic the project will ultimately be successful. Some worry, though, that CROS also might be an example of how PAL could further delay already lengthy procurements.
After working on CROS for four years, Steen resigned in September, citing his frustration with how long the process was taking. He now works for the consulting firm QualApps.
Wallin-Rohmann said they don’t know if PAL will extend procurement timelines, since it’s still being implemented. But she said, “I would rather see those, if you want to call them delays, upfront versus after I’ve signed contracts.”
Though she didn’t offer specific details, Wallin-Rohmann said the state will likely look to reinitiate projects that previously stalled in hopes of giving them new life through PAL. The PAL process now has more built-in accountability too, she said, allowing the state to better monitor projects and extract data for analyses down the road.
Last, the Department of Technology is looking at moving away from the massive IT projects that have caused so many problems in the past, Wallin-Rohmann said. Instead of bringing in large solutions aimed at “fixing” the whole organization, she said, they’ll look at platforms departments can build upon.
It’s a big shift, Wallin-Rohmann acknowledged, with benefits not easy to pinpoint when looking back at 2015.
“It’s slow and steady, but we’re getting there,” Wallin-Rohmann said. “And with support and commitment from everybody else around us, we’ll get to a place where we do things better in government.”
This story appears in the Winter 2015 issue of Techwire magazine.
