PAL, launched in 2016, brought needed structure to state IT project planning. But its stage-by-stage approval process often delayed innovation. Departments sometimes spent years on paperwork, only to find that by the time a vendor was selected, user needs or technologies had shifted.
The PDL model is designed to break this cycle. It encourages agencies to partner with vendors earlier, build proofs of concept (PoCs), and develop minimum viable products (MVPs) iteratively — getting feedback faster and adjusting course before committing major funds.
It’s a smart move — and CDT’s leadership should be applauded. However, as California embraces this promising new path, it’s worth thinking carefully about how to maximize PDL’s success and avoid new risks that could inadvertently arise.
WHAT THE PDL GETS RIGHT
PDL emphasizes prototyping and iterative learning rather than exhaustive planning before any work begins. In a fast-moving world where technology evolves quickly, this flexibility is essential.
Perhaps PDL’s greatest strength is its early focus on defining user needs. The Office of Data and Innovation (ODI) will help departments better frame their problems and build clearer requirements from the start.
This approach mirrors how large infrastructure projects often hire an “owner’s representative” to help define project needs, shape RFPs, advise during procurement, and stay actively engaged through delivery. In California, the Department of General Services (DGS) has long used outside architecture and engineering firms on major construction projects to play this role on complex building projects, supplementing DGS’ internal construction management staff.
By drawing ODI in earlier, PDL takes a major step toward ensuring that IT solutions are built around the true needs of departments and their customers.
RISKS WORTH MANAGING CAREFULLY
While PDL offers promising solutions, some risks deserve close attention. These include:
Vendor Lock-In — Once a department accepts an MVP, it becomes dependent on the vendor who built it for any system improvements. This can reduce the state’s negotiating leverage as it’s difficult to switch vendors without significant disruption or cost.
Incomplete Requirements — If a department moves quickly to MVP without full exploration of user needs, it can risk building a foundation that can’t support future expansion. California’s MyCalPays project offers a cautionary tale. The payroll project failed in part because the underlying complexity of hundreds of union agreements and pay differentials wasn’t addressed up front. In hindsight, the state would have benefited from first examining how to simplify and standardize pay rules to significantly reduce payroll complexity.
Uncertainty for Departments and Vendors — Without full project funding committed up front, departments under PDL may struggle to plan staffing, and vendors may hesitate to invest deeply — fearing they may not win future stages. Such uncertainty comes at high cost.
Staffing and Continuity Risks — Iterative models like PDL demand experienced project teams who stay engaged through changing conditions. California’s history of high turnover on long IT projects could undermine PDL if continuity incentives aren’t built in.
STRENGTHENING THE PDL FOR THE LONG HAUL
California could further strengthen PDL’s success by considering a few additional strategies:
Simplify Before You Digitize — Technology can automate processes — but it can’t fix broken ones. Before building, departments should streamline business processes wherever possible to minimize complexity, mitigate IT project risk and lower costs. However, this largely requires programmatic or even policy changes to happen in advance of IT design and development.
Segment IT Projects by Risk — One size doesn’t fit all. CDT could adopt a more flexible oversight model based on the complexity, cost and risk profile of each project. For example:
- Low-risk projects (i.e., simple commercial off-the-shelf [COTS] software purchases) should proceed independently with quarterly reporting to CDT.
- Moderate-risk projects (i.e., COTS with some customization) could proceed with departments freely combining or compressing PAL Stages 1 and 2, followed by RFP development (Stage 3) requiring final, expedited CDT approval.
- Higher-risk projects (i.e., highly customized systems costing $10 million+) would require full PDL engagement, including PoCs, MVPs, iterative oversight and deeper governance.
- Highest-risk projects (i.e., unique, customer-facing IT projects costing $100 million+) may require a hybrid model that combines PAL’s thorough planning rigor (albeit expedited) with PDL’s iterative flexibility.
Use Owner’s Representatives — Just like DGS supplements construction management with external experts, California IT projects could benefit from IT advisory firms that stay engaged end-to-end: from RFP drafting through project closeout. Given ODI’s limited staffing, supplementing with trusted external partners would ensure projects get the attention they need.
Strengthen Governance and Executive Sponsorship — These aren’t just IT projects — they are program transformations with technology components. Strong executive sponsors must come from the program or business side, not just IT. IT should advise; program leaders must drive business decisions, set priorities and own results.
Incentivize Continuity — Offering bonuses to staff who stick with a project through completion — and holding vendors accountable to long-term delivery — could dramatically improve project stability.
BUILDING FOR TOMORROW
Building a modern IT system is like building a house. You can draft beautiful blueprints, but you also need to build the right foundation to support future additions. Likewise, while planning matters, success often comes down to construction management — when conditions change, problems arise or new needs emerge. That’s why California must strike a balance — faster approvals and smarter designs, but continuous engagement and strong governance during the build.
The PDL offers a promising framework. If we strengthen the foundation now — with the right risk segmentation, process simplification, strong governance and proper incentives — we will build systems that not only meet today’s needs but can flex, adapt and grow for decades to come.
