Prototyping can save government IT projects

When a government builds a bridge, structural flaws can be identified well before the bridge is inaugurated.  

Software, however, is unlike physical infrastructure. The digital landscape shifts constantly, dependencies lie hidden beneath layers of code, and problems often only emerge once citizens start clicking “Submit.” 

Large infrastructure projects typically follow the "waterfall" method, a sequential and linear process that was more suitable for clearly defined problems.

During the development, engineers can physically inspect the progress, bolt by bolt.

Digital systems, however, deal with abstractions. Defining the problems and solutions is not always straightforward due to invisible dependencies, uncertain data availability and quality, and unpredictable user behaviour.  

When governments treat digital projects like traditional infrastructure, the risks of failure can be compounded. 

Sandboxing procurement: A practical solution 

Agile procurement offers a smarter approach, beginning with a proof-of-concept (PoC) that evolves iteratively rather than relying on a risky, "big bang" approach.  

Whether a government builds, buys, or partners to develop digital systems, adopting an agile mindset is essential: 

• Prototype, iterate, and scale. Starting small with a minimum viable product enables governments to test, refine, and ensure the robustness of their solutions before wider deployment.  

• Quality over quantity. Simply adding more manpower (coders, data analysts) rarely solves underlying problems; building on top of shaky foundations only creates bigger risks and costlier fixes in the future, due to “tech debt”. 

One method to encourage a more agile and flexible procurement was sandboxing, which carves out legal and budgetary exceptions for government experimentation on digital systems.  

Key features that can be built into such a regime to make it effective include: 

• Small, time-boxed PoC (proof of concept) contracts with clear thresholds on scope and spending. 
• Outcome-based evaluations to ensure PoCs meet agency goals and deliver user value. 
• Continuation options being integrated into the procurement terms so successful vendors can scale without re-tendering, while unsuccessful ideas end early. 
• Public review to evaluate the effectiveness of the sandbox mechanism itself, whether by disclosing the number of successful PoC conversions or showcasing successful PoC case studies. 

This approach may help reassure stakeholders (clear gates, clear spend), let vendors build credibility through proven results, and provide policymakers with solid evidence before larger funding decisions. 

To subscribe to the GovInsider bulletin, click here.

An example of this is Singapore’s Tech Acceleration Lab (TAL) sandbox, which slashed prototype setup times from six months to two, with nearly half of its PoCs evolving into successful contracts.

Addressing the causes 

While agile procurement is familiar in digitally advanced nations like Singapore and Estonia, most governments in Asia still face challenges in embracing it. This is due to various factors: 

Rigid procurement rules: Many Asian jurisdictions retain inflexible rules emphasising least-cost, single-contract awards. Intended to curb corruption, these rules can inadvertently stifle innovation: 

• Where regulation prohibits agencies from implementing full quality-based criteria to appoint a vendor, the government may end up with a lower-quality, lowest-cost bidder as the winner.  

• Often, there are one-size-fits-all rules that can raise red flags for PoC works in the eyes of government auditors, as the agency breaks a project into several small, iterative works can be mistakenly interpreted as avoiding stricter governance. 

Rigid budget cycles and practice: Governments often lack flexible funding pools for agile sprints. On the other hand, fiscal constraint may make governments dependent on external financing, where its procedures can limit agility: 

• Many treasuries still insist on line item capital budgets that match a traditional “design–build–handover” mindset. 

• Sometimes, external funders such as multilateral international financial institutions are also tied with their processes, where loan-financed government IT projects are executed in hard-coded large tranches, tied to fixed deliverables based on waterfall specifications written years earlier. 

Outdated institutional structures: In many places, organisational design and ways of working can be stuck in the analogue era. The existing accountability frameworks may also reward procedural compliance more than innovation: 

• The majority of governments in Asia have no centralised coordinating body or sufficient institutional capacity or mechanism to manage effective components of a sandbox, such as a cloud environment and pre-approved legal templates. 

• Audit offices judge success by whether officials followed regulations closely, not by whether citizens receive better digital services. This perpetuates a culture of risk aversion within the government.  

Vendor landscape is opaque and incentivises the status quo: Discovering trustworthy IT vendors is challenging as a single logo could mask a daisy chain of subcontractors, making due diligence processes and accountability harder. Established IT vendors also tend to be system integrators that prefer larger multi-year deals over smaller, agile PoCs. 

Political clocks: Election cycles and political timelines often impose a fixed timeline constraint. While political milestones can sometimes become a stretched goal to rally everyone together around a mutual objective, it can be costly for critical system projects. The software codebase can be brute forced to comply, but the quality may not follow. 

Getting started today 

Although governments’ struggles have common themes, the answers on how to be successful in prototyping, iterating, and scaling may be different from one country to another.  

By learning from one another's successes and setbacks, governments across the globe can adapt best practices and lessons learned to their unique contexts.

Encouragingly, some Asian governments are beginning to reform their procurement approach for digital systems. The Philippines’ New Government Procurement Act of 2024 introduces “fit-for-purpose modalities” and electronic procurement provisions that could open the door to smaller, outcome-based contracts.  

The Indian government’s Government e-Marketplace (GeM) had also relaxed rules for startups to level the playing field in delivering services for government agencies. 

The complexity and rapid pace of digital innovation call for more flexible approaches.  

Policymakers should evaluate their current practices, selectively adopting iterative methods where they can most effectively reduce risk and improve outcomes.  

By learning from successful peers and thoughtfully integrating agile techniques alongside existing methods, governments can ensure digital systems better serve citizens from day one. 

__________

The author leads partnership initiatives in Asia Pacific at the Tony Blair Institute for Global Change, bringing over a decade and a half of public policy experience advising governments on digital transformation, budget, and infrastructure.