Beyond the Pilot: Why Governments Must Bridge the Tech 'Valley of Death' to Scale Innovation

For years, the public sector has been great at starting things but notoriously bad at finishing them—at least when it comes to technology. We’ve all seen it: a brilliant pilot project launches with much fanfare, only to wither away once the initial funding dries up or the political winds shift. This phenomenon is known as the "valley of death," and according to global leaders at the Festival of Innovation (FOI) 2026, it’s time for governments to stop being passive observers and start acting as the primary engines of the tech ecosystem.

During a recent panel titled "Lessons from Global Public Sector Leaders," organized by GovInsider, experts from Germany, Malaysia, and Mongolia reached a striking consensus. They argued that for digital transformation to actually stick, government procurement needs a total overhaul. The days of being mere regulators or financiers are over. To truly succeed, governments must evolve into lead customers and active ecosystem builders.

The Government as an 'Anchor Client'

Manuel Kilian, the Founding Managing Director of the Global Government Technology Centre (GGTC) Berlin, brought a sharp perspective to the table. He argued that the government should be the "anchor client" for GovTech solutions. This means not just funding a startup to see what they can do, but actually being the first major customer that uses the product at scale.

Kilian also pointed out that governments need to be the stewards of Digital Public Infrastructure (DPI)—the foundational systems like digital identity, payment gateways, and data-sharing layers. He cited Singapore’s Singpass as a prime example. By using public procurement to ensure high penetration of the service, the Singaporean government created a platform that the private sector could then use to scale their own innovations. Procurement, in this sense, isn't just a back-office function; it’s a strategic tool to channel demand toward real-world solutions.

Breaking the Cycle of 'Pilot-itis'

One of the most frustrating hurdles in GovTech is what many call "pilot-itis"—the tendency to launch endless small-scale tests without ever moving to nationwide adoption. Kilian noted that while the cost of engineering has dropped, making pilots easier to start, the real bottleneck has shifted to adoption and institutionalization.

Scaling from one to many requires more than just good code. It requires rethinking organizational design, adapting legal and regulatory frameworks, and choosing the right governance models. Instead of just focusing on the next shiny pilot, Kilian argued that governments must build internal capacity to understand and absorb technology. We need to measure success not by the number of launches, but by metrics like monthly active users and overall user satisfaction.

The Malaysian Strategy: The Quadruple Helix

Sherry Sokmum, Lead of IP and Policy Management at the Malaysian Research Accelerator for Technology & Innovation (MRANTI), echoed these sentiments. She described MRANTI’s role as an ecosystem builder that connects the "quadruple helix": government, academia, industry, and society.

In Malaysia, the goal is to ensure that public money doesn't just fund R&D in a vacuum, but delivers tangible social benefits while de-risking technology for the private sector. Sokmum highlighted the use of regulatory sandboxes and problem-statement-driven procurement as essential tools. By acting as a "problem owner," the government signals to investors and buyers that a technology is credible. This signal is often what helps a startup finally cross that treacherous valley of death.

However, Sokmum didn't shy away from the challenges. She noted that institutional culture and a lack of funding continuity often kill momentum. When a pilot ends and the Treasury's rigid rules kick back in, the innovation often dies. To combat this, she recommends strengthening cross-agency coordination and proactively engaging private partners to move validated pilots into the mainstream.

Data as the Foundation: Lessons from Mongolia

Ariunbold Shagdar, representing Mongolia’s National Statistics Office, provided a masterclass in how data governance can drive innovation. In Mongolia, the focus has shifted away from technical "point solutions" and toward a comprehensive national data policy. Shagdar’s dream is an ecosystem where interoperable digital public infrastructure allows for secure data exchange across all institutions.

This isn't just theoretical. Mongolia’s Integrated Database and administrative data systems have led to a public services platform offering over 100 digital services. This has improved productivity and led to significant cost savings. Interestingly, Shagdar also emphasized the importance of receiving modifiable source code for local adaptation, ensuring that the government isn't locked into a single vendor.

Yet, even with these successes, barriers remain. Shagdar identified limited institutional knowledge, varying technical capacity across different agencies, and political instability as major risks. To ensure pilots are institutionalized nationwide, he stressed the need for coordinating laws and strategies specifically for Data and AI.

A New Era of Problem-Driven Procurement

The takeaway from the FOI 2026 panel was clear: if we want better public services, we have to change how we buy and build them. The panellists urged governments to stop treating innovation as a series of disconnected experiments. Instead, they need to reimagine traditional tendering processes, turning them into technology-agnostic, problem-driven calls to action.

When governments act as demanding lead customers, they don't just solve their own problems—they create a blueprint for the private sector to follow, ensuring that the next great innovation doesn't just survive the valley of death, but thrives on the other side.