The GovTech Procurement Playbook: How to Stop Buying Problems
Procurement was the quiet protagonist of G4I 2026. Here is the practical wisdom from across the Congress, distilled into a checklist for your next tender. Plus a free downloadable guide!
Ask anyone what derails digital transformation in government and they will probably reach for the obvious answers: legacy systems, skills shortages, risk-averse culture. Few will say “procurement.” Yet across three days at the GovTech 4 Impact World Congress in Madrid, the way governments buy technology surfaced again and again as one of the single most decisive factors in whether digital projects succeed or fail.
It came up in the sessions on AI, in the debates on legacy modernization, and in the discussion of digital sovereignty. And each time, the conclusion was the same. How a government procures determines what gets built, who is allowed to participate, how quickly solutions reach citizens, and whether the public retains control of its own systems. Procurement, the Congress made clear, is not an administrative procedure to be delegated and forgotten. It is a strategic policy instrument, arguably one of the most powerful a government has.
We have gathered the most practical procurement insights shared in Madrid into a single resource: The GovTech Procurement Playbook, a downloadable checklist for public servants writing their next tender. Here is a preview of the thinking behind it.
Procure outcomes, not problems.
The sharpest line of the Congress on this subject came from Idoia Ortiz de Artiñano, Chief Executive Officer, GOBE: “If you are not procuring outcomes, you are procuring problems.” Her argument was that governments routinely write tenders around static feature lists rather than desired results, locking themselves into rigid contracts that cannot adapt as technology and citizen needs evolve. The discipline she recommended is to specify the result you want and how you will measure it — and to leave suppliers room to propose the best way of getting there.
Run the risk logic before you go to market.
Marina Manzoni, EU Policy Officer at the European Commission, described a sequence that belongs at the start of any procurement, not the end: understand what you actually need, what risks you are taking on, whether the solution is genuinely feasible, and whether citizens will accept it, before a contract is signed. Institutions that skip this discipline, she warned, tend to buy problems they cannot later fix.
Beware the “freemium” trap.
Perhaps the most quietly important warning came from Tiago Carneiro, Senior Digital Development Specialist, World Bank (Serbia). A vendor may offer a chatbot or lightweight AI service at little or no initial cost. But once it gains users, absorbs your institutional context, and embeds itself in daily workflows, it becomes very hard to replace, and today’s low prices may be subsidized, meaning costs can climb sharply once you are entrenched. Unlike private firms, public administrations often lack the flexibility to renegotiate once locked in. Procurement discipline, he argued, has to begin at the very first conversation with a vendor.
Don’t outsource the thinking.
Ortiz de Artiñano returned to a second theme that resonated across the Congress: outsourcing delivery is reasonable, but outsourcing judgment is not. Without the internal capability to define, evaluate, and learn from what is being built, procurement becomes an abdication of strategic responsibility rather than an exercise of it.
Procure for sovereignty and scale.
Finally, the sovereignty discussions led by Kelly Ommundsen, Head of Digital Public Goods, World Economic Forum, and Leanne Cummings, Director, Government Digital Service (GDS), United Kingdom, reframed procurement as the place where sovereignty is won or lost. Open standards, interoperability, and modular design are not technical luxuries. They are the mechanisms that keep a government’s future options open. And as Jaume Miralles, Director General of Digital Innovation, Government of Catalonia, reminded the Congress, a procurement that funds a pilot with no path to production is not an experiment but a delay.
The full playbook brings these principles together into seven practical sections and a one-page pre-tender test – a set of questions worth answering before you publish your next tender.
