Belgium's Gov Buys Innovation programme begins with a problem that public services cannot solve well with products already on the market. Officials identify the need before describing a specific challenge. The challenge can then be presented without prescribing a finished product.
Defining that need calls for a capable procurement workforce. Service staff understand daily operations, engineers test technical limits, lawyers protect the purchasing rules and finance staff examine cost. Together they turn a broad frustration into requirements that can be discussed and tested.
Before a tender is fixed, officials can confer with possible solution providers. The OECD says this early contact helps them identify the problem and explore possible answers. Access, questions and feedback must remain transparent, so that one firm does not gain a hidden advantage.
The Belgian procedure may then buy an experiment that tests critical building blocks of a solution. Its stages limit paperwork for firms that do not advance, while feedback and transparency remain central. Within those published stages, procurement specialists can negotiate the experiment contract. The contract covers the experiment itself; it does not promise a finished service.
Buying an experiment gives officials working evidence before they invest further. The disadvantage is that a successful component test can still leave questions about scale, maintenance and performance in everyday use. The experiment narrows uncertainty without removing it.
The same OECD report presents Canada's Build in Canada Innovation Program as a separate example. Under that programme, innovators can test products in the federal government's operational environment. Departments undertake the tests, assess performance and give contractors feedback. This Canadian case describes operational testing, not a later stage of the Belgian procedure.
Ultimately, Belgium's next decision returns to the original need: did the experiment justify investing further in that direction? The Canadian example answers a different question by showing how a government department can assess an innovation in use. Keeping the programmes separate makes the lesson clearer—public procurement can buy evidence in more than one way.