It’s time to go beyond experimentation – Demos Helsinki in the OECD

Over the last years, it has become obvious that governments could benefit a lot more from experimentation. Experimentation helps co-designing solutions, policies, and services together with potential end users. It mitigates risks because it enables the fast process of turning ideas into practice. Without immediate ‘fear of implementation’, governments are able to think more creatively and explore the sphere of unknown. Experimentation also helps in evaluating reliably what works and what does not.

Experimentation does not belong to public sector alone. There are some great examples on how experimentation has been used in the private sector as a tool to accelerate innovation. Mark Zuckerberg has said that one of the things he is the most proud of at Facebook is their testing framework. According to him, there’s never only one version of Facebook, but rather 10 000 versions that are being tested and changed constantly. Jeff Bezos, the founder and CEO of Amazon, has said: ‘Our success at Amazon is a function of how many experiments we do per year, per month, per week, per day…’

Movement towards systematizing the use of experimentation is now taking place in public administrations around the world. Finland is the first country in the world that elevated experimentation to the highest political agenda. The government selected a set of normal policy objectives [ranging from social security to employment, education, and digital services] and now utilizes experimental methods in these policy processes. Canadian government has set a mandate for Deputy Heads to allocate funds for experimentation and report results of experiments transparently. The Prime Minister’s Office of the United Arab Emirates has recently helped horizontal teams to formulate societal missions, and it will be applying experimentation to push towards realisation of these missions.

Experimentation is not about experiments

Improving government’s capability for experimentation is not as much about improving government’s methodological capabilities as we have thought. If we really want to introduce experimentation in government and mainstream it, public governance must not be only improved – it must be transformed.

At the moment, many great experiments are blocked by legal, administrational or structural issues. For example, recently in Finland, a relatively simple detail in data management systems of the employment office led into delayment of several employment related experiments. [Issue was not the fault of the office, but a problem that arose when responsibilities shifted from national level to municipal level, for the sake of the experiment].

If we want to help governments to start tapping into the great potential of experimentation, we need to stop concentrating on experimentation methodology. Instead, we have to start taking a systemic approach to experimentation. This means, for example, that we can’t only concentrate on training new skills to government employees – if there’s no functional financing model for experimentation. We can’t only concentrate on making cultural change happen – if government is unable to collaborate smoothly with relevant stakeholders.

Experimental State – how to get there? Important learnings from the Finnish context

1. Mainstreaming experimentation in government requires mainstreaming an experimental mindset. The simple graph above illustrates some of the crucial changes in thinking that needs to change. Such a cultural change is very challenging to lead. While as government is running first experiments, it must also put emphasis on leading the culture change.

2. Solve what can help government to better interact with the rest of the society. Experimentation is often mostly about stakeholder work: small part of the experiments take place only on federal level, even though they are often steered from ministries or their departments. This creates an imperative for inventing new collaboration models.

3. Innovation requires ‘boring things’. Experiments are often displayed as the interesting, innovative new initiatives. Nevertheless, very few radically transformative experiments will ever take place, if we don’t get the legal frameworks and regulation to support the experimental approach.


These are the topics that we’ll be working on in the OECD’s OPSI Innovation Conference, in a workshop called ‘Systems Approach to Experimentation’ [Tue 20th of Nov 2018, starting from 10:15].

If you failed to enroll in time and are interested of the topic, feel most free to contact: , +358 40 778 6062.

Mikko Annala

Header Picture: Jonas Weckschmied