How to become a resilient government? Introduce the 'Minimum Viable Policy'
Exploring an agile way of working within the public sector
Let's explore the concept of the Minimum Viable Policy. It's not new that the government needs to become more resilient and adaptive due to the rapid pace of technology development and the ever-changing needs of society. In particular, the rise of algorithms speed up the process even more and some of them are far more unpredictable than we as humans can comprehend. At some point we need to accept and embrace that there will always be a gap of (un)known unknowns which we can't foresee. This is why we should approach the future iteratively, to reevaluate our position in cycles and update our regulations where necessary in order to adapt to changes.
We can distract some great examples on the use of the Minimum Viable Product (MVP) from the startup scene. Where most startup’s create MVP’s to test product features, this could also be a great place to start from when creating laws and regulations in an agile way.
About the Minimum Viable Product
Eric Ries (from the book Lean Startup) defines the MVP as that version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort. It’s a mechanism for validated learning, used to test hypotheses and discover what will meet customers’ needs.
Let's give an example. Let's imagine we would have reinvented the car from scratch. If the goal is to create a vehicle that could transport people faster than if they'd be walking. Instead of developing each part of the car with precision, you could test how a person would experience a platform with four wheels (skateboard). Congrats, you made your first MVP. Ah, they need support to stay in balance? How about creating a grip (step)? Nice! Hmm.. Did they say they want to go faster with less effort? How about making bigger wheels? We could try pedals. Boom! We found a great market. Why are people cycling so fast nowadays, do they want a vehicle that requires less effort and takes less transportation time? We could motorize it! See how many people want to buy a motorcycle now? Great! Do you hear them say they want to transport their whole family and want more comfort and safety measures than the motorcycle could provide? Now we have validated proof: Let's build some cars!
Why is this so valuable? Gathering insights from an MVP is often less expensive than developing a product with more features, which increases costs and risk if the product fails, for example, due to incorrect assumptions. Just like the Google Glass was introduced too early, it failed to enter the market as people weren't ready for it yet. Maybe if they'd created a few steps in between as MVP's, to let people experience what it could be like and provide feedback, they might have learned some incorrect assumptions saving a lot of money. Instead of testing solely the functionality, don't forget the empathic design. The Google Glass was functional and reliable perhaps, but there weren't any apps that supported the product, so it didn't score high on usability. And most importantly, people felt uncomfortable wearing such glasses, they felt alienated. So while building your MVP, focus on addressing the four elements:
This means, it is key to involve a group of end users early in the project. By going 'out on the streets' developers could immediately learn what they were getting wrong, then deploy further MVPs improving iteratively with each one until the deployment of a final product — developed with and approved by real users. A further benefit: The users in the test group are thoroughly invested in the project and would evangelize it among their peers. This is how the market grows. Or in the context of policies: it gains support by the public.
Exploring the agile policy context
Regarding the formulation of policies, the tough challenge is that these arise within a political environment. These conversations aren't about being right or wrong, but about the choice to agree or disagree. While a policy is aligned with the opinion of the mass, a product is tailored for the individual. Whereas you choose to buy or lease a car based upon your own values and preferences, a policy formulation on when and how to use your car covers different angles of interpretability. Lowering the speed limit to reduce the eco footprint? Well.. others might value time saving over sustainability. Such a complex ecosystem of human behavior is hard to grasp in a physical product, so the remaining question is: How can we?
With the aim to limit the timespan of the policy formulation as technology gets ahead of us, let’s focus on the ways to gain strong evidence and require a short time to set up and conduct experiments. Below a graph shows various methods to experiment, plotted on the axis of gaining strong evidence and the timespan to conduct experiments.
An agile way to approach policy formulation is by following the three areas as steps in the visual above. These steps correspond to the three phases of experimentation below: provotype, prototype, pilot. Of course, this is not a lineair process, as it also includes making mistakes and collecting valuable insights from these mistakes. So, this process is subject to iteration.
You can see a provotype as a design intervention: "a form of inquiry that is particularly relevant for investigating phenomena that are not very coherent, almost unthinkable, and under-specified because they are still in the process of being conceptually and physically articulated." (False & Boffi, 2016,89) It intervenes expectations and models of thinking by investigating possible future scenarios, new concepts or problematic side effects of existing services. These are great in provoking discussion because they visualize and concretize ideas much further than spoken words or written descriptions.
So depending on the phase whether the aim is to discover, to explore or to conceptualize, the act of policy making could follow this explorative structure. What if the government would create Policy Provotypes and the public could respond to these initiatives? This would be gathering feedback fast, open and in full transparency, right? You'll gain the first insights that would steer how to frame the policy, which could lead to the creation of the prototype: the Minimum Viable Policy. This would be a first attempt of a stated regulation, where you'll need to find an organization that would volunteer to test the new policy with, and measure its effect. After that, you'd have collected enough insights to invest in a long-term Policy Pilot. This pilot could be a regulatory sandbox, where a regulator allows developers to conduct live experiments in a controlled environment under a regulator’s supervision. This would require a larger amount of investment in time, money and resources to be able to execute the test. This could be the last step before implementation, as you took the citizens along in the process and therefore reduce the fuzz around it when the policy is made definite.
Some barriers to overcome
Yep.. there are some challenges. The ‘just do it’ mentality is not entirely the way to go. If the plan is just to see what happens, it’s a guarantee to succeed — at seeing what happens — but you won’t necessarily gain validated learning. An important lesson here: If you cannot fail, you cannot learn. And there it becomes tricky for the government.
The idea of being forced to choose the bare minimum and progressively build over time is challenging for policymakers. Just getting stakeholders on board for a single decision is already an extremely challenging task. Let alone having to do it over and over again. In addition, there isn’t a law to freewheel or anything similar (yet). Such a ‘test to learn’ construction is a minefield from a legal point of view, which makes it hard to explore as the outcome is undefined yet. Finally, coping with uncertainty is risky for the image of the government. Would it be accepted if the government structurally would say ‘we don’t know yet’, ‘we are experimenting’ and ‘we are learning by making mistakes’? What do you think?
But the force to work agile is strong..
If you'd only rely on existing market research or conducting a survey to decide what services to offer or regulation to make, the pitfall is that it might reflect what customers thought they wanted. To emphasize the need for prototyping, let's look at the famous quote of Henry Ford:
"if we would have asked the people, they would have wanted faster horses"
By building and testing a prototype instead, albeit a simple role-play of a policy, the government would learn so much more than by asking questions. The added value of the prototype is the ability to:
- gain accurate data: observing citizens' behavior instead of asking questions
- collect latent needs: by putting itself in a position to interact with the actual citizens to learn about their underlying needs
- reveal new information: by allowing yourself to be surprised by the unexpected ways people behave interacting with the artefact
In addition, currently the pressure comes from outside. Let’s face it, the government is always under the magnifying glass of the media. A publication in the media about the police experimenting with facial recognition, leads to concerned questions in the Second Chamber, leads to a statement by the minister that a explanation letter will follow soon, leads to a short window for policy makers with extreme time pressure to formulate a policy or vision. What if we could reduce that external pressure by inviting the public early in the development phase? Here is a great potential for the government to make use of MVP's. And looking back at the rapid pace of tech development, the faster policymakers could go through the Build-Measure-Learn cycle, the faster they learn what is found desirable by the public, the faster policies could achieve that desired effect.
Exploring the Policy Provotype
Use case example: Regulation regarding autonomous cars. Current stage: Discover phase
We can't foresee all the risks autonomous cars could bring. Creating explicit regulation could be the first step in ensuring our safety on the road. So how to come to the right formulation?
- Describe the context and create a provotype of the potential solution (Build)
- Share online and request feedback (Measure)
- Validate your assumptions: heading in the desired direction? (Learn)
(1) Researchers from Israel’s Ben Gurion University have demonstrated a special ‘hacking attack’ in which an image of a stop sign is displayed on an electronic billboard for several milliseconds. The systems in the cars respond to this by braking. The phenomenon is called a ‘split-second phantom attack’.
Because the image can be seen so briefly, people are not aware that the car detects danger. Moreover, such a manipulated billboard is difficult to switch off preventively. So therefore: "We should stop autonomous car sales, as we can't foresee these kind of disastrous risks".
(2) The text above can be published through the media with the choice to either:
- protect (prevent) — create regulation to preclude unsafe situations or incidents (i.e. exclude autonomous cars or autonomous cars could only be used after a thorough checks)
- mitigate (maintain) — create a plan to quickly mitigate the situation after such an incident occurs
- punish (combat afterwards) — but well.. this is a discussion on its own: who should be declared as responsible if a car would suddenly stop in the middle of the highway?)
(3) Then after a week you collect the insights to validate if (for example) people are willing to wait for autonomous cars to be X% risk-free.
Exploring the Minimum Viable Policy
Use case example: Privacy policy regarding geolocation data. Current stage: Explore phase.
Regarding the current debate on privacy issues, there is a unclear line whether something is perceived as invading one’s privacy. If you want someone’s phone number, bet you don’t like the idea that anyone could look it up in Google Maps to check your address and see a direct link to your phone number. However, this was already the case when the telephone book was introduced. Just the act of connecting these databases to make it online accessible probably creates a terrified gut feeling. But that’s an assumption you’d need to test. So build a MVP to validate your assumptions: build a website which mimics Google Maps that shows your phone number at your address, and observe how people respond to it. Collect the feedback and decide for it yourself: is it something we as society would like to stimulate, adjust or stop?
To summarize, below I created a first attempt to translate the MVP in the context of policymaking. So I'm curious… What are your thoughts?
I hope you enjoyed reading this article. It's work in progress, but I hope it sparks your thinking. This article has been my MVP to test whether there is a demand for further exploring the concept of the Minimum Viable Policy. So, if you'd like to be involved in this further exploration, leave a comment below! Other insights or doubts I didn’t cover in this article? I encourage you to start the dialogue! Let’s work towards an agile government, being supported, empowered, inclusive, and a valued public service. Happy to read your thoughts in the comments!
My questions to the reader:
- What do you think should be stated in the Minimum Viable Policy?
- What are your first thoughts about the Policy Provotype and the Minimum Viable Policy?
- What is your opinion on alternatives how to address these issues?
- Liked the article? Please share it, the larger the audience, the more perspectives, the better the dialogue!
Thank you! Cheers, Michelle
Sources:
- Ries, E. (2011). The lean startup: How today’s entrepreneurs use continuous innovation to create radically successful businesses. Currency.
- https://www.slalom.com/insights/mvp-maximally-misunderstood-term
- https://uxdesign.cc/provotypes-how-making-annoying-things-can-help-you-design-better-64f9a0a7e361
- https://medium.com/anoda-mobile-development-agency/what-is-minimum-viable-product-and-how-to-make-it-right-8b885001c6f5
- https://medium.com/@roshansam12/the-case-for-a-minimum-viable-policy-8a0d16d0f97e
- https://medium.com/@j_camachor/how-design-is-politics-5418d9077df0
- https://hbr.org/sponsored/2020/02/how-the-japanese-governments-new-sandbox-program-is-testing-innovations-in-mobility-and-technology
- https://www.slideshare.net/InnovationAcademy/eia2019hk-prototyping-and-design-hacks-alar-kolk
- http://theleanstartup.com/principles
- https://www.nassiben.com/phantoms
- https://www.rtlnieuws.nl/tech/artikel/5190739/zelfrijdende-auto-gevaarlijk-gevaar-remmen-verkeersbord-borden-reclame
