Establishing, running and closing a public sector innovation lab

[Originally published on the Australian Government Public Sector Innovation Network under a Creative Commons 3.0 BY AU licence]

This post should be read in conjunction with ‘Establishing, running and closing a public sector innovation lab – a reflection on the DesignGov experiment‘. 

What do you need to consider in setting up a public sector innovation lab? What is needed to operate one? And what might you need to consider if closing one?

The DesignGov experiment was an 18 month initiative of the Australian Public Service to apply design-led innovation to cross-agency problems. The following piece is based on the experience of DesignGov and is intended to be a resource for others that might be looking to establish and operate a public sector innovation lab. This piece is by no means a definitive guide, but is offered as a personal contemplation about what should be considered when establishing, running and, possibly, closing a public innovation lab.

This advice should also be qualified by noting the evolving nature of public sector innovation labs and their practice. There is no best practice for a field that is still so new. The material is offered in the hope that it might reduce some of the learning costs. It is also free to be built on and improved from the experience of other labs.

This paper can be read in conjunction with ‘DesignGov – the context for its establishment’ which describes the setting in which DesignGov was developed.

A checklist for a lab

The following provides a checklist for decision makers and those involved in considering establishing a public sector innovation lab. The underlying material is provided in the other sections.


Checklist item y/n
Does the proposal for a lab make it clear what sort of problems the lab will be expected to tackle?
Does the proposal for a lab identify or make explicit what types of problems or projects the lab will not do?
Does the business case identify what aspects of projects the lab will handle? (e.g. will it be responsible for understanding problems, or for identifying solutions, or for testing and implementing the solutions, or will it do all of those things?)
Is it clear who will provide back-up or support to the lab if it runs into unexpected difficulties or it requires access to unexpected skillsets?
Are there identified champions or positional leaders who support the lab?
Are there identified ‘back-up’ champions and supporters who are willing to champion the lab if needed?
Have the skillsets available for the lab been matched with the scope of the lab?
Has an assessment been made of whether the lab will be run as part of the public service, or delivered by the private sector or under a not-for-profit arrangement?
If delivered ‘outside’ of the public service, has a strategy been identified for ensuring that the lab is seen as legitimate and trustworthy by public servants?
If delivered inside of the public service, has consideration been given to how the lab can have the flexibility to test and stretch due process where necessary?
Has consideration been given to how the lab can best draw on the ‘edge’ while still having access to the authority and impetus of the ‘centre’?
Does the proposed timeframe for the lab reflect the proposed complexity and newness of the lab? (e.g. if it is a lab internal to an existing organisation, drawing on existing resources, processes and networks, it will likely need less time to be effective than one that spans multiple organisations, that requires a new business model, or that has to build and establish an identity and capability)
Have the existing relevant networks been mapped?
Is there a sense of how the lab will access or interact with those networks?
Will the proposed governance arrangements provide sufficient authority for the lab to do new things?
Do those involved in the governance of the lab have familiarity/working knowledge of the methodologies and approaches that will be used?
If not, can the governance group include or give access to recognised peers/experts who can vouch for these techniques?
Do the proposed funding arrangements for the lab support the intended business model?
Will the proposed resources be sufficient and provided in time for the intent?
Will the proposed funding arrangements assist the expected work programme of the lab?
Is there a clear sense of the ‘supply chain’? (e.g. where and how the problem identification will come from, and at what stage and to what extent other groups/organisations will take over the problem and the introduction of the response, and who those groups or organisations are)
Is the proposed physical accommodation for the lab suitable for the work of the lab?
Will the physical space for the lab contribute to:
Different ways of thinking
Easy access for likely stakeholders, clients or partners
Flexibility in types of activities (workshops, seminars, training, brainstorming)
Is there a proposed business model for the lab?
Is the business model flexible enough to take account of lessons and changed circumstances?
Have the core assumptions for the proposal been made explicit and tested?
Are there mitigating strategies or options available if any one of those assumptions is wrong?

Running the lab


Checklist item y/n
Is any intention for ‘quick wins’ balanced by with a mechanism or systemic force to ensure experimentation and risk-taking (e.g. a goal of ‘quick failures’)?
Does the lab have access to existing relevant networks and communications channels?
Does the lab have the capacity and capability to communicate for diverse audiences and to address different preferences?
Has consideration been given for how the work of the lab will be ‘embedded’ beyond its immediate team?
Will the lab use secondees/bring in staff on rotation from other areas?
Has thought been given to how to introduce and immerse secondees quickly?
Are there processes to support the secondee on return to their normal work area to apply their new skills/outlook?
Has thought been given to how events can be used to:
Build engagement with the work of the lab
Introduce staff from other areas to new methodologies/insights in a ‘safe’ way
Expose examples and highlight cases and stories that show why new approaches are needed/can be used
Build familiarity and trust networks
Does the lab have the ability to test boundaries and existing conventions?
Is there a strategy to mitigate for any potential ‘responsibility inflation’ where new approaches require permission and approval from high level decision makers?
Does the lab already have significant buy-in/engagement from relevant decision makers?
If not, has thought been given on what methods can be used to encourage buy-in?
Have reporting obligations considered how to balance engaging and informing decision makers with providing flexibility?
Has a developmental evaluation approach been considered for the lab?

Closing the lab

Checklist item y/n
Are there processes in place for capturing the history and experience of the lab?
Are there processes to capture lessons as they are learnt?
Is there the opportunity to revisit lessons after the closure of the lab?
Is there a structured process for ensuring that implications/insights from the model are considered by senior leadership?

Establishing the lab

There are any number of things that should be considered in the establishment of any new initiative, many of which are covered by existing project management frameworks and literature. The following items are offered as being particularly pertinent to a public sector innovation lab.

Defining the intent

What is the intent for the lab? What is it that current practices are failing to address that a lab is intended to address?

There should be an explicit and deep understanding of the intent for the lab and why the current state of affairs is unsuited or incapable of delivering on that intent.

Innovation in the public sector is often problem-led. Unless the problem or existing failing of the current system(s) is truly understood, believed and agreed, the lab will likely face an uphill battle in getting traction.

In addition, without a clear intent it will be hard for a lab to be seen as having a mandate to be involved in problems that are seen as core business by other areas. Alternatively, without clear intent it will be hard for a lab to get others involved in a problem that they do not see as relevant to their core business.

With DesignGov, the case was done in part through the report Empowering Change: Fostering Innovation in the Australian Public Service, the “APS Innovation Action Plan” and internal work through a project undertaken by the senior leadership of the Australian Public Service (the APS 200 project). There was consistent agreement as to the need for something to be done, though in retrospect, it might have been valuable for there to be a stronger case of how it would fit in a broader ecosystem and supply chain.

The business case

What is the case for the lab?

As opposed to the intent which should outline the why and what, the business case of how is also important. Again, much of this is standard project management, however there are some potential differences.

The business case needs to either establish or elicit what the level of support is behind the proposal, including from key areas that may be connected but not within the administrative ‘ownership’ of the lab.

For instance, DesignGov was located within one portfolio (that of the Department of Industry) but had explicit commitments from four other agencies (the Departments of the Prime Minister and Cabinet, Immigration, Education, and the Australian Taxation Office). There were agreements covering the expectations and roles of each of these parties.

If possible the support listed should cover not only financial commitments, but also establish key liaison points (and ensure that successor liaison points can be found if necessary if there are staffing or structural changes within the agency), as well as available capabilities and back-up that can and will be provided.

DesignGov’s operations sometimes suffered for not having an explicit back-up for certain capabilities or ability to call on unexpected capabilities that were needed at short notice. As an experimental initiative, a lab is unlikely to be in a position to know exactly all of the capabilities that are needed up-front. They may vary between projects or because of the staffing profile of the lab or because a project or issue goes in unexpected directions.

Leadership support

Is there explicit and high-level leadership support for the lab? Are there multiple supporters in case of leadership change?

Leadership support is vital for something like a lab, particularly if it sits across multiple agencies, has unusual or multiple reporting lines, or sits across siloed activities. Senior leadership backing is also needed to ensure that the lab has legitimacy with stakeholders and to emphasise to other public sector agencies/areas that the lab’s operations, which may seem unusual or not core business, are important and not a frolic or distraction.

For instance DesignGov had the endorsement of the Secretaries Board (the most senior decision making group for the Australian Public Service) and a governance Board that involved a number of the individual Secretaries.

It is also important to have multiple leadership ‘patrons’. In an increasingly interconnected and changeable world it is not sufficient for innovation initiatives to be championed by a single leader. There is the risk in having a single patron that they may change roles mid-way through the establishment or at a critical juncture of the lab’s development, leaving it without high level support at a key point. Wherever possible (though noting it will not always be possible) there should be a coalition of high level support for the lab prior to its establishment, even if that is led by a primary patron.


What skills do you need to have involved in the innovation lab? How might the operations of the innovation lab differ depending on the skill-sets it has?

This is a very difficult question to answer with specifics. In general, a lab will need to be able to draw on a range of disciplines and techniques, and the relevant mix will shaped by specific projects or initiatives the lab is to concentrate on. However some core skills to consider including in the mix are:

    • Design or design thinking – this is a useful methodology simply because it is a discipline that is based around integrating multiple perspectives. In many ways a lab is about facilitating empathy and understanding of different perspectives. Design can help manage those conversations in a structured and constructive fashion
    • Behavioural scientists, ethnographers, anthropologists and historians – much of what a lab will look at will be about better understanding problems and may need immersive research techniques. Any of the social sciences will likely be able to provide not only the research skills but also the ability to assess and measure qualitative evidence. This will help when dealing with complex problems without a clear causal relationship
    • Policy and service delivery – no matter the specific projects, labs should consider having a full appreciation of the ‘supply chain’ in the public sector, from the policy all the way to delivery. Even if the projects focus exclusively on service design or co-design of programmes, they will still likely have implications for policy development. Each end of the policy spectrum tends to have different ways of approaching issues, and it is advantageous if the lab has an intimate understanding of  both
    • Networking and advocacy – as an initiative that deals with problems that do not fit neatly elsewhere, labs will often attract the attention of other ill-fitting issues and those that are plagued by them. A lab can play a role as connector or broker within a wider ecosystem. To take full advantage of this position in the system though, a lab needs a strong skill set in networking and advocacy, of connecting disparate groups and divergent perspectives and ways of thinking
    • Communications – explaining what and why is vital for a lab. A lab is often going to be seen to be encroaching on contested space. It needs to be able to explain clearly and strongly why it is involved and what value it can bring. Communications will also be vital in ensuring insights and proposed solutions, which by their very nature of coming from an innovation lab will be outside of the norm and may look unconventional, appeal to and are understood by diverse audiences
    • Logistics, event management and all things practical – along with communication skills there is a strong need for the practical and the ability to bring together and run events that involve different players and groups. These will likely run from workshops looking at problems, to bringing together decision makers and facilitating discussion in such a way that different perspectives, insights and forms of evidence are heard and understood, to promotional and training/professional development events. Again, by being a locus for doing things differently, a lab needs to continually engage to help bring people along on the journey
    • Administration – a lab may work across silos, try new approaches, advocate doing things differently, or partner or collaborate with different groups than usual. To enable any of these, the lab will need to engage with and master administrative processes.

The exact nature of the skills needed will, of course, be dependent on the lab’s context and foundation. However the skills base available should also factor in to the scope and ambition of the lab.


Should the lab be purely within the public sector? Can a public sector innovation lab be effective if it is outside of the public sector? Are there any effective partnership or hybrid models that can be applied?

While these questions probably cannot be answered definitely, the place of the lab in the system should be carefully considered.

If it is on the ‘outside’ of the public sector, a lab will need to be aware that it may be seen as ‘other’ by public servants and their organisations. Innovation and experimentation can make people feel vulnerable, and trust and familiarity is an essential ingredient to being effective. An outside group comes with an inherent sense of being from the outside, and not necessarily appreciative of the context, the practices and traditions, the ‘way things work’, or the sensitivities and risks that the public service has to deal with. There may also be confidentiality or political concerns that public servants will not want to share with those who are not fellow public servants. It is also generally easier for a public sector organisation to connect with and leverage the work of other areas and to build a common sense of public purpose/value.

On the other hand, if it is inside the public sector a lab can have many other issues to deal with instead. An experimental body such as a lab can have trouble with traditional approaches to risk management, project management, governance and reporting processes. It can be difficult to find an appropriate balance between due process and dynamic flexibility.

None of these issues are necessarily insurmountable, but they are worth considering as part of the strategy for a public sector innovation lab.

Alternatively there are other models that might emerge, such as the mutualisation of the Behavioural Insights Team in the UK. Being an experimental organisational form, it is likely that labs will develop in a variety of ways, some of which bring together public and private sector partners in new and exciting ways.

Centre or the edge

(If part of the public service) Should the innovation lab be situated at the centre of a public service, or should it be on the edge? How will it best be able to access the insights needed for innovation from the edge, yet have the authority from the centre to implement it?

DesignGov was in many ways on the ‘edge’. It was a partnership between multiple government agencies, had high level backing from senior leadership, and it had involvement from the central agency (the Department of the Prime Minister and Cabinet). But it was at the intersection of multiple rule sets and processes, it was taking a citizen (business) view of government processes in its project on business and government interactions, and its reporting and governance were to the side, rather than with direct access to all of the relevant decision making bodies. It had no direct mandate or political imperative.

This had many advantages. Innovation is fed from the edge – that is where the problems, the issues, the unexpected or extreme cases, the situations that don’t fit the normal procedures, often arise. The edge is more exposed to different voices and different perspectives. It is inherently more open to seeing problems (and possible responses) in a different way. It is more likely to pick up on emerging trends or weak signals that carry the possibility of significant change.

But the centre is where decisions are made, and more importantly where decisions are integrated and implemented into the business-as-usual. The centre can push decisions that the edge can only help to nudge or influence. The centre has the ability to direct attention in a way that the edge can only hope to do.

The centre though is also focused on current activities and current priorities. It can be hard for the centre to engage with risk in different ways or to expose itself and even encourage counter-narratives and voices.

Ideally a lab would find some way to bridge the centre and the edge, drawing on the strengths of both. However this is a tension that probably needs to be considered on a case by case basis, depending on the institutional context and the specifics of the organisation(s) a lab is in.


What timeframe is there for the lab to prove itself? How realistic is that in terms of delivering the achievements and results expected of it?

DesignGov was an 18 month pilot, against an initial recommendation period of three years. In the experience of the team, the 18 months felt insufficient for a number of reasons. It may seem like a long period, but there are a number of tasks that need to be considered:

    • Developing the business model of the lab in practice (as opposed to in theory)
    • Building up networks and connections with various ecosystem players
    • Making sufficient progress to be able to point to achievements
    • Developing sufficient rapport and understanding with the relevant governance group(s)
    • Mapping out and tapping into relevant expertise
    • Essential establishment matters such as recruitment, accommodation, logistics, and administrative practices.

The timeframe for some of these will be dependent upon how much the lab inherits or simply adopts the practices of one (or more) of the organisations that are supporting it, however there will likely need to be some flexibility so that the lab is sufficiently different in practice in order to be effective.

The timeframe required is probably related to the number of organisations or the extent of the scope. If the lab is internal to a single organisation, it is likely the establishment and other matters will be far faster.

For a lab that spans multiple organisations, a period of at least six months is probably needed just for developing the business model of the lab, building networks and connections, and establishing how the lab fits within the broader system. If at all possible, the lab would likely benefit from the first six months focussing on the establishment matters, with limited regard to specific project deliverables or outcomes.


What networks exist and how will the lab connect with them?

An innovation lab generally fits more with a network approach than a traditional hierarchical bureaucracy. A lab rarely has the size, scope or resources to do everything itself (if it did it would not really fit with the notion of a lab). Labs are thereby reliant on networks and being able to access or leverage the capabilities (and possibly resources) of other groups.

DesignGov was heavily reliant on multiple networks:

    • The design community across APS agencies and outside of the public sector within Australia to provide expertise, access and insight
    • Individual internal agency communications networks, particularly for the project on improving business and government interactions
    • A number of other private sector, non-profit and public sector networks that all gave reach, insight and channels for dissemination, research and feedback.

As a small player in a much larger ecosystem, dealing with issues that had multiple stakeholders and agencies with policy/delivery responsibility, DesignGov could not have been effective without tapping into these various networks.

However building relationships with such networks, and identifying the value each bring and that a lab in turn may offer them, takes time. This is aided by mapping out the various networks early on and developing some rough notions about how they connect and where and the lab might fit in the ecosystem.


How will the lab be governed? How does that fit with any intention for the lab to operate flexibly, to undertake experimental activities with uncertain results, and for it to do and deliver things different to what is done now?

DesignGov had a high level Board overseeing it, involving agency heads and an external design expert, as well as the CEO. This Board in turn reported to the Secretaries Board, the most senior decision making body for the Australian Public Service. Towards the second half of its pilot period, DesignGov also had an International Advisory Group of eminent designers and innovation practitioners.

From that experience, what should be considered in developing the governance arrangements for the lab?

Any governance arrangements should establish sufficient authority for the lab. If the lab is reaching across silos, then it needs to not only have the permission to do so, but the mandate and authority to get the participating areas to meaningfully engage and participate.

A balance needs to be reached between seeking approval (and thus authority) and keeping oversight to the minimum needed for the lab to report and be accountable. An experimental initiative will always be doing something new, and if everything new needs approval, the momentum of the lab’s activities will suffer. At the same time, if there is not explicit endorsement for the lab to be doing something new, then there may be resistance from other areas who may question ‘why?’ something is being done differently, and not fully cooperate until that authorisation is given or made clear.

In addition, what experience and familiarity will there be with the techniques and work of the lab in the proposed governance group/arrangements? If the work of the lab will be completely new to them, then consideration should be given to how the members of the governance group will be given comfort that this is ‘within scope’ or that this is how things work. Will they have access to experts or peers that can help them get across the approach of the lab? Can they be given an introduction to the lab and how it will operate by someone that is recognised as being authoritative?

Finally, depending on the scope of the lab, has it been identified how the governance arrangements fit with political ownership or authority for the work of the lab? If the lab is cross-agency, cross-sectoral or cross-jurisdictional this may be significant.


How will the lab be funded? What will that funding be contingent upon? Will the funding arrangements support a collaborative approach to the work of the lab? Is there a way that the arrangements might distract from core work?

DesignGov had a small core budget, supplemented by some in-kind and financial support. Some of this support was given as a result of project prospectus prepared for its business and government interactions project.

The small resource base had the advantages of concentrating efforts and ensuring that external capabilities were actively sought and leveraged, and that all aspects of the work of DesignGov was done on a ‘co’ basis – cooperatively, collaboratively, co-designed, co-delivered. It also meant that DesignGov was not seen as a body that was here to take over or contest issues, but was there as a contributor.

On the other hand, the small resource pool limited the independence of DesignGov and its ability to undertake some core functions. For example, additional investment funding had to be secured through a prospectus to enable professional designers to be recruited for the large demonstration project. This was only achieved over halfway through the pilot period.

The appropriate funding format will vary between institutional settings. In the current environment, most labs will likely be expected to contribute to a broader cost-reduction agenda. Labs can certainly contribute to this, however some care does need to be taken to ensure that expectations are realistic.

Where does the lab fit within the ‘solution’ pipeline?

Is the lab going to be responsible for investigating issues and providing new insights to reframe problems? Is it going to be responsible for prototyping and testing? How will it be involved with delivery of any developed initiatives?

DesignGov was given a broad problem space with the topic of improving business and government interactions. This was an issue involving multiple stakeholders and multiple agencies that had a policy, program, and/or service delivery interest. It was a problem area where any proposed initiative would be building on top of multiple previous initiatives and existing work agendas. Any recommendations would therefore require significant additional work in terms of prototyping and eliciting further feedback about the problem and the needs, in developing the responses, in devising and delivering implementation plans, and identifying where responsibility for these tasks should be given.

Consideration should be given to where the lab fits within this broad ‘solution’ pipeline. At what stage will things come to the lab? When and under what circumstances will things leave/move on from the lab? This is a particular issue when a lab sits across agencies and is expected to bring a citizen or whole of government view. The resultant ideas, by the nature of coming from a different perspective, will probably not fit neatly with existing organisational responsibilities and so the natural ‘owner’ may not be readily apparent. Even if the appropriate owner is clear, that does not necessarily mean that they will readily wish to, or be resourced to, take on this initiative.


Where will the lab be situated physically? How does that fit with what is intended from it?

The accommodation for a lab matters. An innovation lab requires different styles of working, some of which are not so compatible with traditional conventions, practices or spaces.

This is not about having something ‘cool’ for the sake of it. There are very good reasons why there are shared characteristics for many innovation and design spaces. Exploratory and collaborative work that frequently involves external partners really requires spaces other than the standard (for the Australian Public Service) open plan office.

DesignGov had the experience of using or borrowing facilities for events across Canberra (and sometimes interstate) and operated within four different accommodation settings. Each brought its own benefits as well as challenges.

Accommodation for a lab ideally needs to provide for:

    • Supporting collaboration – design and innovation work requires a lot of close collaboration and the need for workshops and the ability to quickly test ideas and put up and modify ideas. Walls with whiteboards and spaces to stick paper, sticky notes, artefacts and drafts are great – whiteboard walls are even better
    • Flexibility – ideally the venue will allow for (easily) changing the shape and size of the space(s) for different types of workshops and settings. The ability for team members to move to a handy discussion space for impromptu discussions is also invaluable, especially at the early stages where many matters will need continual discussion, iteration, and refinement
    • Access – easy access for visitors (such as workshop participants is important). This includes basic considerations like public transport and parking availability
    • Difference (but not too different) – the space needs to encourage and allow people to think and operate differently, or risk people staying in the same mental patterns. At the same time it cannot be too jarringly different or outside of people’s experience so as to make them feel uncomfortable.

In truth, a lab will likely be able to work from most settings – however, due consideration of these sorts of factors can make a remarkably big difference to the effectiveness and the attitude of the lab’s team.

Building the business model

How will all of these elements fit together? What will the business model for the lab be? What is its value proposition and are the preconditions for it being able to deliver on that proposition in place?

If the lab is experimental, it is highly unlikely that these questions will be answerable beforehand. Even if they are, the answers will likely change once the lab is underway and once the available skills, opportunities and capabilities are truly understood. Still, there is value in building a picture of what is intended and what is expected. This should not necessarily be prescriptive, but rather ensure that consideration has been given in the proposal for a lab of a greater variety of issues.

Fundamental assumptions

What are the fundamental assumptions being made in the proposal for the lab? What would happen if any of those assumptions was overturned or changed by circumstance?

The establishment of a public sector innovation lab is not a science.  As an addition to a likely existing complex ecosystem, the introduction of a lab may change the nature of the system itself in ways that are not expected. The existence of the lab may change how other parts of the system interact or inter-relate. It may give stakeholders a different avenue to pursue a particular agenda, or it may reveal new ways of thinking that challenge existing processes and assumptions.

A proposal for a lab can be strengthened by making explicit the fundamental assumptions. This will help test the resilience of the proposal and identify possible weaknesses. Many of these assumptions for instance may be around continuity of existing environmental factors such as key champions, leaders, key staff, or available support. Can the proposal stand as it is if one of those assumptions is wrong? Options for mitigations of such a scenario may be worthwhile, and could be incorporated into the proposal.

Running the lab

This next section outlines some of the considerations involved with running a public sector innovation lab.

Quick wins / fast failures / value proposition

Does the lab need to achieve ‘quick wins’? If it does achieve quick wins, might these limit what the lab can work on in the future? And as an experimental initiative, does the lab have the capacity to celebrate ‘fast failures’?

With any experimental initiative there can often be a pressure (and an appeal) to identify ‘quick wins’, ‘low hanging fruit’ or small victories that demonstrate the value of the work and the approach, which in turn will then build support for the broader work programme or agenda and more ambitious wins.

However a lab should also consider whether aiming for quick wins might encourage a focus on the known and the achievable rather than truly testing new areas. There is a risk that quick wins might shape expectations downwards and ‘contain’ the vision for the lab so that it never actually gets to realise the opportunity for more ambitious aims.

Of course a focus on delivering value as soon as possible will always be a justifiable expectation. A possible strategy might be to balance this with a declared expectation or explicit goal of celebrating fast failures, by identifying small high risk experiments that are unlikely to succeed completely but are worth trying, if only for the insights that they will provide. An approach like this might help the lab balance a drive for results with an equal appetite for risk and experimentation.

Due thought should be given to making sure that the lab has the opportunity to focus on its core value proposition and intent without being too focussed on achieving small scale wins that are likely to fit with more ‘business as usual’ or process improvement approaches.


How will the lab communicate its work?

Effective communications capabilities and skills are essential, as is the ability to tap into existing communications channels. Government stakeholders (internal and external) generally have set patterns for how they relate with a government, and creating new ones can take considerable time and effort. Where possible, a lab should identify and tap into existing channels, using introductions by established/trusted players. This is especially important for any group undertaking project by project type work where the stakeholder groups may differ between stages or for individual projects.

It should also be recognised that it is really quite hard to communicate new approaches as they are being developed and tested. It tends to be easiest to communicate things that are practiced and well established – your audience generally then has ready comparisons and mental models they can use to receive and process that information. When things are fluid and exploratory, it can be hard to articulate where the process is at, let alone to communicate it in a concise, clear and accessible form that appeals to a diverse stakeholder group with varying levels of commitment and connection.

Consideration should also be given to being able to readily communicate across different styles, such as a preference for visual over text, narrative or factual, conceptual or practical. When undertaking experimental activities that might challenge people’s existing patterns or understanding, the more that the messages can be conveyed in their preferred thinking style, the easier it seems to be to help them get across the new idea.

Of course, different decision makers and stakeholders will have different preferences, so it is advantageous to have the capability to provide for multiple styles in the one set of artefacts. Labs that sit across agencies will also have the challenge of dealing with multiple sets of communication preferences as each organisation will likely have a prevailing expectation and style that is favoured by their own senior decision makers.

The experience of DesignGov was that it is also advantageous for communications activities to be built on an accessible and distinct Internet and social media presence. When a lab is doing something new, it needs to be easy to find and have clear and simple messages about what it is doing and why. It is also important to try and be as transparent as possible about what you are doing. When entering a new field it will usually be easier for those with a shared interest to find the lab, than it will be for the lab to find them. In the case of DesignGov the blog was a very valuable tool for the team as a means of making connections that would never have been made without people making contact with the team in response to something that had been written and published.

Secondments, training  and professional development

How will the lab build wider practical understanding of its work? How will it share its skills and understanding of its methodologies beyond its immediate team? What will it do to help legitimise its approach and decrease the ‘otherness’ of the different approaches it uses, and build wider acceptance of alternate ways of doing things?

DesignGov had the good fortune of involving a number of talented and bright secondees from a range of different agencies. This helped DesignGov tap into the expertise, skills and networks of other agencies, and helped ensure that the agencies were engaged with the work of DesignGov.

It also provided an opportunity to immerse the secondees in design and innovation in a way that they would be unlikely to have the experience of in their normal work environments.

If there are a few staff being brought in at the same time or if there is a high rotation, thought should be given to how the seconded staff can be brought up-to-speed on the project and the techniques and approaches of the lab, while also being given the opportunity to contribute and participate as soon as possible.

Another aspect is to consider what happens at the end of any secondments. Equipping people with new skills and experiences and the willingness to apply different methods in their work is the (relatively) easy bit. It is what happens next that is harder and more significant:

    • What support will the secondee have when looking to apply new or different skills when they return?
    • How can they be helped to bring new skills back to their workplace and help their colleagues apply them as well?

Approaches for tackling this might be to ensure that a senior sponsor or mentor has been identified, one who can help make sure the newly learnt insights are not buried underneath the business as usual practices of the workplace.

Alternatively a lab might seek to offer professional development or training activities. Again consideration should be given to how the training can be offered in a way that can be applied back in the workplace and that will integrate with existing processes.


How might the lab use events to help build engagement and further the understanding of its work/techniques/projects?

Events can be a great tool for an innovation lab. Events can be

    • Used to build awareness and engagement – for instance one of DesignGov’s first events was a film night of a documentary about design thinking which was used to both engage with the existing design community, to make introductions and to help identify interested parties and partners
    • Expose the centre to ideas in a safe way – events, particularly workshops, can be a great method for showcasing different techniques and methodologies in a ‘safe’ way. Workshops can help provide practical introductions to new approaches without challenging existing work practices of people in the centre/in priority activity areas
    • Highlight the long tail – events can be used to illustrate or showcase the more extreme or unusual cases that people outside of the lab might be unfamiliar with or not see within business-as-usual work. Public service organisations often have a tendency to focus on the mainstream, whereas an innovation lab will likely be exploring the cases which do not fit neatly into existing categories. Events can potentially help in revealing the wider scene and expand the horizon
    • Build familiarity – events can also be used to help build familiarity, and thereby trust. A lab might consider how it can use events to provide value to key stakeholders (e.g. by exposing them to interesting and relevant insights, practical tools or expanding networks), and thereby build familiarity. This can provided significant value for a lab when it has unusual requests or invite participation in things that may not seem (to the stakeholder) as immediately relevant.

Permission to test the boundaries

Does the lab have permission to test the operational processes that it is expected to operate under?

Any public sector organisation will have to abide by appropriate processes and be accountable and exercise due caution. In applying new techniques, trying new approaches or working in different ways, an innovation lab will likely reveal limitations or constraints offered by existing rules and processes that have been developed for more conventional settings.

DesignGov experienced the intersection of multiple rule sets and processes of partnering organisations. This setting sometimes revealed conflicting or varying requirements and an as yet still developing infrastructure for effective collaboration and partnering. An innovation lab that cuts across any silos is likely to encounter such situations and will either need to be resourced to negotiate this rule set hierarchy, or have the freedom to go with the least complicated option.

There may also be tensions between different methodological approaches and existing agency processes. The value of different approaches is that they provide different insights that come from doing things in different ways. At times these different ways will not always fit neatly with processes developed for other settings (for instance project management is not always a good fit with an iterative design process, or post-hoc evaluations with exploratory endeavours).

An additional complication might be something that was referred to at DesignGov as ‘responsibility inflation’. Whenever a new approach or a new way of doing things is to be tried, especially one that cuts across reporting lines or agency jurisdictions, the decision about approval will usually be escalated up the decision making tree. If an innovation does not fit neatly within a manager’s responsibility, their experience or where they are uncertain as to whether they have the authority to make the judgement, a common reaction is to escalate the decision making to someone more senior.

This can have the unfortunate effect of escalating a decision to the point where the authority to make the judgement call is clear, but the direct involvement or familiarity with the subject matter is removed. In such situations, an innovation lab may need to find an appropriate balance between ensuring appropriate approval and devoting significant time to briefing and familiarising senior decision makers who will likely have significant existing commitments and be time poor.

An innovation lab may need to consider how it can best work with existing processes and also have the freedom to advocate new approaches and seek exemption or exception so as to have a given level of discretion.

Getting buy-in

How will the lab get buy-in from its partners and relevant stakeholders?

One of the strengths of design as an approach is its ability to elicit the stories and real use-cases of people and their experiences. These stories and examples can help encourage empathy and understanding of issues and problems in a way that more traditional forms of evidence may struggle to do. Design methods can help provide a human face to the problem.

Capturing these stories in an effective and accessible way requires careful consideration. As an example, if it is to be done by video, how will this be provided to decision makers? In the Australian Public Service systems and briefing processes tend to favour documents over video.

Some processes seek to involve the key decision makers at key stages – for instance having them listen to the stories of key individuals in person. If this option is chosen, how will you ensure that the decision maker is both available and focussed for the necessary amount of time?

Alternatively, the buy-in might be through requiring co-investment in the project, having the decision-maker commit financial or other resources towards the project. However this approach needs to consider ensuring that the process is not then seen as a ‘fee for service’ type arrangement where the lab is expected to provide ‘the answer’. Any investment arrangement should reinforce the importance of the process being one of co-discovery rather than buying a product.

Whatever the approach chosen, care should be taken in requiring the involvement of key people from the relevant supporting area(s). Otherwise there is a risk that the lab might face a scenario where it puts forward recommendations but the understanding of the underpinning insights is not shared by the recipients.


How will the lab balance evidence and exploration?

If the lab is dealing with the policy space, it will likely be dealing with complex or wicked problems in the context of a larger complex system. A fully evidence-based approach will probably not be suitable. If it is a problem to be given to an innovation lab, then it will either not be well enough understood or the problem has evolved in ways that mean the evidence is yet to be gathered.

In such an environment the lab will not have a firm grasp on the cause and effect related to the problem. It may also be dealing with a large number of players, each contributing and responding to the problem or issue in different ways.

The lab may also face the problem of having vast amounts of data and research about the problem or issue. Given the early stage of most innovation labs, it is likely that other, more traditional avenues will have been applied to the issue first. The problem may also have been around for a considerable time, in which case there may have already been multiple interventions that the lab is expected to build on top of.

In this situation, the lab will need to find a balance between getting across all of the existing research, data and evidence, and discovering and providing new insights that help the problem or issue be seen in new ways.


How will the lab balance reporting to ensure that decision makers, supporters and champions are informed and involved, yet retain flexibility and agility?

DesignGov had a small team, yet in order to usefully connect with key decision makers and supporters from relevant agencies, as well as informing its governance board was fully informed, it had to commit significant effort to its reporting responsibilities.

This can be an issue for many endeavours, but is potentially exacerbated for labs where they involve multiple reporting lines with diffuse connection, where the projects/work of the lab is seen as relevant but not immediate, and where the business model/working style of the lab is still being developed.

The right balance will likely depend on a range of factors and current practices, but a lab will need to consider how it can report in a way that sufficiently informs and engages decision-makers, meets accountability obligations, equips champions and supporters with the information they need to assist/advocate, but that does not prove overly burdensome.

Assimilating different methodologies

How will the lab work to draw together the different methodologies that it uses?

An innovation lab is likely to be open to, require and draw on multiple methodologies and approaches, some of which will be new, practiced but novel to the context, or well-known but evolving.  For instance, DesignGov drew on horizon scanning, design thinking, open innovation, behavioural economics and other methodologies.

A lab is a great opportunity to explore how such varied approaches can be drawn on to reconceive problems and devise new responses. A cross-disciplinary and cross-methodology approach can be extremely enlightening and empowering. A lab can provide the necessary flexibility to test and try combinations of these.

However, the lab also needs to consider how such a divergent approach will be received or understood by other areas used to, and often required, to operate in a more conventional or tested fashion.


How do you know if the lab is being effective?

Assessing effectiveness and impact can be difficult for any innovation, where the existing measurement frameworks and key performance indicators will naturally suit or reflect what has been done before. It can also be particularly difficult in the public sector where there are multiple factors and actors at play and where the effectiveness of any one bit of work can be hard to isolate or assess.

However, if the lab is dealing with an issue in the ‘white space,’ involving but between multiple government agencies and stakeholder groups; where there might already be significant existing work and initiatives; and where the lab is dealing with a system level issues rather than discrete activities, then there is an additional layer of abstraction involved.

It is important for those involved with a lab to recognise that it assessment of its impact will not be easy and that any measures are unlikely to be readily agreed as the right measure by some (or potentially all!) stakeholders.

The experience of DesignGov suggests that labs are more suited to a developmental evaluation approach rather than a traditional ‘after the fact’ evaluation. DesignGov involved a lot of sensing and testing. In such a setting, it may become apparent that a change of course is required. An ongoing developmental evaluation approach can identify those signposts at the time, rather than later, and allow a change of direction when it is needed as opposed to recognising it after the fact.

The DesignGov team conducted fortnightly internal review sessions to ensure potential issues were identified and discussed early on. There are a range of approaches that can be taken, but if possible, evaluation should be a built-in and ongoing feature in the initial phases rather than something that happens at the end or at set periods.

Closing the lab

What is involved in the lab is to be closed? How can the lessons be shared and used to inform other practices? This section covers some of the matters that should be considered when a lab is closed.

Capturing the history

What processes are there in place for capturing the history and experience of the lab?

Just as has been attempted here, there may be value in capturing the history of the experiment and its development.

One of the interesting aspects of trying to learn from innovations is that it can be very difficult for those involved in the experiment to appreciate what will be most relevant or important for someone coming to it from a different angle or context. It is also easy to assume that certain facts or developments will be well known, when this may not be a wise assumption. It may be best to ensure that all the aspects/contextual factors are explained.

At the risk of providing too much information, it may be advisable to provide a fuller context than a shorter one.

Labs should consider how they will document their history from the beginning (as it is easier to do while it is happening, than it is to revisit it after the fact).

Distilling lessons

How will the lessons of the lab be distilled and refined?

The same logic applies to distilling the lessons from the experience of the lab. It is easy for lessons to quickly be taken as given by those involved in the lab, when they may still be new or hard to understand for those not involved.

DesignGov had a fortnightly (mostly) review process which helped identify insights and issues that might be of interest. Maintaining a blog or other sort of official ‘diary’ for the lab can help provide the discipline to continually distil and document lessons, that might otherwise be taken for granted or forgotten due to the pressure of more urgent priorities.

Of course, some lessons and insights may only become apparent or cogent after some time after the lab has closed. There is value in revisiting the experience with the benefit of hindsight and distance which can provide greater clarity and filtering of relevance.

Working with stakeholders to transition

Similarly, a lab will be greatly aided by having effective processes and systems for keeping track of stakeholders and projects. This can aid in transition arrangements or in communicating the end of the lab, and ensure smooth hand over processes for particular projects, artefacts or investigations.

Integrating implications of the model

What are the implications of the model, beyond the immediate context?

DesignGov was an experiment on a number of different fronts, including testing a range of different approaches, concepts and business model ideas. While the pilot itself was not continued, there are possibly significant implications for how the public service works more generally that could be drawn from the experiment.

In situations where this does apply, how will those implications be conveyed to senior leadership? If the lab has been closed, is there a formal channel or process by which the implications can be conveyed even if the lab/experiment itself is not continued? How will this be done?

This is again something that a lab could usefully consider at its inception.

Other matters to consider

The previous three sections outline many of the practical considerations that might be involved in a lab. The experience of DesignGov showed that there are some other factors that do not relate directly to the day to day operation that might be also worth reflecting on.


How might the lab mesh with or conflict with the dominant existing cultures or subcultures within the agencies that it has to work with? If the lab intends to have a different culture, what might that mean if or when the cultures clash?

Innovation is inherently about change and difference. Bureaucracies are about the routine and the perpetuation of existing processes. An innovation lab and other agencies are likely to have differences in cultures (or at least their cultural ambitions).

DesignGov was a small dynamic team involving a range of different disciplinary perspectives and backgrounds in a high pressure environment attempting something very different from the organisations from which it originated. While all of these things will vary between people, wider cultural and organisational contexts, and missions, there may be a some matters that will be of relevance in most situations for labs.

The first is what was described as ‘desk-bound perfectionism’. From the perspective of DesignGov, there seemed to be a cultural bias within the public service within Australia towards officers concentrating on getting something ‘right’ before sharing it or discussing it. They are bound to their desk until something is perfected. A design approach would suggest that for many issues and problems that the public sector deals with that there is no ‘right’ answer, only a range of possible options, the benefits and feasibility of which can only be assessed by involving others.

This is an understandable bias in many ways – for instance professionalism is associated with knowledge and expertise. Competence and ability is associated with being able to fix problems. Admitting that something is incomplete or unfinished can make people feel vulnerable.

But if the lab is working in an area where something is new, part of a dynamic and complex system, where there is shared responsibility for implementing or enacting any proposals or new initiatives, then an approach of desk-bound perfectionism is unlikely to be satisfactory. The process will likely only have legitimacy and engagement if others are given the opportunity to be involved early on.

A second area that became clear was a strong, almost visceral, cultural aversion to failure. This is an important consideration for a lab and its oversight. If a lab wishes to attract established or promising talent within the public service, then there needs to be strong and articulated support that no matter what happens, the experience will be valued and supported. Alternatively, those with relevant skills and valuable input may be put off by the fear of career harm through being linked with something inherently risky.

Again, there is a tension or balance that needs to be considered. If the lab is seen as safe no matter the consequences, it may attract the talented but with an ill-matched risk appetite for the work of the lab. Alternatively, if that safety is not there, the lab may attract those with higher risk appetites or those on the edge of the system/those who do not feel in tune with their home agencies/existing workplaces. This can be a strength, as it gives access to those who might have a stronger preference and aptitude for innovation, but it can also make it harder to translate the work of the lab into the perspectives of more traditional areas of the public service.

A third area is that some organisations and people, professionals who are invested in their work and what they do, may take ideas and opportunities as implied criticisms and threats. This may have to do with the presentation of the ideas. Or it may be that unless the recipients of the idea are brought along on the ‘journey’ and understand the full story, they may not appreciate how the status quo is deficient, and will regard what the lab may regard as a good idea or opportunity, as a criticism.

Diversity of thinking styles

There is nothing like working across different organisations and sectors to demonstrate how differently agencies and people of different academic disciplines see and understand the world. It also helps show how different types of organisations prize or elevate different types of thinking over others.

For instance, within the department responsible for industry policy, the Department of Industry, there is an implicit emphasis on understanding the world from an economic perspective, and with it, a bias towards quantitative data and statistical evidence. Other agencies, for instance those dealing more heavily with service delivery and interactions with citizens, often have a stronger resonance with other paradigms and other forms of evidence.

This is not really surprising, but the degree to which it can introduce difficulties in discussing or exploring issues across agencies perhaps is. While the same language will often be used, the meaning or the intent behind it can often vary. Being in a lab that has a more outside-in perspective, or a perspective from the edge, helps reveal the extent of this mutual misunderstanding.

If this tension is then overlaid with other varied thinking style preferences that can be embedded in organisational processes, reflecting traditions, leadership styles and decision making biases, the differences in how people from different organisations understand things becomes more apparent. This matters more for a public sector innovation lab than it does for many other inter-agency collaborative efforts because the primacy or dominance of a particular set of preferences is less likely to be clear. In many cross-agency activities or whole-of-government initiatives, there is usually a clear sense of who or which agency is in charge or has the political leadership of the matter. In an exploratory and experimental setting this primacy is not likely to be settled or clear to all players. Therefore the usual accompanying acceptance or submission to a particular set of thinking style preferences may not preside.

For instance, drawing on the experience of DesignGov, some of the different styles observed included either:

    • Narrative or factual approaches (with an accompanying respective bias towards anecdotal/qualitative evidence or quantitative evidence)
    • Visual or text based preferences
    • Linear processes or approaches favouring complexity
    • Identifying an objective truth or accepting multiple truths
    • Bias for action or bias for certainty.

These are somewhat crude differentiations, and often there is a spectrum, or different preferences might operate in different settings, however the aim is to illustrate the varieties that exist.

In dealing with innovation, and the accompanying pressures and inevitable stresses that seem to accompany it, these preferences and styles are important to consider, particularly as people often seem to default more strongly to their own approach when feeling challenged.

An awareness of these different preferences and styles can help those involved in a lab understand the reactions and the dynamics of inter-agency/inter-group collaboration, even if it might not help predict or mitigate them.

Being a public sector entrepreneur

A degree of entrepreneurialism can be very helpful with an innovation lab. If a lab is undertaking innovative activity, then there is a good chance that things won’t go exactly according to plan. This suggests two implications – the need for a lab to identify and respond quickly to emerging opportunities, and being able to ensure multiple objectives might be aided by any one activity.

Firstly, a lab, being in an unusual position of cutting across perspectives and networks, will likely be in a good position to identify opportunities that emerge at short notice. A lab needs to consider how it can take advantage of these (in a way that furthers the work and aims of the lab), and to ensure that it cultivates multiple options, in case planned activities do not proceed as expected.

This is particularly relevant for a lab as no matter how organised or prepared some things are, an exploratory process will be an evolutionary one. Different options will present themselves as will unexpected roadblocks or delays. Partners and collaborators will have priorities that will change, they will have unexpected delays or processes with which they have to abide that are outside of the control or influence of the lab. Individuals may move jobs meaning that relationships have to be recreated. Other times circumstances will conspire to make certain things easier and quicker than predicted.

A coping strategy might be to ensure multiple but complementary streams of activity that can be switched between or prioritised as needed. In an eco-system it can be very hard to identify the critical dependencies, so resilience is aided by developing a range of options.

Secondly, on a related note, a lab needs to consider how it can make sure that any one activity might meet several objectives at once. The higher the level of uncertainty that a lab operates in, the less likely it will be able to predict which projects or activities will deliver what is wanted. Therefore any activity should aim to potentially meet several goals, so that if it does not deliver on one aspect, it will at least likely deliver on a different one.

Tolerance for ambiguity and uncertainty

Are the team members of the lab comfortable with ambiguity and uncertainty?

Innovation is inherently about change that is uncertain. Design involves nuance and allowing for multiple interpretations. Either of these elements can be challenging for some personal preferences/styles. An effective innovation capability requires team members to take on different roles and different tasks outside of their normal experience and comfort zone.

Don’t forget to laugh

A public sector innovation lab can be the environment for an exciting and stimulating experience. It can also be challenging and demanding, involve high-pressure situations and timeframes, and bring the team into contact with tensions and existing conflicting or cross-purpose processes and systems. A lab should aim to balance the two sides, as over-investment is a real risk.


The aim of this paper has been to provide a practical understanding of the experience and considerations involved in establishing, running, and closing an innovation lab. It reflects the experience of only one public sector innovation lab however, so should not be taken as necessarily representative.

This material is produced under a Creative Commons By Attribution licence, with the hope that others involved with public sector innovation labs can continue to build on the knowledge base and improve the practical guidance available to those establishing a lab over time.