At Nesta we work towards three mission goals using an array of innovation methods. It’s stimulating, fun and hard work. Not least because a good mixed methods team is one that productively disagrees. Since agreement isn’t the aim of the game, this blog is not a house view on how we do innovation. Rather it is a personal reflection on some of the things I’ve learned about how to do this kind of work well.

My first suggestion is to imitate wherever possible. While we might not always be able to find examples of how to work through specific issues in mixed methods teams, there is almost always someone who has cracked an archetypally similar problem in a different domain.

With that in mind, I’ve identified five mindsets I find it helpful to adopt at different stages of an innovation project.

Most big problems are complex. They are generated and perpetuated by warping economic forces, poorly designed systems and how we behave as a result. In other words it’s very rare to find a problem explained by one driver at one level of a system. This presents a problem: complexity overwhelms us and progress stalls as we fail to make choices about what we will (and won’t) focus on.

This is where thinking like an economist helps. Economists are great at quickly cutting to the core of things. This is partly because they tend to think about outcomes in terms of their production function. That is: of the many factors that produce the problem, which wield the most influence? Approaching our initial analysis this way helps give us a decent picture of the factors we should be focused on. The resolution of this picture does not need to be perfect. In fact, the point is to know when to stop.

Imagine you were given a task to figure out what the first blurred image below was. In the first two pictures it’s impossible to say what the blurred shape is. But by the time you dial up the resolution to the level in the third picture it’s clear that you’re looking at a puppy. Do you want to keep increasing clarity to see the puppy in all its glory? Of course you do! But that wasn’t the task: you have enough information already to draw your conclusion and move on. You must resist the puppy.

This approach of going deep enough but no deeper has been instrumental in helping us set three-year strategies to guide our missions. For example, in our fairer start mission we have a goal to close the attainment gap at age five between children born into deprivation and their peers. The determinants of a child’s development are complex and interact in ways that compound. By taking a pragmatic approach we have been able to identify a handful of key drivers as areas of focus for our work, namely: addressing material needs by giving poorer families more money; improving the effectiveness of parents by finding ways to scale evidence-based parenting support; building better connective tissue in public services by linking up early-years data; and bringing all these insights together to improve national policy in all four UK nations. As we dive into these areas we will dial up the resolution: at any given stage the goal is to keep things as simple as possible.

When you play football you get taught to watch the ball not the feet. The feet and the rest of the body can give hints as to what a player might do but ultimately what matters is where the ball goes. For all three of our missions, behaviour is the ball and we need to think like a footballer to ensure we stay laser focused on it. That is, the outcomes we are chasing are all products of the actions of individuals, be it investing more in child development, eating a healthier diet, or installing low-carbon heating (as mentioned above: a good team disagrees – my wonderful colleague Professor Jenny Gibson pointed out here that I am missing the importance of relationships by only focusing on behaviour).

This doesn’t mean that other factors don’t matter. Quite the opposite: behaviour is a product of everything from economic incentives to cultural norms. By paying attention to behaviours we can figure out what drives them (why they are happening) and, therefore, what would need to change to alter the behaviour. Similarly just because we look at individual behaviour as an outcome doesn’t mean our interventions have to be individualistic (that is targeted at changing the conscious choices of the individuals whose behaviour we are interested in altering).

Our healthy life mission exemplifies this. The outcome is a change in what people eat but almost none of the main mechanisms of change we’d push for are targeted at changing the deliberate choices individuals make. One example of how we have deployed this approach is in our work on mapping the school meals system in Wales. The purpose of this work is to describe how school meals are delivered, giving context on the forces shaping outcomes at various stages. There are many different approaches to producing descriptive system maps like this but by centering the actors involved in the system, be those institutions or individuals, we are able to look at how specific actions (behaviours) contribute to the end result of what ends up in kids’ mouths. This makes it easier to identify the most critical leverage points, that is the actions that matter most and the factors that enable or block them, and design interventions that address them.

In general it’s very easy to underestimate the opportunity cost of doing nothing. This can include doing more thinking or not-having-a-go for fear of failure. Obviously rushing into something with no plan is bad, but in most organisations that value good research skills, it would take a very large overcorrection to end up being too action-oriented.

This is where it helps to think like a playwright. In writing for the stage there is a general rule that you shouldn’t include any scenes that don’t meaningfully develop the plot in some way. The same goes for project activity. Meetings, research tasks, and other activities should be treated in the same way: if they’re not helping you figure out how to move forward then stop and reconsider. We find the designers’ toolkit is especially good at facilitating this mindset because it can help us to identify critical uncertainties early and find practical ways to test them quickly. In our sustainable future mission, we have developed an approach called speed testing, which brings together interdisciplinary teams for a short period of time around a highly focused challenge. This means we can rapidly test and validate concepts. If we find something that shows promise it gets taken forward. If we don’t then we “cut the scene” since it doesn’t enhance the play. Our Visit a Heat Pump service came out of speed testing, as did our interests in financial support for new heat pump installers and group purchasing of heat pumps.

Trying things out in this way generates momentum but it doesn’t have to generate path dependency. If all prototyping does is accelerate a decision that one pathway is not the right one, that's still more useful than building an increasingly high-resolution picture of the problem. In that same round of speed testing we tried and rejected ideas on interim boilers for broken heating and heat pumps as an employee benefit.

To help our ideas mature, we want to increase the resolution of our understanding about how they work. This means testing and refining based on what we learn. To revisit an earlier metaphor: this is where we finally get to see the puppy.

One of the reasons Formula 1 racing is so compelling is the sheer diversity of skills it brings together. Elite drivers are supported by world-class mechanics, automotive engineers make small tweaks each season to shave seconds off a lap and event managers put on a safe and extravagant spectacle time and again in different locations across the world. In our work, we want to think like Formula 1 by understanding how the different types of expertise we have available fit together and who should take the lead at the right time.

Of course, we are dealing with social challenges rather than racing cars. The first kind of expertise we have to draw on is the existing evidence base. All three of the areas we work in have vast amounts of research behind them, which we can benefit from simply by doing our reading. There is no point reestablishing a well-founded insight or trying something others have already shown does not work. But there are many kinds of expertise beyond the ivory tower. In our fairer start work, for example, child development specialists can tell us what kinds of activities are beneficial for children in their earliest years, data scientists can use geospatial mapping to show where services should be targeted,and the parents those services are designed for can tell us why they find most parenting advice too patronising to hear. The trick is to cede space to the right kind of expert in the moment so that we can collectively build something better together.

Similarly, we believe that no one has the monopoly on good ideas. In fact sometimes deep subject matter expertise can hinder our ability to think of new ideas because we have too much familiarity with the status quo. Rather than coming up with ideas behind closed doors we want to open it up to make sure we’re not missing brilliant insights. The designer’s bag of tricks really comes into its own when it comes to getting ideas out of people. For example, getting people to think with their hands or testing prototypes to go beyond the status quo.

We also recognise that many good ideas will already have been invented as hacks to the existing system by savvy practitioners or users. We see it as part of our job to identify, test and scale good ideas that might already live on the fringes of a problem. For example, our Money Saving Boiler Challenge campaign, which reduced emissions by half a million tonnes a year, was in part inspired by boiler installers and other experts who were already hacking high heating bills for their customers by adjusting flow temperatures.

We are an innovation charity. The word innovation tends to conjure ideas of novelty, invention and counterculture. But often the kind of innovation we are looking for is much less exciting. What we want to know is why our work should succeed at scale when others have tried and failed. This means that innovation is often as much about getting something to stick and scale as it is the initial invention.

This means thinking like a startup. Startups succeed when founders are obsessive about scale from the beginning. When we initiate projects we try to assess them based on their ability to have impact at scale. It also means that we think about scaling as an activity in its own right: who do we need to embrace this? How do we persuade them that they should? Will the secret sauce of what we’re testing survive dilution if the implementation isn’t perfect? How many people would our work need to reach for it to meaningfully move us towards our mission goals?

In our fairer start work with local authorities we are continually trying to strike a balance between developing insights and interventions that address the specific needs of our partners but that could also scale nationally if they proved successful.

These five mindsets are not exhaustive but they help guide my work and get me unstuck. That said, I make no claim to have nailed all or even any of these. As ever, it’s a work in progress. But practice makes perfect and each time a new project comes up it feels a little easier to make progress than the last.