Scale up is a deliberate effort 'to increase the impact of innovations successfully tested in pilot or experimental projects to benefit more people'.

Scale up is often challenging because interventions that work in one context may not work in another. There are several reasons why interventions can fail to work when scaled:

• External validity and context-bounding: There may be a mismatch between the context in which an intervention is designed or tested, and where it is needed to be implemented.

• Diminishing returns: The intervention's efficiency reduces over time - evermore resources are required to convert people, for example due to to logistics, or just just to less of the audience being highly receptive.

• Loss of fidelity: As the intervention is scaled it starts to lose some of the qualities that made it effective.

• Mobilising commitment to scale: Research commissioners may take a 'wait and see' approach to trials, which can lead to insufficient investment in setting an innovation up for success at scale.

• Backfire effects: What encourages positive behaviour in one setting reduces positive behaviour in a different setting.

Watch the video to find out more.

The scale up process fits with behaviour change or applied behaviour science approaches, and scale up activities start earlier than you might think.

While there are many models and approaches to behaviour change, most move through several steps: understanding the context and identifying behaviours to change; designing an intervention or solution; and testing and implementing that intervention.

The model for scale up that we have developed from our research in the area also includes several stages. The names we have chosen for these stages are planning, piloting and implementing.

Planning stage

Planning refers to the process of iteratively creating, communicating, and evaluating your ideas.

Though each might use a different label, most behaviour change approaches involve an extended process of identifying, discussing and formalising a plan for creating behaviour change.

Some of the key scale up activities involved at this stage include ensuring that the team has the relevant skills and interest to tackle the problem using a behavioural approach, identifying potential interventions to develop or adapt, prioritising these interventions based on feasibility and scalability and adapting existing interventions to new contexts.

Piloting stage

Piloting refers to the process of iteratively collecting external data to test and evaluate your ideas before you allocate significant resources to implement them.

Many behaviour change approaches mention the importance of quickly testing interventions with a small population to evaluate processes and outcomes.

Some of the key activities involved at this stage include testing assumptions about the intervention, for instance, about the effect, the response of the target audience, the performance of trial delivery partners. Gathering this information is important to identifywhat may need to be changed or adapted about the intervention before it will be successful at scale.

Implementing stage

Implementing refers to the process of releasing the ideas at scale and evaluating and monitoring their fidelity and impact.

Most behaviour change approaches have a stage where an intervention - usually but not always tested for effectiveness - is put into the field, and from that point on managed by a user group, government partner, or organisation.

For example, this could involve taking an implementation that you have tested and validated with a small group of your target audience and partnering with other authorities to make it state or nation-wide.

Notes

These stages don’t always happen in order.

Planning and piloting can often occur in parallel, where both are worked on at once; for example, where some parts of a program are tested while others are being planned.

Planning and piloting can also be alternated, as different stages of a repeat cycle, where you alternate between them.

In contrast, once you move to the implementation stage, you generally don’t go back because you have invested all your time and resources in funding and planning how you are doing to broadcast your innovation.

Successful scale up requires effectively planning, piloting and implementing a potential intervention.

If you have picked the wrong behaviour, audience or intervention, or made a critical error in pricing the cost of your intervention, there will often be nothing that you can do to fix this when you are implementing the intervention.

The most critical factor in a scale up project is ensuring that you get the planning and piloting correct, before you start implementation.

Our toolkit is focused on the early stages of scale up because the planning and piloting stages are when 'uncertainty' and 'flexibility' are highest and 'risk' and 'resource allocation' are lowest.

1. Uncertainty refers to the things that you don't know (some which you don't even know you don't know). For example, a 'known unknown' that you know you need to figure out could be the differences between your old and new audience. The incompatibility between your software and your partners may be an 'unknown unknown', something you don't know you don't know but that will ruin the project unless you identify it early.

2. Flexibility refers to the ability to easily change things. At the planning stage you can change the entire scope of the project at little to no cost. However, this flexibility reduces significantly over time as you move closer to implementation and involve more people and resources. It becomes ever harder to change goals, delay timelines, or the capabilities required.

3. Risk refers to the probability and impact of failure. At the start, risk is low because you haven't invested much time, money or social capital. As the project progresses there is evermore at stake as the time and effort invested increases.

4. Resource allocation refers to the amount of resource invested in the project; relatively little at the start then often vastly more by the implementation phase.