My Process

Integrating multiple sources of evidence to design for positive outcomes.

The social challenges that are most worth solving are those that have the greatest potential for unintended negative consequences. I believe designers have an obligation to ground our work in empirical evidence and collaborate with beneficiaries. We can go beyond creating designs that “should” work to evaluating how a design performs in the messy real world and ensure our work produces positive outcomes, not negative consequences.

Research: What evidence can inform the design so it is most likely to work as intended?

Concept: What form would best address needs identified in the research?

Implement: In the real world, are people able and willing to use the design as intended?

Evaluate: Does the design result in the intended outcomes without causing harm?


1. Start with outcomes.

If the design worked perfectly, what in the real world would change? This can be sweeping or targeted, ranging from “America’s per-capita carbon emissions does not exceed two tons of CO2 per year” to “The school cafeteria wastes 20% less food than it did last semester”. The outcome should be measurable and specific in order to evaluate whether the design performed as intended.


  • Optimism
  • Desire to create a more just world

2. What behaviors support or thwart this outcome?

Use best available evidence to identify behaviors to target in the intervention. This includes published empirical evidence, input from beneficiaries, and qualitative research. What might we help people start doing? What might we help people stop doing?


  • Empirical literature review
  • Expert interviews
  • Qualitative research
  • Target behavior identification

3. Who are you trying to support with the intervention?

One-size-fits-all design rarely addresses social and environmental barriers to behavior change. Select a target audience for the intervention based on potential for positive impact and feasibility of implementation. The tools you have access to, e.g., dedicated staff, digital platforms, print resources, often inform who you are able to reach.


  • Empirical literature review
  • Expert interviews
  • Qualitative research

4. What gets in the way?

Identify barriers and facilitators for behavior change for your target audience. These barriers and facilitators can be at the individual, social, and environmental level and can be as large as systemic inequality and as granular as a form that is difficult to understand. Barriers at every level are relevant for understanding what intervention could support behavior change.


  • Empirical literature review
  • COM-B analysis
  • Interviews
  • Observational research
  • Co-design workshops

5. Make something that addresses these key barriers.

Interventions can consist of making something new, revising existing systems, or removing complexity at the individual, social, or environmental level. Many effective interventions are not flashy. That’s great! Creating the most impact with the fewest resources is commendable. What matters is that it works. Co-design and testing with beneficiaries can help ensure the intervention successfully addresses actual barriers, not assumed ones.


  • Theory of change
  • Co-design
  • Concept design
  • Desirability testing
  • Prototyping

6. Pilot and revise the intervention.

Pilot and evaluate the prototype. Do people use the intervention as intended? If yes, does it produce desired behavior change? Do those behaviors result in the desired short and long-term outcomes? If not, revise as needed.


  • Evaluation and measurement plan
  • Usability testing (qualitative and quantitative data to determine if people are able to use the intervention as intended)
  • Efficacy testing (Pre-post trial or randomized controlled trial to determine whether the intervention produced meaningful outcomes)