Former USAID chief innovation officer visits Pact to discuss social sector impact

January 25, 2019
Ann Mei Chang speaks Jan. 22, 2019, at Pact.

Pact recently welcomed former USAID chief innovation officer Ann Mei Chang to discuss innovation, maximizing impact in the social sector and her new book, Lean Impact. 

Watch a full recording of the Facebook Live event here. Below are key takeaways from Chang's talk. 

In your transition from the private sector to USAID, what were your most eye-opening moments? What are the key differences between innovation in a corporate setting and innovation for social impact?

For one, the ease of running small experiments. In the Silicon Valley tech setting, it took seconds to run small experiments on websites, making small tweaks to drive user-interaction. In the social sector, we don't have that enabling environment that facilitates quick tests. Additionally, the funding structure and the lack of incentives slows down innovation processes, creating an operating environment that challenges rapid iteration.

How do we fall in love with a problem and not the solution, as you describe in your book?

You identify your North Star and you don't give up until you reach it. You ensure that you are crafting metrics, the right metrics, that guide you in the direction of that North Star. If your metrics don't facilitate that path and you end up with "vanity metrics," then you celebrate your solution, not asking the tough questions about your problem. 

How do you manage ethical considerations when running experiments where people's health or livelihoods are involved?

First off, you start with small experiments - think big, act small. By starting with an experiment of only five or 10 people, you have the agility to iterate quickly when something isn't working. Only when progress is obvious do you open your experiment to a larger population. In fact, arguably the way that current large interventions are designed could be considered unethical when a proposal valued at $50 million that includes thousands or millions of beneficiaries fails. Would it not have been better to start small and gradually increase the recipient populations? Additionally, what is the counterfactual? What is the risk of not running this experiment? Of not looking for new, promising solutions to a development problem?

As an organization, how do we know if we are transforming our mindsets, our culture?

You can ask yourselves these key questions: Are you allowed to fail? Do you see failure across the organization? Is there not just permission to innovate but incentives to innovate?