Translating the Lean Language to the Social Sector

By Therese Lim

Recently, I’ve been working with friends in the international development sector to come up with a technological solution that helps bank the unbanked. Most people working to solve this are public policy analysts and academics who haven’t been exposed to lean methods. It’s one thing to have Intuit’s lean-trained Global Business team create AgriNova; it’s another to apply lean when working with people who come from a more comprehensive product development tradition. It has been an interesting personal challenge to translate our class learnings to induce effective collaboration with them.

The social sector doesn’t use lean often for a variety of reasons. First, there is a sense that “social mission equals mission critical.” If we don’t get it right, we’re reducing our impact and possibly destroying lives. However, in my experience this is more of a psychological fear due the high cost of failure in the past when tech-enabled iterative methods were not used, and is usually driven by the threat of losing funding and displeasing donors.

Second, there hasn’t been a direct incentive to be speedy and efficient. Long development cycles are aligned with long funding cycles.

Third, feedback loops are not as rapid or robust as they would be in a typical startup. It’s not as simple as throwing up a smoke test on a website and seeing what happens. One of my friends, for instance, is working on a mobile-based banking solution among rural farmers in the Philippines. Just to test consumer demand, she had to mobilize a lot of community workers to find, educate and monitor users over large geographic distances and shoddy infrastructure.

However, I was encouraged to find that many of the methods these social scientists adopt are in the lean spirit:

1.       Customer feedback: The feedback they collect is statistically rigorous and both qualitative and quantitative. We marveled at Intuit’s Follow Me Home approach to customer observation; the social sector has been doing this for ages through fieldwork and is very aware that stated preferences are not as useful as revealed ones.

2.       Measurement: The sector is very metrics-driven and adopts Randomized Controlled Trials (RCTs) as a gold standard for evaluating solutions to social problems.

3.       Pilots: Essentially the equivalent of an MVP, they roll out programs in small areas first and slowly expand from there. 

I recognize that many of the tactics they use are too comprehensive and arguably no longer lean. RCTs take forever to implement, for instance. So I’ve been trying to streamline the process by (1) destigmatizing mistakes, (2) helping specify true MVPs that are operationally constrained, (3) using tech to lower operational barriers to testing, and (4) emphasizing the need for falsifiable hypotheses and metrics thresholds as opposed to the more common “any usage is better than no usage” approach. I’m using the lean philosophy, but in their language, to be more persuasive.

For the class: I’d love to hear your thoughts on how you’ve been able to integrate lean in atypical product development settings.