Selection: get it right or prepare to fail

So often we don't think about important factors like trust when implementing new products.
But these oversights can make or break us.
Photo: pixabay.com

Shortly after I started working in microfinance, I had a meeting with two researchers and a colleague who began working at BRAC in 1984.  The researchers had a lot of questions about the repayment rate--how was it possible we managed a repayment rate of over 95%, given that our clients were mainly poor women without collateral, in a country with no credit bureau, etc?

My colleague said, "Ultimately, the most important factor is selection.  We spend a lot of time making sure that we only give loans to people who we think can and will pay them back.  If you get selection right, then everything else is relatively easy.  If you get it wrong, no system is good enough to save you."

Over the last year and a half, his words have come back to me over and over.  It was a simple statement, but advice that so many initiatives could benefit from.  Selection rarely gets a lot of time and attention from anyone (unless we're talking about lunch, then it's a really important decision).  Usually we're in a hurry to get started, and anything that stands between us and launching a new product is seen as a delay.  

There are many important selection questions that get overlooked, like:

• Who is the target population?  Are we sure that this product will benefit them?  How will we make sure that they are aware of the benefits, terms, and conditions?
• Who will run this project?  Is this person passionate about it, technically sound, and supported to succeed?
• Who will our partner(s) be, if any?  Do we share a common operating philosophy? Are we on the same page about what success means?

Looking at what I've seen work and fail, often it comes down to basic breakdowns in communication, priorities, and fundamentally, lack of trust.  Ultimately, in situations where trust exists, things may go wrong, but usually they are fixable.  Without trust, things get complicated quickly, require expensive monitoring systems, and are hard to make effective.  And it's a two-way street: trust has to go both ways to be meaningful.

Recent randomized control trials on microcredit (loans only, not savings) have found that its impact is generally positive, but not transformative.  This seems obvious--for some households, access to capital is what prevents them from improving their lives.  But not for all. Figuring out how to best identify those households that are able to benefit economically from loans is an obvious prerequisite for success.  And many more households can benefit from "microfinance-plus", a combination of important financial services like savings and loans, with wraparound services like health care, clean water, education, etc.

This is true far beyond microfinance.  A few months ago I wrote about how BRAC's early work to prevent death from diarrheal disease was ineffective because its educators didn't really believe in what they were teaching.  Just last week the New York Times published an article on "A failed trial in Africa raises questions about how to test H.I.V drugs", describing some of the troubling findings, including that many patients inaccurately reported their adherence and that many hid the fact that they didn't take their medications.

The author centers on the question of whether paying the research participants contributed to the false reporting, and whether it's ethical to pay them at all for participation.  Recipients were paid $10-15 a week for participation, an amount that was meant to at least cover their transportation expenses to and from the clinic (as a point of comparison, I used to get $25 for participating in a 1-2 hour psych experiment in college, and that didn't involve taking any medications).  The $94 million research project involved over 5,000 woman in South Africa, Zimbabwe, and Uganda.

In my mind, the article misses the really obvious point: that the respondents had a deep mistrust in the researchers and by extension, the medication that they were provided. Follow-up studies mentioned in the NYT article confirm my theory:  
"Many admitted to discarding pills or pouring gel down the toilet. One gave hers to a friend working as a prostitute. Another stockpiled them for later, to take if they were found to work....Clinic waiting rooms were often abuzz with rumors — that the drugs would “rot the uterus” or cause liver cancer or infertility, that the blood samples were sold or used for Satanism, that the foreign scientists conducting the study were secretly spreading H.I.V., that a woman could tell from nausea or tingling whether she was getting the placebo."

Researchers now are looking into blood tests to make sure that people are taking the medications properly.  It's good to have additional monitoring, but unless they can do a better job choosing women who have more faith in medical research and build a trusting relationship, it's unlikely to solve the underlying problem.