projecteverest

[Proposed Experiment]: SoCon, Malawi, New Customer Risk Mitigation Experiment

Lean Phase: Customer Segment

Assumption: With the defaulting rate of customers paying back on solar products on a repayment plan being at an alarmingly high 49%, we are testing the assumption that greater risk profiling at POS can mitigate this risk. To do this we are rolling out a risk profiling system which begins with a survey at the POS of repayment plans for solar products. Each answer of this survey adds or subtracts from that person’s ‘creditworthy’ total. If, after all the questions, the total is positive, the person is given a solar product on a repayment plan. In any other case, they are denied.

Time Box: 2 weeks (or time taken to use risk profile 30 new customers)

This can be done in conjunction with defaulting customers risk mitigation experiment 1.

Success Metrics:
1. % of risk profiled customers who pass the threshold who still default.

Criteria:
Success Point: Defaulting rate drops to below 25% of the risk profiled customers

            Green Light: Continue to use and iterate on the risk profiling tool.  
Orange Light Point: Defaulting rate drops to below 35% in the risk profiled customers
             Orange Light: re-evaluate risk profiling tool to see if all information surveyed is available, accessible and relevant and attempt to improve the tool.
Failure Point: Defaulting rate changes is still above 45% in the risk profiled customers.  
             Red Light: evaluate assumption that risk profiling, while important, isn’t the major cause of defaulting rates in our solar customers and assess what other underlying causes could be.

Experiment Build:
1. Employ this survey on 30 new customers buying solar lights on a payment plan.
2. Use the negative or positive weighting assigned to each response to assign an overall risk value to each customer.

3. Trial giving the asset loan based off whether our profiling tool states the customer passes our risk threshold and is hence predicted to pay back their loan. 

4. Track the customers taken onboard with this method and whether the default rate among these customers is lower than 49%.
5. Analyse and publish results.

The questions used are shown below:

  1. Who are they?

·       Name?

·       Phone number

·       Location/Village

·       What loan plan are you interested in?

 

2.              Can they repay?

·       Occupation

·       When do you get paid?

·       Is your income impacted by seasons?

·       What is your monthly income?

·       Weekly expenses related to product

·       Do you have access to Mobile Money

 

3.              Will they repay?

·       Do you have kids who are school age?

·       Have you ever had a loan before?

·       Who was the loan with?

·       Did you repay it?

·       Do you have any loans that are outstanding right now?

·       How much is your outstanding loan for?

edited on 11th January 2019, 08:01 by Rafael Branton

Fiona Aaron 5 months ago

Wow that is very high. Good idea to implement some risk profiling. Just confirming though, this experiment is tied to this one - https://projecteverest.crowdicity.com/post/828730?
If the original experiment on risk profiling proves to be ineffective in determining risky customers, it probably wouldn't work here right? Or am I missing something?

Also, do you have any hypothesis as to why people are defaulting? How do competitors in this space mitigate this risk/deal with defaults?

Reply 0

Share