Project Everest

[Proposed Experiment]: SoCon Malawi - Defaulting Consumer Experiment 1.0

Isaac Penglis
Isaac Penglis | Dec 20, 2018 | in Knowledge Base


Lean Phase: Channels

Assumption: The current default rate from Solar product based loans is 49%. The intention of this experiment is to validate that the price points are not the primary reason for defaulting consumers. The assumptions that SoCon will validate throughout this experiment are as follows:

  1. Energy costs are greater than product repayment costs

  2. Consumer's average income p/week is >6000 MWK


Time Box: 4 weeks (or time taken to survey 30 customers)


Success Metric:

This experiment is designed to measure reasons for the high default rate of customers on the Solar repayment plan. This should be conducted through interviewing 30 customers who have purchased the Solar product through a repayment plan.

The experiment is to determine if cost isn’t an issue for the majority of people who have defaulted - i.e. 50%.

Whether a customer can afford the repayments will be determined by:

  1.  Average energy cost per week

If weekly repayment rate is less than the average energy cost per week, then the customer can afford the repayments

  1. Average income per week

Or average income p/week is over 6,000. If energy costs make up 30% of weekly income then the customer can afford the repayments.

If someone spends more than 1800 MWK per week on energy and has an average income of 6000 then it is safe to assume that they can afford the repayments. If this is validated for the majority of cases then the next step will be to conduct a new experiment attempting to answer; 'why consumers are not repaying the loan considering they can afford it?'. This will likely come down to three points (1) They're unwilling to pay it back (2) they cant be bothered or (3) its too hard.



  • Creating a survey that can accurately determine if a customer has poor spending prioritisation
  • Accessing 30 different defaulted customers within Nancholi without a geographical location.


Experiment Build:

  1. Create survey validating assumptions (draft survey)

    1. Validation of defaulting customer

      1. Do you own a GD light product?

      2. Could we get your phone number? → Cross check in CRM to check balance

      3. Are you part of a repayment plan?

      4. How often do you repay?

      5. Are you aware that you have [amount] on your balance?

      6. Why haven’t you been paying back this loan?

    2.  Customer can afford repayment

      1. How many candles do you purchase in a day? What about matches and paraffin?

      2. What is your occupation?

      3. How long have you been operating in this occupation?

      4. How much would you say you earn a week?

      5. What payments do you need to pay each week?


  1. Interview 30 defaulting customers

    1. Option 1: Hire a translator and call the customersOption 1: Visit the customer's house

      1. This option is much more efficient

      2. Large reliance on the translator

    2. Option 2: Visit the customer's house

      1. This would take multiple days to complete due to the accessibility

      2. The interview would allow us to pin the location of their house in case we needed to revisit them

      3. Face to face interviews are more reliable


  1. Collaborate and analyse results


Attempt to implement option 1 first as this is much more time efficient. If option 2 is an ineffective method move to option 1.


Tagged users

charlie Ford 1 month ago

I admire this article for the well-researched content and excellent wording. I got so involved in this material that I couldn’t stop reading. I am impressed Goole play Giveaway

Reply 0