Project Everest

Adopted Experiment

[Experiment Adopted]: SoCon Fiji - Loan Channel: Product Loan through local hardware store - December 2018

Lean phase: channels


This experiment is to test the assumption that providing product loans (tokens) for equipment will give us relevant and accurate data on the loan behaviours of our customer segment of farmers.


This experiment should be conducted over the space of 2 months to ensure we can monitor repayment over this period.

Success metric (measurements)

  • Adoption rate (no. of products taken out on loan at current interest rate)
  • Default rate
  • Number of risk factors present

Green light: If this experiment proves that we can obtain relevant and consistent data from this channel, we will continue to undertake the experiment and increase the number of products that loans can be given for.

Success point: If 4 people per week take out a loan through this channel then we will be able to draw enough data for it to be a worthwhile experiment. This is assuming that this experiment will run for the period of 8 weeks (2 months of January and February) and n=30 needs to be reached to achieve statistical significance.

Orange light: If successful, we should seek out new hardware stores to expand our reach to further test multiple loan structures.

Failure point: If less than 4 people per week take out a loan for a product then we will cease using this channel as it will not provide us with any data that is statistically significant (n > 30).

Red light: If this channel fails to be useful we will explore new channels by which to test for loan provision such as through village heads.


  1. Organise meetings with local hardware store (S.K. Davey) and determine products that constitute good debt

  2. Customer fills out the application form for us to assess their risk (adoption rate datapoint)

  3. Successful applicants sign the contract for a loan and we provide the finances

  4. Text reminders are sent for them to repay the loan weekly

  5. Transfers are made to us via mPaisa

  6. Final payment is collected (default rate datapoint)

  7. Survey to be completed after the process (risk factor datapoint)

edited on 3rd December 2018, 05:12 by Alex Staples

Lucy Noble Jul 30, 2018

Status label added: Proposed Experiment

Reply 0

Alex Staples Dec 3, 2018

Status label added: Experiment adopted

Status label removed: Proposed Experiment

Reply 0