Project Everest

Work Update

Let's get more data

William Lee
William Lee | Dec 19, 2017 | in Knowledge Base

Why: understanding market price and its volatility/seasonality is an important issue in any market. 

Problem: currently, we collect data in time consuming, labour intensive ways - usually in the form of interviews, and thus fail to get accurate data nor enough sample size to perform meaningful data analysis.

Idea: getting market sellers to text through price of *insert product* on a weekly basis to a phone number managed by PE.

How to: 
1. Recruit a few market sellers in both major and smaller village markets, who have access to mobile phones. They will be compensated for their involvement by PE 
- By recruiting multiple sellers in each market, we can get a "market price" for each product by removing outliers and averaging the other pricing data. 

2. Send weekly automated texts to our market sellers (to acts as a reminder), asking for prices of the crops they sell - and also maybe the crops they can buy at the market (at the end of the day, they are consumers as well).

3. This data is collated on our end - each mobile phone number will act as a seller ID - and can plot out market price movements over time and across geographies.

- Very low involvement on our end - no need to have to go out to each market, each week, in person 
- In fact, once set up, it can be done completely remotely - even back in Aus potentially
- This data will provide inputs to financial modelling
- It may also provide a future platform to make market prices transparent for consumers
- Depending on the extent and accuracy of our data, the data could also be shared/sold to other organisations in the field and empower their projects. This will help connect PE with other powerful organisations, and provide us with something to bring to the table during discussions. 

1) to what extent do our current data collection methods incorporate the above idea?
2) is it feasible to enact this in all the countries, or are there any major barriers?
3) any improvements or suggestions?

Let me know your thoughts.

edited on 19th December 2017, 06:12 by William Lee

Andrew Vild Dec 20, 2017

1. I don't believe we have any standards across countries with respect to collecting data. A frustration is that a lot of teams feel the need to learn mistakes their own way by making a survey and then iterating on it. The main issue here is that an iterated survey essentially makes the initial few surveys invalid and also wastes information opportunities from the members of community we have interviewed, as we can't go back next day with same but slightly different questions.

2. There shouldn't be any major barriers across countries within same projects. You may just need to have a few questions included in a SOP that will have the option for N/A.

3. Definitely, when I have more headspace and am not in the last week of project! Discussion to be continued over beers in Egypt!

Users tagged:

Reply 0

Kate O'Donnell Feb 18, 2018

Interesting point, is it viable for PE to be compensating for such a large volume of participants? I wonder whether respondents would reliably respond to the texts. I do like that you have the ability to get recurring answers without having to actually do the legwork!

Reply 0

Mallory Dobner Jul 1, 2018

Status label added: Work Update

Reply 0