Project Everest

Work Update

AgImpact: Increasing PE’s efficiency

Cec Cameron
Cec Cameron | Dec 18, 2017 | in Agriculture Assessment

To keep up with food demand, production must increase by 70% by 2050 (according to the FAO) or by 1 trillion calories by 2030 (according to GroIntelligence). The challenge in meeting these production goals is heightened when considering that the effects of urbanisation, a growing population and climate change means that this increased supply must occur on the same, or a lesser, area of land and in harsher climates. Ultimately, a second green revolution is required and will be fuelled by progressions in technology. However, to ensure that technologies being designed and implemented will be adopted and not have unintended or unforeseen consequences, market research is vital.


Sometimes, when working in Australia it can be easy to think of agricultural production purely as commercial businesses. However, smallholder farms (<5ha) produce more than 80% of global rice production, 75% of global groundnuts and oil palm and nearly 60% of global millet and cassava production. With such a significant amount of food production coming from these smallholder farmers, there is an enlarged data pool to gather research from.


This month, PE’s Malawian Ag Assessment team has experienced the difficulty that such a large farming population has on the ability to gather research within the sector. Market research is vital in understanding how a business provides value to a marketplace. But with 80% of Malawians (14.5 million people) involved in the agricultural sector, how can a team of 6 people, operating for 4-5 months a year, conduct research such that theoretical saturation occurs?


Of course, defining a sample area will reduce this looming goal. However, perhaps PE could also consider using a greater amount of technology in their data collection. Research conducted by AgImpact (contracted by Australian Centre for International Agricultural Research) has shown that the use of apps for in-field research has potential to improve relationships between researchers and smallholder farmers by improving two-way information exchange in near real time. AgImpact showed that 93% of smallholder farmers gain more value from conducting surveys on a tablet, offering them something more than a hardcopy survey, 73% of farmers found the overall survey experience using apps was positive and that survey times were reduced by approximately 53%. Researchers also found reduced errors in missed, or incorrectly entered data.


Understanding the market that a product is about to enter and the consumer that the product is going to serve is vital. Market research cannot be skipped; however, it is extremely time-consuming. Perhaps adding in apps and technology can improve the market knowledge that PE’s projects have?


Would love to hear your thoughts.



Distribution of farmers across the globe:;


GroIntelligence world food demand:

edited on Dec 18, 2017 by Cec Cameron

William Lee Dec 19, 2017

An initial thought - using market price data collection as an example

Why: given market price volatility/seasonality is a huge issue in Malawi, understanding market prices for different produce week on week is a crucial information point.

Idea: getting market sellers to text through price of crops on a weekly basis to a phone number managed by PE

How to:
1. Recruit a few market sellers in both major (Limbe, Blantyre, etc.) 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 may 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 to have to go out to each market each week in person
- In fact, it can be done completely remotely - even back in Aus potentially
- This data will provide inputs to Ag blueprint modelling
- It may also provide a future platform to make market prices transparent for consumers
- Depending on the extent of our surveys, the data could also be shared/sold to other organisations in the field. This will help connect PE with other powerful organisations, and provide us with something to bring to the table during discussions.

Let me know your thoughts.


Ella Grier Dec 20, 2017

This is a cool idea Will, with minimal set up costs involved in a text-based system, we could definitely look into setting up something like this to have working in the background of the Ag project so we have that first hand research there to analyse when working on the BMC.
I'm curious to see how this information would look as from a buyers perspective, prices can fluctuate not only seasonally, but on a day-to-day basis. And even between stall, prices will vary significantly.

I am wondering if you have any ideas as to how you would extend this beyond the markets themselves for a more comprehensive understanding of the supply chain within Malawi.

Additionally, how would ensure quality and accuracy control when receiving information from an external party. There is a chance that our market sellers would manipulate the price from an external motivation perspective.


Andrew Vild Dec 20, 2017

Would like to see this fleshed out more as an idea on CC! If you have the time, make it happen and it can be a side project over Jan/Feb that will allow us to test whether this is a viable (and valuable) data collection technique that I am sure would have value to an array of stakeholders.


Rachel Chan Dec 20, 2017

Hi William,

This sounds awesome! It would be really valuable in enabling our Agriculture project teams to gain a better understanding of the supply chain, which the Malawian government doesn't even have due to their lack of data collection.

One suggestion I have is instead of recruiting market vendors, there could be a ‘mystery shopper’ system made up of consumers to increase the accuracy of the data collected. This would be similar to how Waze crowdsources its data.

However, this may be an issue in the Malawian context as we have noticed that vendors can change locations in the Blantyre Markets. In addition, it may be difficult to incentivise consumers to submit data if they are unable to obtain much value from it (e.g. knowing where the cheapest produce is). Hence, if valuable information can be conveyed to them via text, perhaps it could be a service made available for a minimal fee, which can then be used to compensate the data collectors.


Alex Piatek Dec 20, 2017

Awesome idea Will!

One idea to avoid manipulation and to ensure market sellers are sending data on a regular basis, you should provide value back to the market sellers.
If we collected the price from different markets, and other produce sellers, we could create a mini report that could be sent back via text, to give insight for the market sellers.

E.g. if market seller sends through prices for carrots, they will receive a text back with the value of carrots from the other markets and produce sellers.

Users tagged:


View all replies (4)

Amber Johnston Jun 30, 2018

Status label added: Work Update