Project Everest

Experiment Results

[Experiment Results]: Channels – Sales Agents – Fuel Fiji July 2019

Experiment Design Post:

Lean Phase:

Distribution Channels


People are willing to act as sales agents using a feasible system 


In order to automate sales of the Buka Stoves in Fiji all year round, the July team has surveyed and presented four options to local Fijians who wish to sell stoves for PEV. These are:


Assign unique reference numbers to each sales agent. Put these numbers onto flyers and give them to the sales agents to pass around. The flyers will direct leads to retailers stocking the Buka Stoves. At the point of sale, the retailer would physically collect the code from the customer and pass it on to PEV. Sales agents will be paid a commission if their lead results in a sale.


Collect phone numbers of people who are interested in the stove and pass the leads on to PEV. Sales agents will be paid a commission if their lead results in a sale.


Sales agents purchase in bulk. PEV trains sales agents to present in villages and market stalls to promote the product and make sales. They may sell the stoves at their own markup price.


Sales agents buy in bulk. PEV will give leads the contact details of these sales agents who will then close sales. They may sell the stoves at their own markup price.


Green Light 

The results of this experiment exceeded our success point (two individuals agreeing to act as sales agents for the Buka Stove using one given system). The distribution system that involved sales agents collecting contact details and passing them onto PEV was the preferred relationship amongst those surveyed. 24 individuals expressed a willingness to become a sales agent. 

Validated Learning:

The results of this experiment supported our assumption that people are willing to act as sales agents using a feasible system. 

The metrics measured were the number of sales agents recruited over the experiment’s timebox (1.5 weeks) and the preferred relationship to PEV. The leads were further sorted according to willingness to become a sales agent. 13 of the 23 potential sales agents confidently expressed interest in the role while the remainder agreed but showed hesitation. 

The four relationships offered differ in the level of commitment required from sales agents, and automate different parts of the sales pipeline. See the Experiment Outline for more detail on these. Although sales agents were required to rank the options from most preferred to least, it was found that they only liked Option 2: provide leads to PEV for commission on successful sales. All other options were not considered. Only the community in Kabisi Village considered buying in bulk. 

Approximately 11 potential sales agents recruited were from villages. The rest were from Sigatoka town. The sales agents recruited from villages included past customers, those looking to purchase a stove and those that knew about the stove. They were recruited through phone calls and village visits. Due to language barriers and the complexity of what we were trying to pitch, it was found that village visits (or meetings in person) were the most viable option to recruit sales agents. Even past customers we called requested we meet them in person. 

The sales agents recruited from Sigatoka town consisted of market stall vendors, taxi drivers and shop and restaurant owners. PEV’s presence in Sigatoka town made the process of recruiting people with no previous relation to PEV or knowledge of the Buka Stove easier.

Next Move:  

As Option 2 has been strongly validated (and all others invalidated), the fuel team will proceed to trial this distribution channel to establish whether it is effective. However the reliance on a human element means there will be variables requiring control.

We will need to provide training to ensure the sales agents can qualify leads. We plan to invite those who are interested to workshops we will run next week. This will still need to be confirmed. The workshop will go over Project Everest Ventures, the Buka Stoves, the sales pipeline, the groups of people to target, the Project Buka Facebook page, how to pitch and monetary rewards. A script incorporating the sales pipeline and spreadsheet have already been created as part of a starter pack. This will be handed to sales agents at the workshop to take with them.

Although the sales agent channel has been validated, PEV still needs a way to automate the second half of the sales pipeline. The team has discussed the possibility of using Natasha, our cook, to stock Buka Stoves in her market stall and close the leads in person there. Natasha is ideal since PEV trusts her and she is very familiar with the stove. If the sales agents have followed the script closely, they should be providing quality leads. These will be submitted through the Facebook page. PEV will then be able to call and direct them to Natasha’s stall in Sigatoka Market. We will also be able to inform Natasha beforehand the number of stoves she should stock for the week.

On a separate note, the other two channels which still need to be tested this month are Headman and Retailers. These could also work with Natasha’s stall as distribution channels to complete the sales pipeline process.

edited on 17th July 2019, 21:07 by Andre Werror

Andrew Vild 6 months ago

Hi Andre,

Good post and glad to hear you have had success in this channel pursuit!

Can I request that you briefly outline what the options are and what you have validated. Trying to go into the experiment build still doesn't clearly tell me what the 4 options were that you tested and which was successful. All I know is that "option 2" was successful.

I eventually found what it meant by delving into the rabbit hole and downloading an excel spreadsheet.

To make this easier to read and navigate, reducing bulk prose and involving dot points or a basic summary at the top of the page.

Reply 1

Andre Werror 6 months ago

Thanks for the feedback, Andrew. I have updated the post to include more context of the options we presented to Sales Agents.

Reply 0

Eugenia Muñoz 5 months ago

Status label added: Experiment Results

Reply 0

Eugenia Muñoz 5 months ago

Status label added: Customer Segment

Reply 0

Eugenia Muñoz 5 months ago

Status label removed: Customer Segment

Reply 0