Adopted Experiment

[Experiment Adopted]: Fuel Fiji - December Goal 1: Iterate on the prototype to implement current customer feedback and ensure possibility for scalable manufacturing - December 2018

Lean Phase: Prototyping - Solution


Assumption: That the Buka 5.0 accurately addresses the problem space like the 4.0 did, whilst enabling scalability and cost reduction in manufacturing.

This is an important assumption to validate as the redesign of the Buka stove in July 18 made critical alterations to the stove design. Whilst the underlying principles were held the same and the rough design of the 4.0 kept, the change in materials alters many aspects of the stoves.


Time Box: This experiment is intended to run throughout the month of December, possibly longer as design iterations continue to be optimised. Initial results and iterations to the stove should be aimed for the start of week 3 in December.


Throughout the Fuel project, it has become clear that there are two types of experiments concerning the stove prototypes: Internal and External. Things like fuel consumption, emissions and particulate matter are difficult to measure in country, and must be done internally by PEV. Two stoves are in Australia to be tested for these aspects in the near future. Other metrics such as user experience, qualitative efficiency, portability and myriad others can and should be tested with customers, as their perceived value is more important than ours. This experiment focuses on these kinds of metrics that can be tested in December Week 1 of project.


The Key Metrics to be measured include the following:

1. Immediate appeal to customers.

One strength that the Buka 4.0 possesses is that customers were drawn to the product. This can be gauged subjectively but recorded quantitatively as either a strong or weak and positive or negative initial response to the product. For example, when customers first interact with the 4.0 the initial reaction was one of impressed interest. Gauging reactions is important due to the intention to distribute stoves through hardware stores, where making a statement on the shelf will contribute to the success of the product in that distribution method.

This reaction should be measured as many times as possible, which can be relatively simple to do as an addition to many of the other tests on the new prototype with customers. Something to note is that strength/loading capacity of the stove was something many customers were wanting to test initially. Some sort of evidence of the high static load capacity could be useful here.

Ideally this testing is done with a variety of different customer segments to gauge any differences in opinion across those segments. For some this will be more important than others due to frequency at, and prevalence in, hardware stores or similar.


2. Usability of the stove.

The 4.0 design is intuitive to use and given the move in distribution to include indirect sales (hardware stores or other distributions methods that are not PEV), the intuitiveness of the design is crucial for customer satisfaction. This can be assessed through letting customers attempt to use and light the stove with no instruction given and see how they fare and if there are any commonly occurring issues that need to be addressed.


3. Qualitative Efficiency

For obvious reasons of standardisation and personal preference when cooking, efficiency is extremely difficult to accurately measure with customers. One thing that is important for referrals and customer satisfaction, is the experienced and perceived efficiency of the stove. This is where the true value of the product lies rather than in controlled lab testing, and this will be touched on in the internal testing metrics.


4. Durability

The main way to assess durability is through long exposure testing. This is one of the main aims of the case studies we have prepared to run from July to December, with more information on these able to be found here.


5. Assessing whether handles are necessary or not.

The prototype does not currently have handles due to time constraints on the prototyping. Testing this feature with customers or whether portability through the stove being lightweight is enough. If a handle is desired, would they be best like the previous Bukas, two symmetric handles, or one handle on the side like on a suitcase. Would they only need to slot in and act as handles when being used, like a hook?

Green Light - Each metric above that is being tested will have slightly different success points. A rough approximation is: 60% of customers have a positive response to the specific metric.

This may differ and some metrics be time based however, especially regarding durability. In general 60% is a good target.


Orange Light - Anywhere between 40-60% of positive responses indicates that the product is doing okay on the metric being assessed, but can do with some improvement. Design alterations should be considered and  


Red Light - Should less than 40% of feedback from customers be positive, then a re-design of the stove should be considered. Whether this be the entirety of the 5.0 or just certain aspects to address the feedback will depend on customer responses.

Experiment build:

There is the potential for significant overlap between this goal and several other goals the involve interacting with customers, both existing and potential. To specify a precise experiment method would be unnecessarily restrictive for the December team. The only main consideration is any bias or framing that is introduced when gaining feedback on certain stove features. What metrics are tested when and in what combinations should be iterated on based on the success and findings while surveying. It is to be expected that there will be issues when trying to gather such a range of details, especially if conducted in conjunction with one or more of the other goals.


Ensuring that the feedback is gained across the different customer segments (goal 5) as they are identified is crucial to remove bias from a minority of the overall market. Individual segment preferences are useful to note down, for as scale grows there is potential for differentiation between stove models.


Tagged users
edited on 2nd December 2018, 21:12 by Haziq Ahmed

Andrew Vild 10 months ago

Status label added: Proposed Experiment

Reply 0

Haziq Ahmed 6 months ago

Status label added: Experiment adopted

Reply 0

Haziq Ahmed 6 months ago

Status label removed: Proposed Experiment

Reply 0