UX Case Study: Klinik Tani — An App to Manage Agricultural Agents.
Klinik Tani is an app to help its users (the agricultural agents) obtaining local agricultural problems, discussions related to scheduling of agricultural education, and manage the delivery of agricultural education from formal agricultural institute to agricultural cadres then finally accepted by the farmers.
The objective of this project is to create a managerial app as a portal for the formal agricultural institutes that needs human resources and young agents (from students and local communities) who are willing to become extensions between the formal agricultural institutes and the farmers.
As a result, Klinik Tani gets 5 main features including 1) Confirmation and rescheduling of agricultural education schedules; 2) Post and read educational materials; 3) Monitoring of farmers; 4) Counseling; 5) Community. This case study became a finalist in the national UX competition organized by the Directorate General of Research of the Republic of Indonesia in September 2020.
I took the role of UX Researcher. Besides that, I also worked on the UX Research with Wina Hafidh. In UX research I mostly lead and work in the early stages to get quantitative and qualitative data, while Wina leading role in the last stages, usability testing. And our other teammates, Edo Novanto does more business-related research roles.
The food agriculture sector is an important aspect to achieve food sovereignty and security as well as the people’s welfare. This is because in terms of production, food agriculture is the second most influential sector in Indonesia. In order to support it, the government through the Indonesian Ministry of Agriculture launched a PPL (Agricultural Extension Program) to help farmers increase their production output. The agricultural agents are assigned to different areas to provide direction, guidance, and counseling on a sub-district administrative basis.
Even though it has been running for quite a long time, the performance of PPLs in several areas is still insignificant, such as the inactivity of PPLs to go down among farmers to the quality of PPLs that do not meet qualifications. Some of these shortcomings were further exacerbated by the emergence of the Covid-19 pandemic which affected food agriculture activities in many rural areas.
Therefore, an idea emerged from us to build an agricultural education ecosystem called Klinik Tani. Klinik Tani is an application of technology to make it easier for farmers and agricultural agents to gather on one platform. Later, Klinik Tani will collaborate with related agents from the district to the smallest scope (the village). This is because each region has different agricultural potentials, so that the Klinik Tani ecosystem will be able to reach it efficiently and on target.
The method used in solving the problem is Design Thinking. Design Thinking was chosen because the process is iterative, flexible, and focuses on collaboration between designers and users.
First Phase: Empathize
The first phase of design thinking process is Empathize. This stage aims to understand the human needs involved. We gain quantitative data analysis by shared an online survey via Google Forms and qualitative data analysis by In-Depth Interviews. Both data are intended to validate our concerns and gain insight from our potential users.
To prevent Invalid Participation, steps are used in the form of Data Screening Protocol Reducing Capacity (Elizabeth N, Sheila RW, 2008). This step refers to the Proportional Stratified Random Sampling technique. This technique aims to obtain a representative sample by looking at a heterogeneous population and classifying them based on strata.
Our target potential users are:
- Students and the community as cadre candidates (extension of the formal agriculture institute).
- Formal agriculture institute actors such as PPL (Agricultural Extension Program) actors.
The final sample size decisions:
- Population (N): 42 respondents
- Target (n): 38 respondents
- Respondent’s strata (i):
- Student/general people (30 respondents) → we used this for the questionnaire
- Agriculture agent (12 respondents) → we used this for in-depth interview.
After we gain the size, then we validate it by calculate with this formula:
We managed to get 30 people as our participants. with 85% of them are students who currently pursuing “Diploma / S1 (Bachelor) / S2 (Master) / S3 (Doctor)” education and 60% of them have a discipline related to agriculture. The reason why we take 30 participants is in accordance with the opinion of Singarimbun and Efendi (1995) who say that the minimum number of questionnaire trials is 30 respondents. With a minimum number of 30 people, the distribution of values will be closer to the normal curve.
After that, we wanted to validate the need for awareness of agricultural issues and the desire to be involved in the agricultural sector in a real way. Here are the findings:
From the result, we can see that the majority of respondents have a relative/environment close to agriculture. Even though some of them are not from an agricultural background either, they still have an interest in agricultural issues and have a concern to help with agricultural issues around them.
For interviews, we used the focus group interview type and targeted 8–12 agricultural actors as the sample. The subjects included in the focus group were homogeneous subjects (one field). For that, the subjects must have been selected before the interview so that the subjects were homogeneous.
We wanted to know the actor, regulation, flow, and perspective of the formal agricultural institute. These are what we found:
- Agricultural empowerment actors involved:
Formal Agricultural Institution, LPP, Technical OPD, NGOs, CSR, PPL.
- Classification of types of farmers:
rural and urban farmers. But more so in rural farmers.
- Activities performed:
Education of rice field & land intensification, product processing, marketing (there is target status).
We also wanted to know the difficulties they (the formal agricultural institute) faced in explaining agricultural education to farmers. There are:
- Lack of motivation
- Some farmers do not have sufficient insight to be able to understand their problems
- Lack of human resources.
- The agricultural education is not well-organized and does not work well.
- During the Covid-19 pandemic, access to counseling was getting narrower. The difficulty of meeting in person is that farmers are not proactive in coming to extension services, and rarely use digital technology.
Second Phase: Define
This stage aims to design a persona that represents the target demographics of the target user.
Target User (Persona)
From our previous findings, we collected several behaviors, pain points, and goals that our potential users have. Then, we categorized the patterns found in our findings into two different personas:
After getting the persona, we created an affinity map based on the 3 conclusions of the solutions we adopted, including:
After obtaining an affinity map, the features are divided by priority using the following Eisenhower Method:
Through the priority matrix above, we decided to implement the features in the Important and Urgent and Important but Less Urgent matrix. The reason these features were important and needed is that we wanted to take an intersection from the MVP Klinik Tani so that even though it is still the simplest version, we can still convey the value of Klinik Tani and keep running optimally.
Third Phase: Ideate
At this stage, we initiate a solution to user problems. In the first step, we designed the application concept using a wireframe that pays attention to both design and psychological aspects.
This stage aims to implement all application flows in a visual form that is easier for the user to understand. Hifi design and prototyping are made based on the concepts that have been compiled.
Third Phase: Prototype
In this phase, we turned our ideas into a form of a prototype. We used Figma to create the prototype. Later, this prototype will be tested on some potential users.
Notes: all picture which are used in this design have a COC license.
Based on the prototype, there are several main features of Klinik Tani :
Homepage, Profile Page, Notifications
Fifth Phase: Test
In this phase, we wanted to validate whether our solution works out or not.
The results of the quantitative data obtained from this experiment were verbatim participants when filling in the UEQ questionnaire and interviewing them after completing the task. The data will be analyzed using the Focus Group Discussion method between the three research teams and formulated into recommendations for the future development of the Klinik Tani software for young cadres.
a. Task Completion Rate (TCR)
It appears that there are several tasks that failed to be done by some participants. Obtained some input and suggestions related to this, which will be discussed further in the qualitative analysis section.
b. User Experience Questionnaire (UEQ)
UEQ results will be analyzed using descriptive statistics. Descriptive statistics provide a description of the data seen from the mean, maximum, variance, and standard deviation values. This method is used to provide an overview of the dependent variable, in this context the UEQ scale (attractiveness, clarity, efficiency, accuracy, stimulation, and novelty).
Based on the UEQ testing of experimental participants, the mean value of six UEQ scales was obtained. Attractiveness Scale is 2,167, Clarity is 2,000, Efficiency is 2,000 too, Accuracy is 2,250, Stimulation is 1,333, and Recency is 0.667. Highlight colors represent values (green for good / above average, yellow for medium, and red for bad). To determine the value of each scale, it is enough to follow the following rules: if the result is> 0.8 then the product can be said to have a good UX, if it is between 0.8 to -0.8, then the product can be said to be neutral, and if <-0.8 then the product can be said to be bad.
Furthermore, benchmarking can be done using the standards provided by UEQ, namely by comparing the results of Klinik Tani prototype testing with existing standards. Benchmark values help provide conclusions about the quality of the product in comparison to the products generally on the market.
It can be seen that the first five scales of the Klinik Tani have fairly good scores (Good and Excellent), but apparently, for the novelty scale, Klinik Tani has a fairly low score (Below Average). The reasons for this value will be discussed again in the qualitative analysis below.
The results of the qualitative data obtained from this experiment were verbatim participants when filling in the UEQ questionnaire and interviewing them after completing the task.
The data will be analyzed using the Focus Group Discussion method between the three research teams and formulated into recommendations for the future development of Klinik Tani.
Here are some points that we recommend for the next development iteration of Klinik Tani’s product:
- Uniform the back button on the profile page like the others (in the top left corner, not in the bottom center) to improve learnability and consistency.
- Changes the “view log” button to become more visible when it can be pressed and becomes a navigation tool.
- Replace the (+) button on the top right for important and frequent activities such as submitting an extension schedule and adding monitoring data with other better options such as the CTA (Call-to-Action) button.
- Klinik Tani is considered to be able to solve agricultural problems, especially regarding the education of farmers through young agricultural cadres and agricultural offices in the local area in dealing with Food Security Fluctuations during the Covid-19 pandemic to agriculture in the future.
- Klinik Tani is very capable of being applied in all lines of extension programs or activities. The results of research and testing of the Klinik Tani application will later become suggestions for improvements for a better user experience.
- The design is not finished only at the design stage. Even though it has been tested, we still have to make improvements for future design development.
Fao.org. 2020. FAO — News Article: Joint Statement On COVID-19 Impacts On Food Security And Nutrition. [online] Available at: <http://www.fao.org/news/story/en/item/1272058/icode/>
Nosen, E. and R. Woody, S., 2008. Online Surveys: Effect of Research Design Decisions on Rates of Invalid Participation and Data Credibility. Graduate Student Journal of Psychology, [online]10,p.1.Available: <http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.522.1857&rep=rep1&type=pdf>.
Singarimbun, M dan Efendi,. 1995, Metode Penelitian Survey, Jakarta : PT.Pustaka LP3ES