See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/303323831

Usability challenges in an Ethiopian software

development organization

Conference Paper · May 2016

DOI: 10.1145/2897586.2897604

CITATIONS READS

0 76

3 authors, including:

Yvonne Dittrich
IT University of Copenhagen
146 PUBLICATIONS 1,148 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Urban Data 2 Decide: http://www.urbandata2decide.eu/ View project

All content following this page was uploaded by Yvonne Dittrich on 23 May 2016.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
 2016 9th International Workshop on Cooperative and Human Aspects of Software Engineering

Usability Challenges in an Ethiopian Software

Development Organization
Degif Teka Yvonne Dittrich Mesfin Kifle
IT Doctoral Program, Systems and Software Section, Department of Computer Science,
Addis Ababa University IT University of Copenhagen Addis Ababa University
Addis Ababa, Ethiopia Copenhagen, Denmark Addis Ababa, Ethiopia
[email protected] [email protected] [email protected]
ABSTRACT
Usability and UCD are becoming more important in software
Usability and user centered design (UCD) are central to software
development [6, 17, 23]. Software engineering methods like Agile
development. In developing countries, the gap between IT
Software Development Methods (ASDMs) are promising to
development and the local use situation is larger than in western
integrate usability. However, there is not a straight forward way to
countries. However, usability is neither well addressed in software
do so. ASDMs are increasingly adopted in industry and also in
practice nor at the policy making level in Ethiopia. Software
Ethiopian software industry.
practitioners focus on functional requirements, meeting deadlines
and budget. The software development industry in Ethiopia is in In countries like Ethiopia where ICT industry is not well
its early stage. The article aims at understanding usability established, there are additional problems such as economic
practices in an Ethiopian software development company. resources and skills. Most of the software companies in Ethiopia
Developers, system analysts, product owners and users were are young and inexperienced [4][12]. Another challenge is that
studied. In this first phase of the research, participatory there is no formal education on usability and human computer
observation, a workshop and interviews with practitioners and interaction (HCI): the curriculum of higher institutions do not
operational staff were analyzed. Informal discussions have been have even introductory course in this area. In August 2014, prior
observed to outweigh formal meetings for sharing experience and to the research presented here, an interview study was conducted
ideas. Practitioners’ internal configuration, their experience, with two interviewees from government IT organizations and two
cultural knowledge and common sense regarding the users’ from private software companies. The analysis revealed that
situation guided the design. Prototypes and fast delivery of usability and UCD practices are considered important to address
working versions helped in getting user feedback even if early user needs but provide challenges for the practitioners.
user focus proved to be a challenge as communication between
developers and users suffered from several layers of indirection. This article reports on the initial phase of an action research study
Further challenges are the heterogeneity of users to be supported, addressing the questions: How is usability addressed in the
a lack of awareness of usability methods, and lacking resources. context of agile software development? And what are the specific
challenges when addressing usability in an Ethiopian company?
Keywords
User focus; usability; ASDMs; UCD; agile team communication; The research presented here investigates usability practices and
product owner; operational staff; PC. challenges in an Ethiopian company in the context of agile
software development. The projects subject to the field study
1. INTRODUCTION develop special purpose applications of mobile services
For interactive systems, usability is utterly important. In Ethiopia connecting rural communities to the capital. They are designed for
in many cases, ICT and mobile services are leap frogging paper two different organizations, with different customers and users,
based administration. Especially in rural areas, part of the geographically separated, speaking different local languages and
population is illiterate or semi-literate. Many can only afford low- culture. One of the application is for the northern region of
end mobile phones. Usefulness and usability are of core Ethiopia, while the other is developed for the southern region of
importance for this part of the population to take advantage of Ethiopia. However, the two projects have similarities in the
ICT services. application domain and, to a large extent, share the same code
Usability has been defined by Nielsen as a property of the user interface base.
with the attributes efficiency, learnability, memorability, error free and
The next section discusses related work. Section 3 details the
subjectively pleasing [24]. The operational definition followed in this
article is the widely used one given by ISO 9241-11, which defines research methods applied. The detail of the research is presented
usability as “the extent to which a product can be used by specified users in the section 4. The discussion in section 5 relates the issues
to achieve specified goals with effectiveness, efficiency and satisfaction in encountered in section 4 to the issues in the related work and that
a specified context of use” as cited in [19]. way develops the contribution of the research. The conclusion
Permission to make digital or hard copies of all or part of this work for summarizes the findings.
personal or classroom use is granted without fee provided that copies
are not made or distributed for profit or commercial advantage and 2. RELATED WORK
that copies bear this notice and the full citation on the first page. Latest with the introduction of personal computers in the 70ties,
Copyrights for components of this work owned by others than ACM usability and HCI have become an issue for software engineering
must be honored. Abstracting with credit is permitted. To copy and made their way into the curriculum in many European and
otherwise, or republish, to post on servers or to redistribute to lists, North American universities. With the diversification of the
requires prior specific permission and/or a fee. Request permissions
research discourses though, the integration of software
from [email protected].
CHASE'16, May 16 2016, Austin, TX, USA engineering and HCI maintained a disputed area. The related
© 2016 ACM. ISBN 978-1-4503-4155-4/16/05…$15.00 research section first discusses software development in the
DOI: http://dx.doi.org/10.1145/2897586.2897604

114
contents of ICT and development. The few articles that are development: Dyba and Dingsøyr [16] presented systematic
published on the topic indicate the need of iterative development review of empirical works in agile software development up until
to accommodate user participation and the use of early prototypes the year 2005. Their report indicated that agile software
to finalize the design. As agile development is one of the development is increasingly adopted. Furthermore, their review
recommended approaches and many of the companies in Ethiopia indicated code quality, developer satisfaction, the increased
and also the case company use ASDMs, we present the state of cooperation with the customer, and customer satisfaction. They
the art on agile development and usability. have indicated that empirical studies in this area remain scarce
and called upon more empirical investigations. However, they do
The discussion in the ICT for development (ICT4D) community not mention the end user engagement, and do not problematize
focuses very much on the design, implementation and use of ICT that the on-site customer is rarely the end user. A follow up study
in developing contexts. The research in this area has by partly the same authors [11] states that quality and quantity of
acknowledged that not only the design results need to take cultural research on agile methods increased. The authors summarize
specificities into account, but also the design methods have to be research contributions about appropriateness of agile methods,
adjusted to the specific context [31]. The same is proposed for adaptations and reconciliations of agile methods. Again, user
software engineering [4] and HCI [30]. The software development centered design in agile methods is not mentioned.
process is by and large ignored. An exception is an article by
Doerflinger and Dearden [15] on “Evolving a software Addressing usability and hence user involvement has several
development methodology for commercial ICTD projects, advantages for software development [20]. Flexible ways of
Information Technology and International Development”. The addressing user needs are crucial for software companies who
authors proposed a close collaboration and cooperation between exist in a competitive market. UCD and hence user involvement is
project stakeholders involving multiple parties, emphasizing a important when developing useful and usable software. Dittrich
systematic approach to study end users using local personnel and Lindeberg, report that the involvement of users in software
instead of relying on researchers who are new to the local development in combination with control of schedules and budget
community. The agile approach they propose combines is possible and improves the software [13]. A systematic literature
evolutionary and iterative development and the reflective action study indicates that the participation of users has demonstrable
research of the researchers that generates input and feedback for advantages [1].
the design. They emphasize that every project has a unique
makeup in terms of scope, use situation, IT skills, availability of The case study by Hansson et al. [18] shows that flexible
resources and other contextual factors that have to be taken into development practices can successfully facilitate active user
account when setting up the software engineering project. The involvement. A combination of UCD techniques and agile
necessity of an iterative process and explicit user focus has also approaches is possible as has been reported by Chamberlain et al.
been documented in a case study on two projects by Pade-Khene [8]. In their examination of professional practice based on
et al. [25]. Sustainability of such projects depends on that the interviews of UCD practitioners involved in agile software
projects are community driven, initialized with pilot projects in development, the authors reported that ASDMs have a distinct
the community, and iterative and incremental in design and culture that at first glance seems to conflict with UCD [21].
development to help identify the requirements and challenges for However they also report that the use of agile methods can result
the community (users). in improved usability.

Dörflinger and Dearden propose ASDMs as a starting point to In two recent SLRs, the research on the combination of UCD and
integrate user participation and feedback when developing ICTD ASDM has been sumarised. Salah et al. [26] report that agile
applications. The use of ASDMs is also wide spread in the methods suffer from lack of allocated time for system design and
Ethiopian IT sector. ASDMs are flexible, iterative and user research. Some studies propose upfront design though
incremental methods focusing on customer collaboration, extensive upfront design and planning is considered as waste in
individual interactions and responding to changes [2, 10]. There agile development and is against the agile values and principles.
are many success factors identified for software development Salah et al., therefore, recommended for the close collaboration
organizations with ASDMs [22, 9]. between the development team and designer. The SLR also
indicated some of the usability testing methods that could be
The main focus of agile processes is how to organize the required carried out in agile development. Brhel et al. [7] looked into the
tasks in a flexible way to reach at the overall goal of delivering current state of user centered agile development (UCASD). In the
working software [3, 27]. Project management, team organization, studies they reviewed, usability work relies on the team members’
design and coding techniques are organized to support flexibility own initiative and understanding, user involvement by and large
and adaptive planning. While delivering working software is a takes place in an ad hoc manner. The study proposes five
required condition of usable systems, ASDMs focus more on principles related to processes and practices of software
efficient coding; usability issues might be ignored as an explicit development and recommends future research in this area should
user-centered focus is lacking [5]. look into people, social and technological aspects for UCASD.
The open issue up until to date is who should be responsible for
The focus on engineering and project management aspects of usability and quality requirements: usability specialist or a cross
ASDMs can prevent an orientation towards UCD [5]. In many of functional team. The authors require a clear definition of who are
the methods, an on-site customer is expected to represent the the users in the scrum agile method and to clearly place the
users’ perspective when asked to elaborate on the user stories and responsibility for usability and for the quality of the products.
implementing acceptance tests. However the customer may not
know the needs of the actual users and could be from a different The analysis and discussion presented below is thus dealing with
organization than the actual users are. Moreover, none of the two sets of issues: the general challenge to combine UCD and
major ASDMs explicitly incorporates usability-engineering ASDMs and the specific challenges of doing so in a developing
practices. This ‘blind spot’ also becomes visible in the systematic country. The challenges of integrating usability and UCD into
literature review (SLR) on empirical research on agile ASDMs mentioned in the literature are: lack of early and

115
continuous focus on users, developers aiming at meeting that there are heterogeneous users who are geographically
deadlines and budget and, hence, developers considering usability dispersed, speaking different languages, with different cultures
as something that could be addressed later in development if time and hence usability matters. The project team is composed of
permits, lack of clear definition of the user perspective, lack of roles product owner, three developers (software engineers),
awareness of usability by developers and managers, lack of project manager / scrum master and two operational officers (i.e.
organizational support and also that ease of use and futurity, the product training officers). Operational officers are responsible for
ability of the product to support evolving needs, are not deploying successive versions at the users’ site, training users, and
considered [6, 29, 7]. Software development in the developing communicating users’ challenges to the developers and the
countries is characterized by low infrastructure, lack of manpower product owner. Due to the non-disclosure agreement (NDA), we
and technical problems or knowledge. This goes hand in hand are not able to disclose the application domain and functionality
with a lack of the ICT knowledge by the majority of users of IT of the system.
products. In addition, developing countries like Ethiopia suffer
from the lack of HCI education and trained HCI practitioners. 4.1 The Development Process and Artifacts
This is reflected in a lack of awareness on the policy making The projects considered for this research are ongoing contract
level, as for example the procurement and acquisition processes in projects and have been implemented using an adaptation of
Ethiopia do not consider usability as a criteria. As already SCRUM, one of the most widely used agile development
discussed above, the culture in developing countries might require approach.
different ways of product development that fit the cultural context
Upon agreement with clients, the product owner and operational
[31, 4, 30].
officers visited the client site for detailed discussion and
3. METHOD presentation of the company profile and experience particularly
The research has been carried out in a young private software related to the project. The operational officers are both bachelor
company, which is actively involved in the software development degree graduates in business and marketing. They learnt to work
in Ethiopia. The company mainly develops Software as a Service with software practitioners and users through experience. Both
(SaaS) including products for public private partnership (PPP) but operational officers and product owner met users and customers at
also contract projects for specific customers. In addition to the the customer site for interviews, observation and documentation
PPP projects with giant public organizations in Ethiopia, it is also of the as is situation to gain an understanding of the domain
oriented towards digital financial support of the rural people. knowledge during the initial period. The operational officers’ role
is later to train users, to follow up with users on their challenges
The research presented here is part of a PhD study focusing on and to report to project manager and product owner. The product
integrating usability and UCD into agile methods inspired by owner (PO) used paper sketches and other tools for eliciting
cooperative method development (CMD) [14], an action research further requirements. During the succeeding periods the PO got
approach combining qualitative empirical research with software comments, challenges, request for change and other requests from
engineering method improvement. The research reported here is the operational staff who collected field data and user support
meant to inform the deliberation and implementation of requests. The IT support personnel were recruited and trained
improvements in cooperation with practitioners involved in the during the first deployment as detailed in section user contacts.
development process. They were meant to support the rural users but turned out to
operate the system on behalf of the actual end users who are
The first author has been involved from the beginning of the farmers and chair men of farmer cooperatives.
projects, observing and documenting the software development
practice two days per week in the company in order to understand The product owner working with the project has a master degree
the actual software development practice and usability challenges. in economics. She is an experienced IT user and learnt the
Additionally, the research consisted of initial interviews with technicalities of how to work with software practitioners through
developers, the product owner responsible for the project, and an meetings and experience. The PO developed the requirements that
operational officer supporting the deployment of an early version, form the product backlog and continuously enhanced and
attending project meetings, and observing and interviewing prioritized the product backlog. The backlog is recorded in a table
operational staff and users in the rural area. As recording of the format. User stories are captured in the form of paper sketches as
meetings was not possible, the analysis relies on field notes. One shown in figure 1 or as graphs represented on a spreadsheet. The
workshop with software developers, product owner and PO then discussed with developers on the requirements, prepared
operational officer has been carried out in addition to the user stories, time estimate for each user story as shown in table 1,
observation. Initial analysis has been done identifying issues that similar to sprint backlog and sprint planning. User stories are
are meant to inform future action research. This initial phase has prepared in collaboration between the programmers, the project
been performed starting from December 2014 to June 2015. manager (scrum master) and the PO. User stories are assigned to
team members based on their preference and expertise. Three
4. AGILE DEVELOPMENT IN AN software engineers (SE1, SE2 and SE3) were involved in the
ETHIOPIAN SOFTWARE COMPANY projects. Sprint planning meetings took place once every sprint
This section presents the initial analysis and the findings from the before the start of the sprint. The PO prepared requirements in
empirical research. We start with describing the development ways understandable for both developers and customers as for
process and presenting the development team. The following sub example shown in figure 2.
section then presents the findings regarding contact with the users A sprint usually takes three weeks but sometimes may go on for
and customers. The section concludes with identifying related up to a month. After the sprint planning meeting, the PO follows
challenges. up on the progress. The PO and developers have frequent contact.
The purpose of the observed projects is to provide services for The meetings are sometimes initiated by the PO and sometimes by
rural communities that connect them to the center. This implies the developers in order to clarify questions regarding the

116
functionality of the system. These meetings are not regular or have master degree in computer science. All developers in the
scheduled. The developer team, PO and operational staff all have case company have a minimum of either bachelor or master
their offices on the same floor. degree in computer science or information systems. Usually,
different teams work with a number of projects and there is often
A sprint ends with sprint review meeting which also includes the not a one to one match between teams and projects. Team
sprint retrospective. Usually, the PO and the team participate in members may be moved from one project to another if necessary.
these meetings. The PO verifies the completeness of the user Not all projects, especially not the offshore projects, are
stories for the sprint. The team and the PO discuss what went implemented using agile development, as the practice is not yet
well, what the challenges were and what has to be improved in the mature.
successive sprints. After delivery, the team, the project
manager/scrum master, the operational officer and the PO meet Internally, the PO represents the customer and tries to follow up
and discuss the issues with the deployment of the software. the progress of the project, clears the doubts of the developers.
The PO keeps contact with customers for any additional
requirements or change of requirements. On top of this, the team
also receives information about the user and use context from
operational staff.
The project manager acts as a scrum master. He follows up on the
progress of the development team, discusses any obstacles and
guides the junior team members. The scrum master also interacts
with the PO during the project for change requests and new
requirements approval.
Total purchase (in Birr) per primary cooperative (PC)

PC 6-Oct-14 13-Oct-14 20-Oct-14 27-Oct-14

Figure 1. Sample Interface design by the PO PC1 23069 19258 36123 48001
PC2 16500 26300 25000 32012
PC3 18900 13600 15200 26985
Table 1. Sample user stories for the mobile application
Estimate PC4 15838 17500 38695 26036
User stories (hour) Responsible PC5 56369 25856 86756 49856
Add Loading/Unloading interface to
the mobile application. fields
required(type, variety, quality,
amount, driver name, driver’s phone Total purchase (in Birr)
number, truck’s plate number,
destination point, distance from or to 100000
(...) (in km), transportation fee (only
when unloading) 16 hrs SE1 80000
The Loading/unloading interface 60000
should also have date and time
stamp. It should also have the option 40000
to edit or delete. 8 hrs SE1
20000
Add Address(location) information
of PC at the back end web 0
application 8 hrs SE3 pC1 PC2 PC3 PC4 PC5
Create database table for
loading/unloading 8 hrs SE2 41918 41925 41932 41939

Add SMS service to transfer the

loading/unloading data to the server 16 hrs SE2 Figure 2: Sample material supporting user stories
Send SMS notification for the driver
The message should include the total
4.3 User Contacts
amount of crops delivered, date and The system is meant to be used by different kinds of users: The
time, and the pickup/destination back-end system will be used by well-educated staff at the
point 8 hrs SE2 headquarters and the warehouses of the customer. The mobile
application interfacing with the system is meant to be used by
Build test data repository 8 hrs SE3
transport agents and IT support personnel and farmers in the rural
Ethiopia. The initial contact between PO and the customer also
4.2 The Development Team included interaction with the users at the headquarters and the
The development team is collocated in the same room, where they warehouses.
can easily turn around and help each other on any development
issues or challenges. Two of the team members have a bachelor The contact to rural users first took place after deploying a first
degree in computer science, one of them and the project manager version of the application. ‘User training’ then was offered by

117
operational personnel to the support personnel, who were list of challenges was prioritized for further action. The challenges
recruited to support the introduction and initial usage at the range from difficulties with public infrastructure, like mobile
countryside. These IT supporters have a 3 year education related network and electricity in the rural setting, to the interface design
to IT. The IT supporters ended up operating the system on behalf of the application.
of the users due to the lack of IT skills and lack of high end
mobile phones in the rural communities. Each such IT supporter An example for usability issues caused by the lacking
interacts closely with a so-called core farmer who is a member telecommunication infrastructure is raised by the IT support staff:
and contact person of the primary farmers’ cooperative (PC). The The mobile app communicates with the server via SMS messages.
IT supporters upload data into the system on market days i.e. This problem has two sources: not only is the mobile network is
when transactions take place, accompanying the PC to the places not always available in the rural area but also the messaging
where crops is collected and loaded onto a lorry to the regional platform of the company happens to fail to operate. The central
hub. A PC has several members supplying agricultural yield to be example for UI related challenges are language localizations and
collected. During the pilot phase, the system is used in about 10 the understandability of error messages.
PCs in the southern project and 5 PCs in the northern project. The According to the participants, the most important challenge is
assumption is that in the future the PCs themselves will learn how related to the development process: the lack of prioritization of
to operate the system. Training for the IT supporters on how to usability issues due to the focus on functionality and on meeting
use the mobile app was offered by training officers, or operational deadlines and budget, lack of practice for usability testing and the
personnel. The IT supporters operating the mobile app cooperated long chain of representations between end users and developers.
with the PC in recording data and performing the transactions Another highly prioritized challenge was raised by the scrum
mainly on market days. Usually, there are two market days per master: “We developers need single point of contact”. He
week in one location. There can be more than one market at the complained that product owners and operational staff have user
same day in different locations. contacts and talk about requirements to the developers. Proper
The operational staff reports bugs, user challenges and additional way of communicating usability requirements to the team was
requirements discovered during the training to the project also raised as one of the challenges. An additional development
manager and product owner. The IT support staff work with the related problem is that developers are assigned to several projects
PCs and reflect the PC’s feedback and requirements. The first at the same time. Often a specific project is paused and the
version is the result of the first sprint of the development process. developers work on other projects.
The operational staff also logs the support calls from users and IT These challenges need the understanding and cooperation of the
supporters at working days from Monday to Friday. management and the project members. The next step is to work
The interview with the operational personnel revealed that after for action and improvement in the next development phases.
the deployment of the first version, user respectively IT supporters
recorded a number of challenges: understanding the terminology
5. DISCUSSION
In line with the related work on UCD and agile development [6,
on the interfaces; need of customized report generation which is
21, 29], the initial research presented above confirms that it is not
specific to each IT supporter’s PC, for example, to let the user
straight-forward to integrate both methods, though the iterations
view only suppliers to the specific PC; the difficulty of
in ASDMs open up for including user feedback into the
understanding application generated error messages; and the need
development. Developing software for users in rural Ethiopia
of language support i.e. Amharic for the northern project. The
increases the challenge.
challenges were communicated to the operational personnel on
her visit to the field to receive comments after deployment on the Though developers and users are living in the same country, the
northern project. The operational officer visits the field usually distance and difference between software developers in the capital
once per month i.e. when a new version is deployed, interacts with and users in rural communities is difficult to bridge. Additionally,
users and discusses their challenges. The PO visits the user sites to the IT skills and literacy level of the intended users provide
discuss with users about their challenges and to gather additional challenges not only for the communication about the functionality
requirements to later discuss with the client. Additional but also for the usage of the early versions. As a result, we
requirements and change of requirements are communicated and observed a chain of intermediaries between the intended users and
logged and considered on the next sprints. the development team. As using the early versions turned out
difficult due to the lack of high-end mobile phones and IT skills of
For the development team the PO represents the users, both the
the intended users, the software provider recruited IT support
corporate users and the farmers’ cooperatives. She communicates
staff. The IT supporters used the software on behalf of the
the needs of users and also proposes concrete interface sketches.
intended users. With other words, the intended users are
She also gives comments on the interface the developers designed
represented by proxy users. These representatives then provided
with respect to its ease of use. The PO uses the interface design
the ‘user feedback’ for the team. The operational staff and the
for her discussions with development team. In most cases, it has
product owner further translated this feedback to the team, but
been observed that the PO follows up meeting deadlines and the
they also represented other stakeholders and the central customer
fulfillment of the functional requirements as with the development
for the project. This turned out to be confusing for the team as the
team.
scrum master complained: ‘I need a single point of contact about
4.4 Usability Challenges user challenge reports’.
During the process a number of challenges for usability and An interesting observation though is that for the team, the
functionality has been encountered. In a half day workshop held feedback by proxy users, customer representatives and the PO all
in the case company’s office involving participants from the was talked about as user feedback. Further, in the discussions, the
developers, product owner and operational staff, the observed heterogeneity of users connected through the service was not
challenges were listed and complemented by the participants. The reflected in the way the team talked about ‘the user’. This can turn

118
into a problem: Rather than relating to concrete users with We started out by raising the research questions: How is usability
concrete difficulties, ‘the user’ might become a ‘scenic feature’ addressed in the context of agile software development in
[28] in the discussion of the team. In the projects observed by Ethiopia? And what are the challenges regarding usability in a
Sharrock and Anderson the real users were not involved. Sharrock typical Ethiopian company?
and Anderson [28] conclude their investigation of how the
designers referred to the user: “However, this does not mean that The results can be summarized as follows: Usability and user-
the user were not present in the designing. … we want to say that centered design are not addressed in any systematic manner in the
‘the user’ was a ‘scenic feature’ of the design process in that what observed projects but relies on common sense of the involved
users would want, what they might do, what they would be practitioners. Usability issues are gathered based on the
willing to accept were treated as significant and sometimes even deployment of a first version. As the use of the mobile
decisive” [28, p. 11]. The issue, though, is that if the assumptions applications turned out to be cumbersome, the company employed
about ‘the user’ are not realistic, usefulness and the usability of IT supporters using the software on behalf of the actual users. The
the product suffer. feedback of these supporters provided the main feedback
regarding usability of the software. On the side of the SCRUM
The difficulty of representation of the heterogeneous user groups team, the heterogeneous user groups are not distinguished. The
that became visible in our fieldwork as well as the developers’ use danger here is that the user becomes a ‘scenic feature’ rather than
of informal and common sense representation of user needs has a concrete person with specific skills and challenges,
also been mentioned in [7]; the user perspective is of an ad hoc
nature in an agile-scrum project requiring clear definition of user Based on our fieldwork and a workshop together with the project
perspectives. The literature study further requests future research members and managers of the company, we identified and
in the area of the social aspects of user centered agile prioritized a number of challenges when addressing usability. One
development working in other domains like organizational science of the questions to be addressed in future research is to what
and sociology. extent the challenges are a result of agile development not taking
heed of usability issues and to what extent the challenges are due
Given these initial observations and the discussion of at the to the specific conditions of software engineering in a developing
workshop, a number of challenges become visible: First, how can country context. In many cases both causes might contribute:
the rural communities as intended users be included in the design ‘normal’ problems of integrating UCD in ASDMs are aggravated
of complex IT services? Though of utter importance, this question due to the specific circumstances. The fieldwork indicates that
lies outside the research presented here; it would constitute a PhD normal usability challenges are emphasized due to literacy level,
thesis in its own rights. Interested readers might refer e.g. to [31] technical skills and access to high-end mobile phones of part of
or to the article by Zewge et al. [32] with respect to the Ethiopian the intended users. Different local languages and lacking
context. Second, how can the intended users and their input or telecommunication infrastructure require additional technical
feedback be represented as comprehensive and as authentic as considerations. It is less clear how the cultural context affects the
possible for the development team? Third, how can the software development. That management prioritizes functionality
heterogeneity and diversity of users and use situations of complex and budget, feedback on usability issues is not systematically
IT services be represented for the development team in a way that communicated to the developers, or the need to work with several
allows the developers to take in the right usability input at the projects in parallel impairs the intention to emphasize usability
right time? Fourth, how to coordinate the cooperation between the resembles very much findings from the developed world. Also
development team, the PO and the staff at the operational offices that the feedback by the intended users is filtered through a long
of the company? chain of ‘representatives’ might to some extent be also observable
in other contexts but, like the usability issues, is aggravated due to
Inspiration on how to organize close cooperation and the heterogeneity of users and the distance of the situation of users
collaboration of users can be taken from Dörflinger and Deaden: and the situation of the developers. The next steps in the action
local change agents who are familiar to the users’ situation and research will help to understanding to which extend the challenges
who can act as intermediaries to communicate user challenges and are specific for the Ethiopian context: How methods are adopted
use situation to the developers; the use of presentation of context and adapted reflects on both, the method as well as the challenges,
related knowledge like process descriptions, scenarios and which they are meant to address.
personas has been proposed by Doerflinger and Dearden [15].
The beginning action research explores the use of personas and
However, the challenge regarding the coordination and other representations of the users and their contexts. Such
cooperation between the development team, the PO and the representations could address the difficulty to design for
operational staff and also other challenges listed in the section heterogeneous user groups and could support the understanding of
‘Usability Challenges’ makes visible that a solution requires not the users and their situation by the PO and the development team.
only changes to the development processes, methods and tools, The challenge here is not only to introduce the work with
but also organizational support for usability as a central quality. personas, but also to adapt their usage to the agile development in
The next phase of this research aims at proposing solutions to the the company. Discount usability practices like heuristic evaluation
observed challenges, implementing them together with the and other light-weight usability evaluation methods that take into
practitioners and observing the improvements in practice. account the local context will be explored.

6. CONCLUSION AND FUTURE 7. ACKNOWLEDGMENTS

We would like to thank our collaborating company for its support
RESEARCH and cooperation. Thanks to the development team and operational
In a country like Ethiopia where its IT services are leap-frogging
and support personnel for their openness and participation. We
paper based administration, usability is of utter importance.
also would like to thank the reviewers for their supportive and
constructive criticism.

119
 8. REFERENCES Systems Design in the Software Development Process.
[1] Abelein, U. and Paech, B. 2015. Understanding the Influence Softw. Process Improve. Pract., 2003; 8: 111–131.
of User Participation and Involvement on System Success–a [18] Hansson, C. Dittrich, Y. and Randall, D. 2004. Agile
Systematic Mapping Study. Empirical Software Engineering, Processes Enhancing User Participation for Small Providers
20(1), 28-81. of Off-the-Shelf Software. XP 2004, LNCS 3092, pp. 175–
[2] Agile Alliance. Manifesto for Agile Software Development. 183, Springer-Verlag.
2001, Available: http://www.agilealliance.org, Accessed on [19] Jokela, T. Iivari, N. Matero, J. and Karukka, M. 2003. The
March 2015. Standard of User-Centered Design and the Standard
[3] Beck, K. 2000. Extreme Programming Explained: Embrace Definition of Usability: Analyzing ISO 13407 against ISO
Change. Addison-Wesley. 9241-11, CLIHC '03: Proceedings of the Latin American
conference on Human-computer interaction. (pp. 53-60).
[4] Biru, T. 2008. Reflective Steps: A Collaborative Learning ACM.
Oriented Approach to Software Development and Process
Improvement. Doctoral Thesis. Universität Hamburg. [20] Kujala, S. 2003. User involvement: a review of the benefits
and challenges. Behav Inform Technol. 22(1), 1-16.
[5] Blomkvist, S. 2005. Towards a model for bridging Agile
development and user-centered design. In Human-Centered [21] McInerney, P. and Maurer, F. 2005. UCD in Agile Projects:
Software Engineering—Integrating Usability in the Software Dream Team or Odd Couple? Interactions. 12(6), 19-23.
Development Lifecycle. pp. 219-244, Springer Netherlands. [22] Misra, S. Kumar, V. and Kumar, U. 2009. Identifying some
[6] Boivie, I. Aborg, C. Persson, J. and Lofberg, M. 2003. Why important success factors in adopting agile software
usability gets lost or usability in in-house software development practices. The Journal of Systems and Software.
development. Interacting with Computers 15(5): 623–639. 82 (2009). 1869–1890, Elsevier.
[7] Brhel, M. Meth, H. Maedche, A. and Werder, K. 2015. [23] Nebe, K. Zimmermann, D. and Paelke, V. 2008. Integrating
Exploring principles of user-centered agile software Software Engineering and Usability Engineering. in
development: A literature review. Information and Software Advances in Human Computer Interaction. Pinder S. InTech.
Technology. 61 (2015) 163–181, Elsevier B.V. 2008, PP. 331-350.
[8] Chamberlain, S. Sharp, H. and Maiden, N. 2006. Towards a [24] Nielsen, J. 1993. Usability Engineering. Academic Press.
framework for integrating Agile development and user- [25] Pade-Khene, C. Mallinson, B. and Sewry, D. 2011.
centered design, In Extreme Programming and Agile Sustainable rural ICT project management practice for
Processes in Software Engineering, pp. 143-153, Springer developing countries: investigating the Dwesa and RUMEP
Berlin Heidelberg. projects, Information Technology for Development, 17:3,
[9] Chow, T. and Cao, D. 2007. A survey study of critical 187-212, DOI: 10.1080/02681102.2011.568222, Rouledge
success factors in agile software projects. The Journal of Taylor & Francis group.
Systems and Software, 81(6), 961-971. [26] Salah, D, Paige, R, and Cairns, P. 2014. A systematic
[10] Cockburn, A. 2002. Agile Software Development, Addison- literature review for agile development processes and user
Wesley. centered design integration. In Proceedings of the 18th
International Conference on Evaluation and Assessment in
[11] Dingsoyr, T, Nerur, S, Balijepally, V, and Moe, N. 2012. A Software Engineering. (p. 5). ACM.
decade of agile methodologies: Towards explaining agile
software development, Journal of Systems and Software. [27] Schwaber, K. and Beedle, M. 2002. Agile Software
85(6), 1213-1221. Development with Scrum. Prentice Hall.
[12] Dino, H. 2012. A Framework for Integrating Software [28] Sharrock, W. and Anderson, B. 1994. The user as a scenic
Usability Into Software Development Process. Masters feature of the design space. Design Studies. 15(1), 5-18.
Thesis, HiLCoE School of Computer Science & Technology. [29] Silva, S. T. Martin, A. Maurer, F. and Silveira, M. 2011.
[13] Dittrich, Y. and Lindeberg, O. 2003. How use–oriented User-Centered Design and Agile Methods: A Systematic
development can take place. Information and Software Review. In AGILE(pp. 77-86).
Technology 46.9 (2004): 603-617, Elsevier. [30] Smith, A. Bannon, L. and Gulliksen, J. 2010. Localising HCI
[14] Dittrich, Y. Rönkkö, K. Eriksson, J. Hansson, C. and practice for local needs. Proceedings of the 2010
Lindeberg. O. 2008. Cooperative method development. international conference on Interaction Design &
Empirical Software Engineering. 13(3), 231-260. International Development. British Computer Society.
[15] Doerflinger, J. and Dearden, A. 2013. Evolving a software [31] Winschiers-Theophilus, H. 2009. Cultural Appropriation of
development methodology for commercial ICTD projects. Software Design and Evaluation. In: Whitworth, Handbook
Information Technology and International Development. 9 of Research on Socio-Technical Design and Social
(3), 43-60. Networking Systems. IGI Global
[16] Dyba, T. and Dingsøyr T. 2008. Empirical studies of agile [32] Zewge, A. Dittrich, Y. and Bekele, R. 2014. Software
software development: A systematic review, Inform. Softw. designing methodology for ICT4D domain. Proceedings of
Technol., doi:10.1016/j.infsof.2008.01.006. the 13th Participatory Design Conference: Short Papers,
Industry Cases, Workshop Descriptions, Doctoral
[17] Goransson, B. Gulliksen, J. and Boivie, I. 2003. The Consortium papers, and Keynote abstracts-Volume 2. ACM.
Usability Design Process – Integrating User-centered

120

View publication stats