SAP Usability in Bangladesh's Textile ERP
SAP Usability in Bangladesh's Textile ERP
in Bangladesh
Wai-Peng Wong
Universiti Sains Malaysia
Vito Veneziano
University of Hertfordshire
Imran Mahmud
Universiti Sains Malaysia
Abstract
Information technology (IT) is increasingly playing a crucial role in managing business processes across all
indus- tries and organizations but business managers and analysts seem to underestimate the impact of
usability of IT solutions on processes and people. This paper focuses on the usability of one of the most
popular business process management software systems, SAP, which comes as an integrated solution
that incorporates the key business functions and processes of an organization. The paper critically
analyzes related issues and implica- tions using the System Usability Scale (SUS) and semi-structured
interviews. Several considerations and sugges- tions are drawn in terms of rethinking and pursuing
usability in training when applied to Enterprise Resource Planning (ERP) and other systems
supporting business process management.
Keywords
enterprise resource planning, SAP, human computer interaction, information system management,
system usability, user-centered design, interface usability, Bangladesh
Lack of proper training and communicativeness of the SAP software continue to affect
its usability.
these techniques to improve business processes and Survey participants and sample size
operations. Table 1 provides a brief description of
each of these five criteria and some sources from The survey was conducted in a renowned garment
which the criterion is derived. manufacturing company in Bangladesh (henceforth
this company is identified as VG).
VG is one of the leading multi-dimensional busi-
ness conglomerates of Bangladesh, with spinning,
Research design, strategy and settings knitting, dyeing, accessories and printing facilities.
Goal and research questions It emerged as an established group in 2002. VG is
The primary research question of this paper is, ‘How now focused on further developing their product
usable is the SAP ERP software from the end user’s range by diversifying their business into different
perspective?’ sectors. The company expects the present turnover to
However, a number of other objectives will also double in 3 years’ time riding on a 30% yearly
be addressed. They are: growth. VG could be considered a thriving enterprise
because it started as a fashion and garment
1. to identify whether there are still usability manufacturer and now has gained a foothold in the
problems in ERP logistics, agriculture, power and energy sectors.
2. to measure end users’ points of view on the The target population for our survey comprised
SAP interface 140 VG employees who use different modules of
SAP ERP on a daily basis and therefore have accrued
3. to identify and explain the gaps between ERP
significant experiences of the system. A System
frameworks suggested by Singh and Wesson
Usability Scale questionnaire was distributed to 127
(2009) and the System Usability Scale on
(83% male, 17% female) users from 140 who had
ERP.
been using the ERP software for at least 6 months
and 24 participants were randomly chosen from these
We first investigated and carefully compared sev- 127 users for interview. By also considering the
eral appropriate qualitative and quantitative tech- outcomes and attributes of user modeling activities
niques for a mixed approach before adopting them and procedures, represented by a specimen result in
to evaluate ERP user interfaces and report on the Table 2, we obtained a predictive evaluation of real-
usability of SAP ERP in a manufacturing enterprise. world tasks aimed at capturing some aspects of
We explored possible data collection sources and users’ understanding, knowledge and inten- tions,
con- sidered research constraints when pursuing and their actual ability to process information using
several research strategies which comprise surveys their ERP modules.
and inter- views. Surveys, in particular, were Following Kules (2004), we derived from the ERP
deployed to dis- cover usability issues perceived by system interface and specifications a user model con-
ERP consulting organizations and ERP specialists. taining ‘‘all information that the system knows about
Table 2. List of attributes derived for user These research methodologies are mostly
modeling purposes (A specimen). grounded on cognitive psychology and ergonomics,
User class: SAP user Attribute and have been developed on the assumption that
cognition is a distributed process, occurring not just
Age Range 25-35 within the human mind, but also between the human
English literacy Average mind and any repre- sentational instance of
Computer knowledge Average information provided in the world. We implemented
Training Completed
convergent mixed method suggested by Creswell and
Education Diploma and bachelor
degree holders Clark (2007), as two differ- ent approaches could
Other system used MS Office Package, result in a new paradigm of usability. The mixed
Web Goal Operate SAP modules method approach to usability test- ing is also applied
in several researches of Scholtz et al., (2010a,
2013b) and Thavapragasam (2004).
the user’’. This model, usually initialized either with
default values or by querying the user, is maintained Semi structured interview and qualitative content
by and within the system, and users are allowed to analysis. Semi-structured interviews were adopted to
review and amend it by editing their user profile in under- stand the users’ expectations, goals and
the demographic information (see Table 2). problems. Before selection, participants who were
VG invests heavily in SAP and they are pleased to willing to take part were asked to complete a short
recognize that with the implementation of SAP, the questionnaire about their name, job title, years of
company has reached a significant milestone. Specif- experience in using the SAP ERP system and three
ically, VG is using eight different modules of SAP: frequently used functions of the software to ensure
SD (Sales and Distribution), PP (Production Plan- we had a represen- tative end user base.
ning), QM (Quality Management), PM (Plant Qualitative data analysis of the responses from the
Mainte- nance), HR (Human Resources), FI open ended usability questionnaires was performed
(Financial Accounting), CO (Controlling) and MM using content analysis, whereby the text is
(Material Management). Based on VG’s categorized or coded into themes or categories
achievement, we assume that software users at VG (Kolbe and Bur- nett, 1991). Noting Singh and
have played a cru- cial role in the company’s Wesson’s (2009) use of qualitative analysis and their
expansion and consolida- tion plans. suggestion that no validated questionnaire was
From our investigation, we assumed that SAP needed for quantitative analysis, we followed their
HCI experts had developed user interfaces for their approach. The text responses were analyzed and
ERP system and considered similar criteria (age, coded thematically according to the five usability
computer literacy, English literacy, familiarity with criteria of navigation, presentation, task support,
other IT systems)when profiling users. learnability, and customi- zation suggested by Singh
and Wesson (2009) and Scholtz et al. (2009a,
2013b). Within these themes, the content was
Set up and protocol
analyzed in greater detail and further coded into sub-
As stated above, the two main instruments for collect- themes (see Table 3.)
ing data in our study were a round of interviews using
an open ended questionnaire and a survey question- System Usability Scale (SUS). The SUS has several
naire for end users. We applied a convergent mixed attri- butes which make it a popular choice for general
method to identify any new dimensions of usability. usability practitioners. According to Bangor et al.,
The first instrument was a semi-structured interview, (2008), SUS assessment has a reliability value of
mainly developed by referring implicitly to the ‘usabil- 0.85, which is relatively high.
ity criteria’ adopted in Singh and Wesson (2009). Brooke’s (1996) starting point was usability,
Following Perlman (1997), ‘‘Questionnaires have defined by ISO 9241-11 as a combination of
long been used to evaluate user interfaces’’, our effective- ness, efficiency and satisfaction.
ques- tionnaire instrument then uses the System Unfortunately, since these ‘‘classes of metric can
Usability Scale (Brooke, 1996) which has its vary widely’’ (Brooke, 1996), he employed a new
reliability already established. and simpler scale which does not evaluate the system
on individual dimensions of effectiveness, efficiency
and satisfaction, but rather scores the overall
subjective experience of the user
Table 3. Code and sub-themes. Data analysis and findings
Items Sub items Qualitative data analysis
Navigation Ease of use As mentioned before, we applied content analysis to
Helpful navigation the interview outcomes of SAP users.
Shortcut key
Search option Navigation. The lessons learned about navigation
Like/Dislike from the interviews are the positive comments and
Presentation Font style
sugges- tions for improvements. The answer from
Language
Lay out
one of our participants summarizes what participants
Like/Dislike thought of navigation in SAP:
Task support Documentation
Productivity ‘‘It is very swift to work with and it takes only a few
Time to get sec- onds if there is a need to retrieve huge data. It is
support easy to access any function from the initial screen by
Learnability Understandability using T- codes. Overall navigation is excellent and
Time to learn easily understandable.’’
Customization Customize
layout Flexibility However another participant suggested that
‘‘Less unnecessary data and more proper training as ‘most usable’. A partial result is reproduced in
could improve my skills.’’ Table 4.
We take note that the SUS scores are not for indi-
Learnability. Responses from the interviews on vidual items, which are less meaningful on their own.
under- standability and intuitiveness to use are within Instead, the SUS scores are a composite measure of
an acceptable range. Two participants emphasized overall usability.
how previous knowledge of computers, IT skills and The reliability score (Cronbach’s Alpha) of 0.819
train- ing have helped new users gain confidence is greater than the recommended 0.70 (Nunnally and
with the system: Bernstein, 1994), indicating that all items possess
good reliability. The mean score of 58.97, relatively
‘‘For me getting confident with the system took about lower than the acceptable score of 65.0, is also statis-
one week, and it is that easy because of the course I
tically significant (t 5.014,
¼ p <.001).
attended during which I obtained my ORACLE
certification.’’ A t-test was used—to measure the significance of
‘‘If you have basic computer knowledge it should the usability responses of the SUS instrument, similar
not take longer than two weeks.’’ to Bangor (2008). The t-test value of -5.014 shows
sig- nificant evidence of poor usability (see Table 5).
Customization. When the participants were asked The results also show that ‘unacceptable’ usability
about flexibility and ease of customization of user ratings were recorded by 59.8% of total participants
interface, most gave favourable responses. One (see Table 6).
respondent clarified that The highest variations were identified in Q4 (SD ¼
0.98) and Q10 (SD ¼ 0.94), which reflect the assis-
‘‘Yes, I often customize the screen according to my tance of technical personnel (Q4) and learning (Q10),
needs. Yes, the system is very flexible to interact with and we noted that these results were similarly identi-
because it allows the updating, duplication and high- fied in the interviews. Some issues from the mixed
lighting of required fields or the deletion of data when response results will be rationalized with evidence
necessary.’’ from other researchers in the following section.
These interview results guided us in our
questionnaire survey of all end users to get a broader Discussion
picture of ERP usability for a mixed response The software quality model, FCM (Factor Criteria
approach in evaluating learnability and Metrics), proposed by McCall et al. (1977), focuses
supportability. Therefore, we sur- veyed all VG SAP on three criteria which are applied only to the
ERP users using the System Usability Scale usability factor found in software quality character-
instrument. istics. The three criteria are: operability, training
and communicativeness.
Operability is associated with the user’s effort for
Quantitative Data Analysis operations and operations control (for example
In using the SUS questionnaire, we are able to derive mouse support and macro-commands) (Stefani and
more rigorous results from quantitative data analysis. Xenos, 2001), which is applicable in our SAP usabil-
The 5-point Likert scale, upon which the tool was ity testing. Training is associated with the effort
designed and built, allowed us to eventually compute required to teach the use of software to the user,
a final score ranging between 0 as ‘least usable’ to
100
Table 5. Overall System Usability Scale (SUS) score chart.
Question
No Detailed Question Mean Value Standard Deviation
Q1 I think that I would like to use this software frequently 2.74 0.90
Q2 I found the system unnecessarily complex 2.44 0.86
Q3 I thought the software was easy to use 2.22 0.87
Q4 I think that I would need the support of a technical 1.88 0.98
person
to be able to use this system.
Q5 I found the various functions in the system were well 2.70 0.83
integrated.
Q6 I thought there was too much inconsistency in this 2.42 0.82
system
Q7 I imagine that most people would learn to use this 2.18 0.89
system
very quickly
Q8 I found the system very awkward to use. 2.48 0.87
Q9 I felt very confident using the system. 2.48 0.78
Q10 I needed to learn a lot of things before I could get going 2.01 0.94
with this system.
Table 6. Frequency distribution System Usability The first theme is concerned with the development
Scale (SUS). and introduction of a usability-oriented method
Usability status Frequency Percentage related to how users participate in eliciting and speci-
fying requirements, with an explicit focus on how to
Not Acceptable 76 59.8 fully exploit the skills of technical communicators
Acceptable 45 35.4 and other stakeholders.
Excellent 6 4.7 The second theme is about usability improvement
Total 127 100
by means of user training. This is a reinterpretation
of the second theme originally suggested by Carl-
while communicativeness is associated with the shamre (2001),), who argued that usability considera-
effectiveness of the software to communicate to the tions should drive planning in software product
user the purpose for which it has been developed and development, indicating that planning of the
the method of using it. develop- ment methodology should be designed by
The remaining two factors of FCM, training and usability criteria of a specific information system.
communicativeness, have yet to be explored. User training should be prioritized when planning for
Our interview results and overall usability as mea- the educa- tional needs of individual users (Esfandi,
sured by the SUS questionnaires are somehow 2010). SAP does not go live as a complete software
contra- dictory. This provides scope to consider and package; it was implemented module wise in
apply two main investigation themes on SAP ERP different depart- ments . As a result, we suggested to
usability inspired by the work of Carlshamre (2001). create a more structured training plan for the
organization.
User participation during development agree with Badamgarav (2010), who stated that ‘‘user
We adopt user participation in the first theme from involvement during system development is thought
Har- ris and Weistroffer (2008). User participation is to lead to greater user commitment, user acceptance,
referred to as observable behaviors and activities that behavioral intention, usage, and satisfaction with the
the target users or their representatives perform in system’’ (Amoako-Gyampah and White, 1993; Hart-
the system development process (Barki and wick and Barki, 1994).
Hartwick, 1989; Hwang and Thorn, 1999; Lin and Hartwick and Barki (1994) explicitly referred to
Shao, 2000), as opposed to the psychological three dimensions of user participation for successful
constructs of user involve- ment and user attitudes software development, which are identified and
(Lin and Shao, 2000), which are vague, not visible, designed from the developers’ viewpoint and for
and empirically challenging. their benefit. They are:
It is clear that none of our users have participated
1. Responsibility (i.e., the performance of activi-
in the development of SAP ERP, nor have they had
ties and assignments reflecting overall leader-
any a posteriori exposure to the development
ship or accountability for the project).
process.
2. User-IS relationship (i.e., the performance of
Several authors (Hartwick and Barki, 1994; Kap-
development activities reflecting users’ formal
pelman, 1995) agree that participation and involve-
review, evaluation, and approval of work done
ment are important for the success of information
by the information system (IS) staff, and
systems implementation (for example, in terms of
3. Hands-On-Activity (i.e., the performance of
users’ satisfaction). SAP is a proprietary software
specific physical design and implementation
and it operates in Bangladesh as a commercial off-
tasks).
the- shelf product; as a result, it is clear that users in
VG were not involved – not even in ‘simulation’ – in
the elicitation and specification process of the ERP
mechanism and the overall structure. Training structure
ERP vendors do not collect requirements from Our interpretation of the second theme originally
individual customers. While it is safe to assume that sug- gested by Carlshamre (2001), is about
the SAP developers have properly and legitimately improving usability of an ERP project through user
interviewed customers to define their requirements training. This is considered as a dual process
so as to provide clarity and understanding in the occurring to and from the user during the
development of the software, customers who use the implementation process of the sys- tem. Therefore,
SAP ERP system worldwide might not have bene- the user is not seen as a merely pas- sive recipient of
fitted from this process. In the best scenario, custom- the training.
ers appear to work on requirements on a largely In spite of the fact that the SAP ERP was
‘review only’ mode. In the worst and most common developed with consideration of usability features
scenarios, system requirements might not have (navigation, presentation, learnability, supportability
crossed the minds of the customers. and customi- zation) as proposed by Singh and
ERP practitioner Kimberling (2012) stated that it Wesson (2009), the feedback from our investigation
took nearly 20 years or more for large companies to frequently showed that the lack of proper training
adopt ERP systems because of complexity and risk. and communicative- ness of the software continue to
Medium organizations are still learning about ERP affect its usability.
implementation phase. Research reports have shown how learnability
This could explain why, although the SAP is the (i.e., how easy it is for users to learn and master
best selling ERP software worldwide, it still receives system functions and are able to transfer this learning
a severe judgment in terms of poor usability and low to real situations) (Kushniruk et al., 2009) and
satisfaction by our operators at VG. additional general usability (i.e., a measure of how
Engaging users during system design activities not easy it is to use a system) have a close relationship.
only means obtaining valuable first-hand knowledge Some authors (Dix et al., 2003; Linja-aho, 2006;
about the actual use of a system for the benefit of Rogers et al., 2011) even consider learnability as one
developers (Kujala, 2003); it also means that these of the key compo- nents of overall system usability.
same users should be credited with the success of the Again, we argue that issues and problems in
software system they have helped to develop. We learnability are mostly due to the limited help and
support found in the ERP itself and would cause the
industry to expend enormously
for training after the system has been implemented. 4. Ensure that training includes the HCI of the
Researchers have identified that training for ERP system and the overall business processes and
costs ranged up to 30% of total project costs (Beatty procedures which are affected directly by the
and Williams, 2006). Failure of well-organized train- introduction of the new system. Users have
ing will lead to project failure (Brunjolfsson and Hitt, to map the system command language against
2000). the business process key words in order to
In considering the requirements for effective train- cre- ate a common dictionary for all members
ing, our recommendations would be: of the organization.
5. Ensure that users are committed to and
1. Plan training first and ensure initial training is
engaged in training. Employees must be made
provided alongside and not after the actual
to feel the relevance of the training needs to
implementation of the system. Ensure training
their jobs and their personal and professional
involvement is not limited to the final users,
satisfaction. They should also be able to
but extended to the developers too, in order
identify the objec- tives and benefits of the
for them to gain further insights on how the
training to facilitate their contribution to the
system is to be used and how it should be
ERP training process.
developed.
6. Customize the delivery of training on users’
2. Consider training as a communication process
skills and competencies. Training must be
with an open channel from the users to the
delivered to develop individual competencies.
developers, and ensure that users are made
In an organization, ERP training could be
fully aware of this, with the hope that users
delivered to several groups comprising man-
will start to feel themselves to be responsible
agerial personnel, key-users, end-users and
actors and not mere passive recipients of deci-
the trainers. Developers and other stakeholders
sions made by someone else. User-centered
who might benefit should also be trained even
design should be pursued by implementing
though they do not have hands-on interaction
an open communication channel.
with the new system.
3. Allocate a proper budget for training. The
7. Schedule training in parallel with
costs should be incorporated with total budget
implementa- tion. It is important to
for the software implementation especially if
synchronize appropriate and critical usability
it is a large organization intending to deploy
training during imple- mentation. This would
ERP. The awareness and training programmes
require a proper docu- mentation outline
should be extended to top management and
listing the training procedures and explaining
therefore additional costs must be imputed.
the interactive fea- tures that would support
(Esteves et al., 2002, Esteves and Boho
changes.
´rquez, 2007).
Appendix 1