=Paper=
{{Paper
|id=Vol-2490/paper4
|storemode=property
|title=Using i* to Reason about Employee Behavior on Public Social Media
|pdfUrl=https://ceur-ws.org/Vol-2490/paper4.pdf
|volume=Vol-2490
|authors=Vik Pant,Eric Yu
|dblpUrl=https://dblp.org/rec/conf/istar/PantY19
}}
==Using i* to Reason about Employee Behavior on Public Social Media==
https://ceur-ws.org/Vol-2490/paper4.pdf
Using i* to reason about Employee Behavior on Public
Social Media
Vik Pant1 and Eric Yu1, 2
1
Faculty of Information, University of Toronto, Toronto, Canada
2
Department of Computer Science, University of Toronto, Toronto, Canada
vik.pant@mail.utoronto.ca
eric.yu@utoronto.ca
Abstract. The use of social media by employees can be both beneficial and det-
rimental for their employer. Favorable posts by employees can bolster an organ-
ization’s respectability while criticism can undermine its reputation. Many em-
ployers adopt policies to influence social media activities of their employees.
Such policies encourage behavior that is favorable to the organization and dis-
courage conduct that is damaging. Employees must make tradeoffs between
their professional objectives and personal aspirations while using social media
because the two sets of goals may conflict. This paper demonstrates the applica-
tion of i* modeling to develop an understanding of employee behavior on public
social media. A hypothetical industrial scenario drawn from scholarly literature
and professional press is used to explain this approach. i* modeling can be used
to support employers in formulating effective policies for influencing employee
behaviors on public social media. It can also be used to promote a shared under-
standing that can help to guide employee participation on public social media in
a manner that balances their professional and personal objectives.
Keywords: Social Media, Visualization, Information Representation, Visual
Modeling, User Behavior.
1 Introduction
The use of public social media (SM), including Facebook, Twitter and LinkedIn, is common-
place in organizational settings [1,2]. Public SM offers organizations with channels of focused
and targeted communication for interacting with their external stakeholders (e.g., customers,
partners, suppliers, vendors) [3]. Organizations leverage public SM in many functional areas
including marketing, sales, customer service, and recruitment. Organizations permit their em-
ployees to act on the employer’s behalf for posting content that is relevant for external stake-
holders. Moreover, many employees also share their opinions and commentary about their
employers through their personal accounts on public SM.
While the use of public SM by employees can be beneficial for an organization – it can
also expose their employers to risks and uncertainties [4,5]. This is because an employee may
intentionally or inadvertently post content on public SM that may be damaging to their employ-
er. When an organization is negatively impacted because of such deleterious behavior by its
own employee then it is likely to take punitive measures. This paper demonstrates the applica-
tion of i* modeling to support the comprehension of employee behavior on public SM. A hypo-
thetical industrial scenario drawn from scholarly literature and professional press is used to
explain this approach. This approach can be helpful for the employee and the employer to un-
derstand each other’s motivations.
2 Case Example: Social Media Participation by Employees
The role of employees in many departments, including Marketing, Customer Service, and Re-
cruitment, entails interacting with external stakeholders such as vendors, customers, and job
applicants. In the context of their job, an employee may expect public SM participation to lead
to career progression. Outside of work, many people also use their personal SM accounts to
Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons
License Attribution 4.0 International (CC BY 4.0).
share their opinions on public fora. In the context of personal use, a user may intend for their
participation on public SM to support the establishment of their reputation for independence.
In many cases the aims associated with job-related use and personal SM use may not
be compatible and may even conflict with one another. The behavior of an employee on public
SM reflects their choices to accommodate such tradeoffs which are inherent in their objectives.
The trade-offs in an employee’s objectives emerge due to their relationships with other actors
as well as the relationships among those actors. Each actor pursues its own interests and the
interests of an actor may not be compatible with the interests of other actors. For example, if an
employee openly lambastes their employer on SM then those posts may be accessed by the
competitors of their employer. Rivals monitor public SM to mine competitive intelligence (CI)
[4-8] that is useful for furthering their own interests at the expense of the employer (e.g., by
poaching its customers).
An employer may tie the SM behavior of employees with their performance apprais-
als to prevent damaging outcomes such as loss of customers. That employer might use a car-
rots-and-sticks policy to influence the behavior of its employees on SM. Such policies encour-
age employees to post favorable content about their employers on SM services and discourages
them from posting damaging content. Posting of favorable content might be rewarded with an
increase in the performance rating of an employee while posting of damaging content might be
penalized with reduction in the performance rating. However, the employer must be careful
because, while rewards will increase employee goodwill, penalties will decrease it. Reduced
employee goodwill may result in additional damaging content from disgruntled employees
which may repel other stakeholders such as customers, vendors, and job applicants.
Employee behavior on SM will also impact their own professional and personal
goals. If an employee engages in frank SM behavior and posts content that is critical of their
employer then that employee may succeed in establishing their reputation for independence.
However, that employee will then forego the opportunity for career advancement as their per-
formance rating will be reduced by their employer since they posted damaging content about
that employer. Conversely, if an employee engages in restrained SM behavior and refrains from
posting any content about their employer or only posts uncritical content then that employee
may succeed in advancing their career. However, this will impede that employee from estab-
lishing a reputation for independence as they will be regarded as being evasive towards contro-
versial topics related to their employer or unceasingly sycophantic towards that employer.
As shown by this example, employee participation on public SM can have beneficial
and deleterious effects for employers and employees. Both parties face nontrivial choices that
are characterized by complicated tradeoffs relating to employee participation on public SM.
Therefore, employees and employers can benefit from a methodical approach for understanding
each other's intentions. Such an approach can help employees to predict the consequences of
their SM behaviors. It can also support employers to formulate an effective policy that links
performance appraisal with employee SM behaviors.
A shared understanding on the part of employers and employees can help to promote
the interests of both parties while also accounting for the objectives of other stakeholders with
whom they have relationships. Employee participation on public SM in an ad hoc or unmethod-
ical manner may neglect or underemphasize important aspects of relationships that employers
have with their external stakeholders. It might also yield a performance appraisal policy that
omits or overlooks crucial tradeoffs made by employees when they use public SM. Such mis-
takes and errors are likely to impair or block the achievement of the objectives of employers
and employees. The next section applies a structured and systematic approach to support a
shared understanding of employee behavior on public SM for the employer and the employee.
3 Using i* to reason about Employee Behavior on Public Social
Media
A crucial requirement for understanding employee behavior on public SM is the ability to ar-
ticulate and analyze relationships among various stakeholders such as employee, employer,
social media service, etc. In this paper, we use i* to represent this phenomenon because i*
treats actors and goals as first class entities [9]. Fig. 1 presents an i* SD (Strategic Dependen-
cy) diagram of the case example in section 2. To streamline description of the model in the
following text, instances and classes of i* entities are written as: instance (class).
Continuing
business
List of
differentiators of
Vendor
Analytics
about self
Vendor Customer
Access fees
for Analytics Content about
about self firms
User traffic
List of
employee SM
List of
accounts Contribution be differentiators of
recognized Social Media Competitor
Service
New
Effective social
business
platform
Competitor
Analytics
User
Competitor
generated
Analytics
content
access fees
Employee
Competitor
Legend
Actor
Goal Softgoal Task Resource
Actor Actor Boundary Goal Softgoal Task Resource
Task
Help Contribution Hurt Contribution Decomposition
Means-End Dependency Link
Link Link Link
Link
Fig. 1. i* SD diagram showing relationships pertaining to employee SM participation
This diagram depicts five actors, namely Employee, Vendor, Social Media Service, Customer,
and Competitor. The Employee (actor) depends on the Social Media Service (actor) for an
“effective social platform” (softgoal) while the Social Media Service depends on the Employee
for “user generated content” (resource). Similarly, the Employee (actor) depends on their em-
ployer, (i.e, Vendor (actor)), for SM participation to be acknowledged (“contribution be recog-
nized” (softgoal)) while the Vendor (actor) depends on the Employee (actor) for a “list of
employee SM accounts” (resource). This is necessary for the Vendor (actor) to attribute contri-
butions by SM posts of that Employee (actor).
Fig. 2 presents an i* SR (Strategic Rationale) diagram of the case example. The Em-
ployee (actor) intends to interact with stakeholders of their employer and, in this diagram, this
is depicted as “stakeholders be engaged” (goal) within Employee (actor). To satisfy “stake-
holders be engaged” (goal) an Employee (actor) may perform “reach out to stakeholders via
social media” (task). To complete “reach out to stakeholders via social media” (task) an Em-
ployee (actor) will need to satisfy “professional opinions be expressed” (sub-goal). This is
because customers may be interested solely in the professional opinions of their business con-
tacts and not in their personal opinions (e.g., related to sports, politics, or religion).
An Employee (actor) can choose among two options for satisfying this requirement
which include either selecting “share employer-related topics” (task) or “avoid employer-
related topics” (task). “Share employer- related topics” (task) is comprised of “stance be cho-
sen” (sub-goal) which refers to selection of a public voice by the Employee (actor) vis-à-vis
their posts on SM about their employer. An Employee (actor) can choose among two options
for meeting this requirement which include either selecting “include criticism” (task) or “ex-
clude criticism” (task). That Employee (actor) can compare these options by assessing the
impact of each option on their quality criteria (i.e., softgoals).
Run Promote service
business sold by self
Continuing Service
Analyze own business be used
presence on social
media Procure Reputation
service of vendor
Favorable Inspect Damaging
List of be superior
content be content about content be
differentiators of Select
encouraged self discouraged Contract
Vendor respectable
be offered
vendor
Appraise Highlight
content by favorable Analytics Service
employees content about self disruption
Continue with be avoided
Reward Penalize Incumbent
praise with complaints Content about
bonus with demerit Access fees Switch to
firms
for Analytics Competitor
Goodwill be about self
earned Vendor User traffic
Increase Reduce
Customer
performance performance
rating rating List of
differentiators of
Competitor
List of New
employee Social Media business
SM accounts
Contribution Effective social Service
be platform
recognized
User
generated
content
Employee
Stakeholders
be engaged
Competitor
Analytics Win
Reputation for
Career be business
independence
advanced Competitor
be established
Analytics Incumbent
Reach out to
access fees vendor be
stakeholders via
Freedom of replaced
social media Appraisal be Analyze rival
expression be
excercised positive presence on social
media
Professional opinions
be expressed
Frank Restrained Highlight
behaviour disadvantages of Favorable
behaviour
service by rival content about
rival be excluded
Inspect damaging
content about rival
Share employer- Avoid employer- Content from
related topics related topics employees of
rivals be shown
Stance be
chosen Competitor
Include Exclude
criticism criticism
Fig. 2. i* SR diagram showing distributed intentionality about employee SM participation
An Employee (actor) may “avoid employer-related topics” (task) which may help
that Employee to exhibit “restrained behavior” (softgoal) rather than “frank behavior” (soft-
goal). Conversely, that Employee may “share employer-related topics” (task) which may help
that Employee (actor) to demonstrate “frank behavior” (softgoal) as opposed to “restrained
behavior” (softgoal). By themselves these quality criteria may not provide a substantial basis
for choosing among options but their hierarchies can be used to discriminate between alterna-
tives.
“Restrained behavior” (softgoal) is likely to result in an Employee’s (actor) SM per-
formance “appraisal be positive” (softgoal) which is likely to support that Employee’s (actor)
professional growth (“career be advanced” (softgoal)). Similarly, “frank behavior” (softgoal) is
likely to support that Employee’s (actor) “freedom of expression be exercised” (softgoal)
which is likely to bolster that Employee’s (actor) “reputation for independence be established”
(softgoal). The prospect of career advancement and reputation building, reflected by higher
level quality criteria, are sufficient for comparing options.
The decision by an Employee (actor) to pursue either of these courses of action im-
pacts and is impacted by the decisions of their Employer (i.e., Vendor (actor)). The Vendor
(actor) has an existing relationship with its customers (i.e., Customer (actor)) that depend on it
for “continuing business” (softgoal). To satisfy this dependency from Customer (actor) to itself
the Vendor (actor) must “run business” (task). While performing this activity the Vendor (ac-
tor) will need to “promote service sold by self” (sub-goal) to retain existing Customer (actor).
This can be fulfilled by “analyzing own presence on social media” (task) because
self-awareness about social media content can help to stave off incursions into the customer
base by rivals (i.e., Competitor (actor)). Analysis of own SM presence requires a Vendor (ac-
tor) to access a report on “analytics about self” (resource) from Social Media Service (actor).
As payment for this report the Vendor (actor) pays “access fees for analytics about self” (re-
source) to the Social Media Service (actor). “Analyzing own presence on social media” (task)
will require the Vendor (actor) to “inspect content about self” (sub-task) which includes exam-
ining sentiment and tonality of SM content about oneself.
This will require the performance of “highlight favorable content” (sub-task) and
“appraise content by employees” (sub-task). The former is necessary for providing a “list of
differentiators of vendor” (resource) to Customer (actor) so that Customer (actor) decides to
maintain business relationship (“continuing business” (softgoal)) with the Vendor (actor). The
latter is essential for ensuring that “favorable content be encouraged” (softgoal) and “damaging
content be discouraged” (softgoal) are satisficed. “Reward praise with bonus” (sub-task) and
“penalize complaints with demerit” (sub-task) make positive contributions to these quality
criteria. “Reward praise with bonus” (sub-task) is further refined into “increase performance
rating” (sub-task) while “penalize complaints with demerit” (sub-task) is further refined into
“reduce performance rating” (sub-task).
However, the Employer (actor) also has quality criteria related to the goodwill of its
employees (("goodwill be earned" (softgoal)). Therefore, it must be careful because, while
“reward praise with bonus” (task) will help "goodwill be earned" (softgoal), “penalize com-
plaints with demerit” (task) will hurt it. This may result in additional damaging content on SM
from disgruntled employees, greater employee dissatisfaction leading to attrition, and avoid-
ance of employer from various stakeholders including customers, vendors, and job applicants.
These outcomes are not shown in the model as they are tangential to this case example.
The Vendor (actor) rewards or penalizes favorable or damaging SM posts by Em-
ployee (actor) because SM content by Employee (actor) impacts Vendor’s (actor) relationships
with its existing Customer (actor). Existing Customer (actor) must “procure service” (task) to
meet some requirements (i.e., “service be used” (goal)). To do so the Customer (actor) must
“select respectable vendor” (task) which is defined as “reputation of vendor be superior" (soft-
goal) among available vendors. The Customer (actor) performs due diligence and obtains in-
puts from multiple sources to determine the respectability of available vendors. It examines
“content about firms” (resource) directly on Social Media Service (actor) by itself and in the
process generates “user traffic” (resource) for those SM services. The Customer (actor) also
instructs the incumbent Vendor (actor) to supply a “list of differentiators of Vendor” (resource)
and solicits a “list of differentiators of Competitor” (resource) from Competitor (actor).
Procuring a service from the incumbent Vendor (actor) or any Competitor (actor) re-
quires “contract be offered” (goal) to the relevant actor. The options for this that are available
to the Customer (actor) include “continue with incumbent” (task) or “switch to competitor”
(task). A Customer (actor) factors into this decision its intention for “service disruption be
avoided” (softgoal) which is helped by “continue with incumbent” (task) and hurt by “switch to
competitor” (actor). However, this intention is subordinate to the higher-level intention of
“reputation of vendor be superior” (softgoal) or otherwise the Customer (actor) would always
“continue with incumbent” (task) and not consider “switch to competitor” (task) as a viable
option.
The opportunity to poach the Customer (actor) of an incumbent Vendor (actor) serves
as motivation for Competitor (actor) to remain vigilant on SM. A Competitor (actor) can “win
business” (actor) by enabling an “incumbent vendor be replaced” (sub-goal) by scanning SM
(“analyze rival presence on social media” (task)). Analysis of competitive intelligence on SM
requires a Competitor (actor) to access a “competitor analytics” (resource) report on Social
Media Service (actor) and as payment for this report the Competitor pays “competitor analytics
access fees” (resource) to the Social Media Service (actor). This report is useful for responding
to Customer (actor) solicitation of “list of differentiators of Competitor” (resource). Based on
this report the Competitor (actor) can “highlight disadvantages of service by rival” (sub-task).
In doing so the Competitor (actor) uses the criteria of “favorable content about rival be exclud-
ed” (softgoal). Moreover, the Competitor (actor) can “inspect damaging content about rival”
(sub-task) to prioritize “content from employees of rivals be shown” (softgoal). This can serve
to highlight weaknesses of the incumbent Vendor (actor) directly from an authoritative source
that knows the Vendor (actor) – which is complaints and criticisms by its Employee (actor).
4 Related Work
The research presented in this paper is related to a rich body of academic literature that pertains
to the modeling of business strategy. In [10-12] we show models of strategic pivoting by
startups and large enterprises. In [13, 14] we outline the requirements for modeling and analyz-
ing strategic coopetition (i.e., simultaneous cooperation and competition) among organizations.
In [15-17] we show the design and analysis of a win-win strategy in a multi-actor setting using
a combination of i* strategic actor modeling with game-theoretic techniques. In [18] we show
the expression and evaluation of strategies, based on the notion of value, by combining i* and
e3Value modeling. The application of i* modeling to focus on employee behavior, as a specific
consideration in formulation of business strategy, are novel contributions of this paper.
5 Discussion and Conclusion
This paper demonstrated the application of i* modeling to develop an understanding of em-
ployee behavior on public social media. We showed that i* is conducive for representing and
reasoning about strategic relationships among stakeholders in the context of employee partici-
pation on public SM. Employers and employees can refer to the models presented in this paper
to construct analyses that are specific to their own contexts. Employers can use those artefacts
to design performance appraisal policies that factor in employee SM behavior. Employees can
use those artefacts to guide their decision-making vis-à-vis their participation on public SM.
We faced some limitations of i* in the process of developing models of this example. These
include the lack of support in i* for: temporal reasoning, depiction of conditionality, and ex-
pression of negative dependencies. These subjects are the focus on ongoing scholarship by
researchers of i* modeling including ourselves and will be addressed in our future work.
6 References
1. Barnes, N. G., & Jacobsen, S. (2013). Adoption of social media by fast-growing companies: Innovation
among the Inc. 500. Journal of Marketing Development and Competitiveness, 7(1), 11-17.
2. Meske, C., & Stieglitz, S. (2013, June). Adoption and use of social media in small and medium-sized
enterprises. In working conference on practice-driven research on enterprise transformation (61-75).
Springer, Berlin, Heidelberg.
3. Eyrich, N., Padman, M. L., & Sweetser, K. D. (2008). PR practitioners’ use of social media tools and
communication technology. Public relations review, 34(4), 412-414.
4. Vuori, V., & Väisänen, J. (2009, November). The use of social media in gathering and sharing competi-
tive intelligence. In 9th Internafional Conference on Electronic Business.
5. Dey, L., Haque, S. M., Khurdiya, A., & Shroff, G. (2011). Acquiring competitive intelligence from
social media. In Proceedings of the 2011 joint workshop on multilingual OCR and analytics for noisy
unstructured text data (p. 3). ACM.
6. Rothberg, H. N., & Erickson, G. S. (2002, January). Competitive capital: a fourth pillar of intellectual
capital?. In World congress on intellectual capital readings (94-103). Butterworth-Heinemann.
7. Jin, T., & Ju, B. (2014). The corporate information agency: Do competitive intelligence practitioners
utilize It?. Journal of the association for information science and technology, 65(3), 589-608.
8. Fan, W., & Gordon, M. D. (2014). The power of social media analytics. Communications of the ACM,
57(6), 74-81.
9. Yu, E. (2001). Agent orientation as a modelling paradigm. Wirtschaftsinformatik, 43(2), 123-132.
10. Pant, V., & Yu, E. (2018) Conceptual Modeling to Support Pivoting – An Example from Twitter. In:
Proceedings on Advances in Conceptual Modeling - ER 2018 Workshops AHA, MoBiD, MREBA, Onto-
Com, and QMMQ, Lecture Notes in Computer Science. Springer, Verlag.
11. Pant, V., & Yu, E. (2018). Conceptual Modeling to Support the “Larger Goal” Pivot–An Example from
Netflix. In: IFIP Working Conference on The Practice of Enterprise Modeling (pp. 394-403). Springer,
Cham.
12. Pant, V., Yu, E., & Tai, A. (2017, November). Towards Reasoning About Pivoting in Startups and
Large Enterprises with i*. In: IFIP Working Conference on The Practice of Enterprise Modeling (pp. 203-
220). Springer, Cham.
13. Pant, V., & Yu, E. (2016). Coopetition with Frenemies: Towards Modeling of Simultaneous Coopera-
tion and Competition among Enterprises. In: IFIP Working Conference on The Practice of Enterprise
Modeling (pp. 164-178). Springer, Cham.
14. Pant, V., & Yu, E. (2018). Modeling Simultaneous Cooperation and Competition Among Enterprises.
Business & Information Systems Engineering, 60(1), 39-54.
15. Pant, V., & Yu, E. (2018). Getting to Win-Win in Industrial Collaboration under Coopetition: A Strate-
gic Modeling Approach. In: Intl. Conference on Business Informatics Research. Springer, Cham.
16. Pant, V., & Yu, E. (2019). A Modeling Approach for Getting to Win-Win in Industrial Collaboration
under Strategic Coopetition. Complex Systems Informatics and Modeling Quarterly, (19), 19-41.
17. Pant, V., & Yu, E. (2018) Generating Win-Win Strategies for Software Businesses under Coopetition:
A Strategic Modeling Approach. In: Intl. Conference of Software Business. Springer, Cham.
18. Pant, V., & Yu, E. (2017) Modeling Strategic Complementarity and Synergistic Value Creation in
Coopetitive Relationships. In: Intl. Conference of Software Business (pp. 82-98). Springer, Cham.