Empirical Task Analysis of Data Protection Management and Its
Collaboration with Enterprise Architecture Management
Dominik Huth
a
, Michael Vilser, Gloria Bondel and Florian Matthes
Chair of Software Engineering for Business Information Systems, Department of Informatics,
Technical University of Munich, Boltzmannstr. 3, Garching, Germany
Keywords:
Data Protection Management, General Data Protection Regulation, GDPR, Enterprise Architecture
Management.
Abstract:
The General Data Protection Regulation has forced organizations worldwide to rethink their processing ac-
tivities of personal data. One of the key difficulties of ensuring GDPR compliance is the scope of the regu-
lation and its interdisciplinarity: Data protection management (DPM) has to address challenges on the legal,
business and technical level over the entire organization. Enterprise architecture management (EAM) is a
well-established discipline that follows a holistic approach to strategically develop the enterprise architecture,
consisting of people, processes, applications, and their interrelationships. Thus, DPM can be considered a
stakeholder in the EA management process. In this paper, we report on a survey with 38 data protection of-
ficers that investigates the main challenges for DPM, as well as the collaboration between DPM and EAM
during the implementation of the GDPR.
1 INTRODUCTION
The GDPR has entered into force in 2018, replacing
the previous EU directive from 1995 - a time when
less than 1% of the world population used internet
services (Miniwatts Marketing, 2019). Besides its up-
dated definitions and the extended data subject rights,
one of the key changes of the regulation are the dra-
matic fines (Christina Tikkinen-Piri et al., 2018). Re-
cent announcements by national data protection au-
thorities, e.g. e 50 for Google, GBP 183 million
British Airways or GBP 99 million for Marriott (CMS
Hasche Sigle, 2019), underline the need to maintain
compliance with the regulation.
However, according to a recent industry report,
less than half of privacy professionals consider their
organization to be “fully compliant” or “very com-
pliant” (Ernst & Young and International Associa-
tion of Privacy Professionals, 2019). A key problem
that is associated with the regulation is its enterprise-
wide scope and the interdisciplinarity that this brings
with it. While data protection management (DPM) is
most frequently conducted by legal functions (Inter-
national Association of Privacy Professionals, 2019),
the GDPR’s accountability principle creates a need to
a
https://orcid.org/0000-0003-2924-8598
work more closely with IT and business departments.
Enterprise architecture management (EAM) is a
holistic approach to strategically develop and align
the technical and business architecture of an enter-
prise to realize cost saving potentials and increase ef-
fectiveness (Farwick et al., 2013). Recent research
has stressed its applicability to security and privacy
topics as well (Larno et al., 2019; Burmeister et al.,
2019).
This work investigates the possible support of
EAM for the data protection management tasks. To
this end, we define the following research questions:
RQ1: Which DPM activities are necessary to
achieve GDPR compliance?
RQ2: What are the most severe problems when
conducting these activities?
RQ3: How are DPOs collaborating with EAM
and how do they evaluate the helpfulness of
EAM?
We first provide an overview about the existing lit-
erature that addresses GDPR tasks and concepts de-
scribing the interrelationship between GDPR tasks
and EAM in section 2. Section 3 describes our re-
search approach and section 4 presents the results ob-
tained from our research. We conclude and point to
future areas of work in section 5.
656
Huth, D., Vilser, M., Bondel, G. and Matthes, F.
Empirical Task Analysis of Data Protection Management and Its Collaboration with Enterprise Architecture Management.
DOI: 10.5220/0009342506560665
In Proceedings of the 22nd International Conference on Enterprise Information Systems (ICEIS 2020) - Volume 2, pages 656-665
ISBN: 978-989-758-423-7
Copyright
c
2020 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
2 RELATED WORK
2.1 Data Protection Tasks - Academic
Contributions
From the academic body of knowledge, we identified
a range of analytical and qualitative contributions.
An analytical comparison between the GDPR and
the previous Directive 95/46/EC was conducted by
(Christina Tikkinen-Piri et al., 2018). Based on the ar-
ticles of both the directive and regulation, the changes
between them are identified and the consequences for
organizations processing personal data derived. The
authors identify twelve implications for companies
that process personal data, of which we consider ten
as DPM tasks.
(Sirur et al., 2018) interviewed twelve cyber secu-
rity experts about the difficulties of implementing the
GDPR’s requirements. They show that these issues
are mainly based on the large scope of the regulation,
the enactment of its qualitative recommendations and
the necessity of mapping out the organizations’ com-
plex data networks.
The work of (Almeida Teixeira et al., 2019)
presents a literature review concerning the critical
success factors of GDPR implementation. They
found out that, among other factors, GDPR analysis,
risk identification, process documentation and train-
ing awareness are seen as enablers of GDPR imple-
mentation. On the other hand, interpretation of the
regulation, lack of practical guidelines and standard
procedures are some of the barriers identified in their
study.
2.2 Data Protection Tasks - Industry
Contributions
There exists a substantial amount of industry reports
on GDPR implementation, mainly by software ven-
dors or consulting agencies. These reports are often
focused on the need to act, with less detailed practical
recommendations. We only review a small selection
here.
A study that was published shortly before the
GDPR entered into force mentions the reengineering
of systems and processes as one of the main concerns
of respondents (CIPL and AvePoint, 2018). The au-
thors conclude that clarification on how to implement
various provisions are needed, among them the con-
ditions for processing, impact assessments and third
party processing agreements. According to another
global study, less than half of the respondents report
being compliant (Ernst & Young and International
Figure 1: Research Approach.
Association of Privacy Professionals, 2019). As the
top responsibilities of DPM, the study cites privacy
policies and internal trainings, higher than address-
ing compliance of single processing activities. In
turn, roles outside of the core DPM team are mostly
involved in data mapping and addressing process-
ing activities. Recent privacy regulation worldwide
seems to support a trend for single global data protec-
tion strategies (International Association of Privacy
Professionals, 2019). The most frequent operational
DPM task from this study are privacy policy changes,
followed by privacy impact assessments and records
of processing activites. An observation that seems
consistent with other studies is the limited occurrence
of data subject access requests.
2.3 Collaboration between EAM and
DPM
(Burmeister et al., 2019) derive a privacy-driven EA
metamodel from an analysis of the GDPR provi-
sions, which represents EA elements and their rela-
tionships across six layers. The authors conclude that
EA models are useful for achieving compliance with
the GDPR and suggest additional research on the in-
tegration of EA, privacy and security. One of the
DPM tasks that has been addressed in EA research is
the record of processing activities (RPA) (Koc¸ et al.,
2018; Huth et al., 2019). Other researchers use Archi-
Mate for modeling security principles, e.g. (Larno
et al., 2019).
3 RESEARCH APPROACH
The goal of this study is to derive and validate a set of
reccurring tasks in DPM on an organizational level. In
detail, we aim to rate each task in terms of complexity
and time consumption, discover frequently emerging
problems and evaluate the status quo of the collabo-
ration between DPM and EAM at this point in time.
We illustrate our research process in figure 1:
First we conducted an in-depth analysis of the
GDPR, as well as secondary literature from academia
and industry that is presented in section 2. The result
of this step was an initial list of DPM tasks.
In the second step, we created a questionnaire that
consists of four categories:
Empirical Task Analysis of Data Protection Management and Its Collaboration with Enterprise Architecture Management
657
Descriptive information about the participant
The list of tasks, for which we asked the partici-
pants to rate their complexity on a five-point Lik-
ert scale and to assess their time consumption rel-
ative to the participant’s total work time
Problems that arise in the context of addressing
the tasks
The participant’s position on collaboration with
EAM
We validated and developed our task list and question-
naire iteratively together with four different DPM ex-
perts. Three of the experts work as external DPO and
one works in DPM in a large organization. We always
incorporated the feedback before conducting another
interview and did not observe any major changes in
the last two iterations.
Fourth, we implemented the questionnaire in the
survey tool Questback
1
and distributed the survey in
data protection interest groups in professional net-
works, as well as among personal contacts of the au-
thors.
In the fifth and final step, we visualized and ana-
lyzed the survey data. We present the results in the
next section.
4 RESULTS
The survey was conducted from August 15th to
September 30th of 2019 and resulted in 38 complete
responses from data protection officers. 15 of the re-
spondents work as internal DPOs, while the remain-
ing 23 work as external DPOs. When grouping the
sample based on the collaboration with EAM, only
12 respondents already collaborate with EAM, leav-
ing 26 DPOs that did not collaborate with EAM at that
point in time. Figure 2 shows the (absolute) distribu-
tion of participants by company size and the propor-
tions of respondents who collaborate with EAM. The
collaboration between DPM and EAM is more pro-
nounced in larger organizations, reflecting the obser-
vation that EAM is a discipline that is more prevalent
there.
4.1 DPM Activities
As described before, we conducted a literature anal-
ysis of primary and secondary sources to identify a
valid list of DPM tasks. Choosing the right granular-
ity for these activities poses a challenge, since each of
the activities itself consists of multiple steps, which
1
https://www.questback.com/
Figure 2: Collaboration with EAM Depending on Company
Size (N=38).
are executed by different stakeholders. We opted to
define nine activities. Since we evaluated them with
four DPM experts we are confident that they repre-
sent a valid characterization of data protection man-
agement. Further, our questionnaire provided the op-
tion of adding further activities. Only one participant
used this option to point out certification mechanisms.
We argue that these can be included in the verification
of compliance and in conducting audits. Our final list
of is as follows:
Awareness-raising and Schooling within the
Organization: Informing individuals within a
company about the GDPR and its implications. It
is the responsibility of the DPO to organize train-
ing and provide information material to Data Con-
troller, Data Processor and Management.
Verifying Compliance of Existing Data Han-
dling Processes: Processing activities that have
already been established before the GDPR might
pose a risk for compliance. This also means that
if a certain data processing activity is not regula-
tion compliant, measures need to be taken to en-
sure lawful processing, which is the obligation of
the Data Controller. The Data Processor is con-
sulted to actually verify if its provided processing
activity is GDPR compliant while the DPO and
the Supervisory Authority are only contacted for
supporting or clarification purposes.
Creation of New Data Handling Processes: re-
lates to verifying new processing activities that are
developed regarding regulation compliance. This
activity contains not only setting boundary condi-
tions during the planning phase, for example spec-
ifying which kind of data and for what purpose it
may be used, but also constantly supervising the
ICEIS 2020 - 22nd International Conference on Enterprise Information Systems
658
progress of the development. Similar to the prior
activity, those tasks often require a thorough un-
derstanding of legal obligations imposed by the
GDPR. The DPO is usually strongly involved in
this activity. In summary, this activity is mainly
about applying the data protection by design and
by default principle specified in Art. 25 GDPR.
Identify Need & Conducting DPIAs: As de-
scribed in (Bieker et al., 2016), the data protec-
tion impact assessment (DPIA) consists of three
stages. In the Preparation stage, it is determined
if a DPIA is necessary at all. If that is the case, the
scope, involved actors and targets are defined and
legal requirements are identified. During the Eval-
uation stage, the total risk of a processing activity
is evaluated. This is done by identifying protec-
tion goals, potential attackers and their motives as
well as determining evaluation criteria and bench-
marks. In the last stage, the Report and Safeguard
stage, appropriate safeguards to mitigate the iden-
tified risks are selected and implemented. Lastly,
a report documenting the evaluation results is cre-
ated and published, which can then be evaluated
by a third party. Once a new processing activity
is developed or extended, the process is repeated.
For all this, the Data Controller is the responsible
actor while relying on the support of the DPO and
may consult the Supervisory Authority for clarifi-
cation purposes.
Cooperation with Supervisory Authority: in-
cludes the DPO acting “as the contact point for
the supervisory authority on issues relating to pro-
cessing, including the prior consultation referred
to in Article 36, and to consult, where appropri-
ate, with regard to any other matter” as required
by Art. 39 GDPR. However, also the Data Con-
troller and Processor are required to cooperate
with the Supervisory Authority if necessary (Art.
31 GDPR).
Maintaining Records of Processing Activities
(RPAs): Article 30 describes the information that
is required in the record of processing activites
(RPA), such as the responsible person, the legal
basis for processing, or the categories of data sub-
jects and personal data. The DPO supports in
compiling the information about all processing
activities.
Conducting Audits: Based on a guide published
in (UK ICO, 2018), the scope of an audit may
cover, inter alia, data protection governance and
accountability, data protection training and aware-
ness, and risk management. In summary, an audit
is used to verify that all other activities described
in this section are carried out correctly and to dis-
cover areas of noncompliance. An audit may be
initiated by the Data Controller or by the respon-
sible Supervisory Authority and is performed by
the DPO (Art. 28, Art. 39, Art. 58).
Dealing with Data Subjects: Besides reacting to
requests based on the Data Subject’s rights (Art.
12 through 22), the communication of a personal
data breach to the data subject is also addressed by
this activity. The responsible actor thereby is the
Data Controller, although usually the DPO acts as
the first point of contact, given that his/her contact
details normally are publicly available as specified
in Art. 37.
Report to Management: is not directly required
by the GDPR itself. However, during the ex-
pert interviews two interviewees suggested to in-
clude reporting to Management as a separate ac-
tivity. Since it is accountable for the execution of
all the activities described above, it is necessary
that an organization’s Management is constantly
informed about the current situation in DPM.
Note that it is not possible to strictly separate the ac-
tivities: supporting the creation of processing activi-
ties can involve considering a DPIA and preparing in-
formation about the processing activity for the RPA.
Nonetheless, we consider these tasks to be sufficiently
delineated so survey participants could assign them
without further explanation.
4.2 Task Complexity & Time
Consumption
Figure 3 shows the distribution of responses regarding
task complexity, where 1 indicates the lowest level of
complexity and 5 indicates the highest level. We ar-
ranged the activities in descending order of their aver-
age complexity (x
c
), with “Identify need & conduct-
ing DPIAs” being rated as the most complex and “Co-
operation with supervisory authority” being rated as
the least complex activity. Based on the sample distri-
bution of the five complexity levels for each activity,
we assign the activities to three groups:
The first group consists of “Identify need & con-
ducting DPIAs” (x
c
= 3.47), “Verifying already ex-
isting data handling processes regarding compliance”
(x
c
= 3.45) and “Creation of new data handling pro-
cesses” (x
c
= 3.32) and can be described as the group
of “most complex” activities because of the large
amount of participants rating these activities with a
complexity level of 4 or 5.
The group of “moderately complex” activities is
composed of “Maintaining records of processing ac-
Empirical Task Analysis of Data Protection Management and Its Collaboration with Enterprise Architecture Management
659
Figure 3: Complexity Distribution of Activities (N=38).
tivities (RPAs)” (x
c
= 3.05), “Dealing with Data Sub-
jects” (x
c
= 2.89) and “Conducting Audits” (x
c
=
2.87), given that complexity level 3 is prevalent here.
Finally, Awareness-raising and schooling within
the organization” (x
c
= 2.76), “Report to Manage-
ment” (x
c
= 2.55) and “Cooperation with supervi-
sory authority” (x
c
= 2.39) build the group of “least
complex” activities, which is based on the high count
of complexity level 1 and 2 ratings. Although as-
signed to the group with the least average complex-
ity, Awareness-raising and schooling within the or-
ganization” was rated as very complex by four partic-
ipants.
The average time consumption of each activity is
shown in figure 4. To prevent a distortion of the re-
sults, we decided to exclude the answers of one partic-
ipant from this examination, given that the time con-
sumption of “Awareness-raising and schooling within
the organization” was rated with 100%.
With on average just over 20%, “Verifying already
existing data handling processes regarding compli-
ance” was rated as the most time consuming activ-
ity by far. While most of the activities consume be-
tween 10% to 14% of a DPOs time, “Report to Man-
agement”, “Dealing with Data Subjects” and “Coop-
eration with supervisory authority” were rated as less
time intensive.
4.3 Most Frequent Problems
The results of the problem analysis for the whole sam-
ple are summarized in Table 5. Per row, the relative
frequency of the 38 participants that selected a prob-
lem as one of the two most severe ones is shown. Con-
sequently, the upper limit for the sum of each row’s
values is 200%. However, it is important to note that
the total amount of selected problems differs in same
cases significantly from each other. For example, with
an average of 1.7 selected problems per participant,
“Verifying already existing data handling processes
regarding compliance” shows the highest number of
problems of the nine activities. In contrast, usually
less than one problem per participant was selected
for “Report to Management”, resulting in the lowest
number of ticked problems of all activities. For all
nine activities on average 1.3 problems were selected.
In the following, the characteristics of the problems’
frequency distributions are addressed for each activ-
ity.
For the activity “Awareness-raising and schooling
within the organization” especially “Lack of per-
sonnel” and “Missing practical guidelines/ stan-
dard procedures” often pose a problem.
When looking at the activities “Verifying already
existing data handling processes regarding com-
pliance” and “Creation of new data handling pro-
cesses”, it is apparent that their frequency dis-
tribution of problems is very similar. In both
cases, “Lack of personnel” and “Missing practical
guidelines/ standard procedures” again are an is-
sue. However, also “Finding the right contact per-
son(s)” and having “Insufficient information on
single data processing activities” pose a challenge
with almost equal frequency.
The problems’ frequencies for “Identify need &
conducting DPIAs” are quite evenly distributed
with around 20% of participants selecting them as
most severe. Only the absence of practical guide-
line and/or standard procedures seems to pose a
challenge more often, whereas on the other hand
the lack of authority and a holistic view seem like
less important problems in that context.
The “Cooperation with supervisory authority” is
mostly hindered because of “Missing practical
guidelines/ standard procedures”, “Lack of per-
ICEIS 2020 - 22nd International Conference on Enterprise Information Systems
660
Figure 4: Mean Time Consumption of Activities (N=37).
Figure 5: Frequencies of Problems Rated as Most Severe (N=38).
sonnel” and difficulties with “Finding the right
contact person(s)”. The remaining problems
thereby only play a subordinate role for most par-
ticipants.
For the activity “Maintaining records of process-
ing activities (RPAs)” the problem “Missing prac-
tical guidelines/ standard procedures” was se-
lected most frequently by the participants. Al-
though showing lower frequencies, the issues
“Missing holistic view on system landscape” and
“Insufficient information on single data process-
ing activities” chosen by 21% as well as “Lack of
right tools/technology” and “Lack of personnel”
selected by 18% of the participants should not be
neglected.
Looking at the activity “Conducting audits”, the
issue “Missing practical guidelines/ standard pro-
cedures” reaches its highest frequency of all ac-
tivities with 32% of respondents rating it as one of
the two most severe ones. However, also a short-
age of personnel was reported by 24% of the ques-
tioned DPOs.
When it comes to “Dealing with Data Sub-
jects”, most often “Insufficient information on
single data processing activities”, “Inaccuracy
of European legislation” and “Lack of right
tools/technology” were chosen as most serious
issues. The remaining problems were selected
equally frequent by 8% of the participants.
For the last activity, “Report to Management”,
the “Lack of Authority” is by far the most severe
problem with a selection rate of 26%. Further-
more, a lack of additional personnel was chosen
by 16% of the participants. All other problems
show a frequency of under 10% in each case.
The problems “Lack of personnel” and “Missing
Empirical Task Analysis of Data Protection Management and Its Collaboration with Enterprise Architecture Management
661
practical guidelines/ standard procedures” were espe-
cially common for most of the activities and were
therefore often directly addressed in this section.
However, since the frequencies are in general rather
evenly distributed, less often selected issues should
not be neglected. For example, although the two
aforementioned problems show high frequencies in
Awareness-raising and schooling within the organi-
zation”, their share of the total number of selected
problems for this activity is only 42%.
4.4 Collaboration with EAM
In the last part of the survey, the current status of the
collaboration between DPM and EAM was evaluated.
The questionnaire was forked to examine the value
of EAM support if EAM is already supporting DPM
and to discover reasons for a missing collaboration if
that is not the case. We therefore use a similar struc-
ture for this section by first inspecting the results of
the 12 EAM collaborators followed by the reasons for
non-collaboration provided by the remaining 26 par-
ticipants. Finally, the (potential) support of EAM in
general is analyzed for both groups.
The participants who are already collaborating
with EAM were first asked to rate the support for each
of the nine activities. We present the results in fig-
ure 6. The activities are presented in decreasing or-
der depending on the average value of EAM support
based on the answers for the different levels of help-
fulness. For the calculation of the arithmetic mean
(x
s
) the amount of people not using EAM for an ac-
tivity were therefore excluded. The different levels of
helpfulness were weighted as follows: (1) Not very
helpful; (2) Somewhat helpful; (3) Very helpful; (4)
Extremely helpful. The majority of participants who
receive support rated the same as either very or ex-
tremely helpful.
For the activities “Verifying already existing
data handling processes regarding compliance” (x
s
=
2.67), “Creation of new data handling processes” (x
s
= 2.58), Maintaining records of processing activities”
(RPAs) (x
s
= 2.58) and “Identify need & conduct-
ing DPIAs” (x
s
= 2.33) each of the 12 participants
said they are supported by EAM. Except for DPIAs,
the support was rated as very or extremely helpful
by at least half of the DPOs. Although the number
of EAM supported DPOs decreases for the activities
Awareness-raising and schooling within the organi-
zation” (x
s
= 2.33), “Dealing with Data Subjects” (x
s
= 2.00) and “Conducting Audits” (x
s
= 1.91), in the
case of the former two activities 75% and in the case
of the latter activity 92% of participants still collabo-
rate with EAM. Only the activity “Cooperation with
supervisory authority” (x
s
= 1.80) is usually not sup-
ported by EAM and if it is, the value of the support is
also rated rather low when compared to other activi-
ties.
The 26 participants that stated they are not col-
laborating with EAM were asked to justify that by
providing one or more reasons. Around half of
the respondents (n=14) answered that “EAM does
not exist in the organization”, by far the most fre-
quently selected reason. Other explanations are
“I don’t know if EAM exists in the organization”
(n=4) or “Missing contact (persons)” (n=4), followed
by “Other goals/objectives/level of detail” (n=3),
“No Time/Resources for my part” (n=3) and No
Time/Resources on the part of EAM (n=2). Each of
the remaining justifications were only selected once.
Although the participants were given the opportunity
to add their own reasons, only one stated that “The
soft- and hardware is provided by an external organi-
zation”.
In the last section of the questionnaire the respon-
dents were asked to rate the (potential) support of
EAM for DPM, and vice versa, using a 10-point rat-
ing scale
2
. As shown in figure 7, the participants who
already collaborate with EAM rated the support pro-
vided by EAM generally higher than the participants
that do not. However, also two respondent who col-
laborate with EAM evaluate it as “Very low”.
Examining the support of support of DPM for
EAM (see figure 8 we see a similar trend. While the
distribution of ratings of non-collaborators is almost
equal to the one of the general support of EAM for
DPM, the DPOs receiving support rate the use of their
discipline for EAM lower than the other way around.
Nevertheless, collaborators still see a higher value of
DPM for EAM than non-collaborators. This time,
only one person of the collaborator group rated the
support with the lowest possible value.
5 CONCLUSION & OUTLOOK
In this work, we empirically evaluated DPM tasks
with respect to their complexity, the time consump-
tion and the main problems that occur when conduct-
ing these activities. We did this with literature analy-
sis, four qualitative expert interviews and a quantita-
tive survey with 38 DPOs. Additionally, we evaluated
the current or potential future support with EAM from
the DPO’s perspective. Thus, our three research ques-
tions can be answered:
2
Due to the different sample sizes (12 to 26), we use
relative frequencies in figures 7 and 8.
ICEIS 2020 - 22nd International Conference on Enterprise Information Systems
662
Figure 6: Distribution of EAM Support Effectiveness (N=12).
Figure 7: (Potential) Support of EAM for DPM (N=38).
Figure 8: (Potential) Support of DPM for EAM (N=38).
RQ1: Which DPM Activities are Necessary to
Achieve GDPR Compliance?
Based on related work and the input of four DPM ex-
perts, the spectrum of tasks in DPM was grouped into
nine activities which were rated in terms of complex-
ity and time consumption by the survey participants.
Based on the distribution of rated complexity levels,
each activity could be assigned to a group of either
rather complex, moderately complex or rather easy
activities. For an activity’s time consumption its rela-
tive average share of a DPOs total work time is used
as a measure. In the following, the final list of ac-
tivities with their respective rating in complexity and
time consumption (TC) is presented:
Most Complex:
Identify need & conducting DPIAs
(TC 10%)
Verifying already existing data handling pro-
cesses regarding compliance (TC 20%)
Creation of new data handling processes
(TC 14%)
Moderately Complex:
Maintaining records of processing activities
(RPAs) (TC 13%)
Dealing with Data Subjects (TC 7%)
Conducting Audits(TC 11%)
Least Complex:
Awareness-raising and schooling within the or-
ganization (TC 13%)
Report to Management (TC 7%)
Cooperation with supervisory authority
(TC 5%)
Given that the nine activities were selected and eval-
uated during the four expert interviews and subse-
quently verified by 38 DPOs during the questionnaire,
it can be said with confidence that these activities may
be used to reliably cover the range of tasks in DPM.
Furthermore, the activities’ complexity and time con-
sumption characteristics may be used to determine
the activities where support is most urgently needed.
Following this line of thought, especially the activ-
ities “Verifying already existing data handling pro-
cesses regarding compliance” and “Creation of new
data handling processes” should be considered first
when designing support measures, which in turn may
focus on mitigating the respective problems for these
activities.
Empirical Task Analysis of Data Protection Management and Its Collaboration with Enterprise Architecture Management
663
RQ2: What are the Most Severe Problems when
Conducting these Activities?
In the second part of the survey, the participants were
asked to rate the two (or less) most severe problems
for every activity. As we show in figure 5, the fre-
quencies of the different problems vary considerably
depending on the underlying activity. Nonetheless,
we found that a lack of personnel is for many ac-
tivities one of the most severe issues which might
be rooted in organizations seeing GDPR compliance
more as a hindrance that must be overcome with as
little resources as possible, instead of viewing it as
an opportunity for achieving a competitive advan-
tage (Almeida Teixeira et al., 2019). Furthermore,
implementing guidelines or standard procedures may
be supported by an integrated tool support approach
and might reduce the effort of performing certain ac-
tivities. An official guide or self assessment as pro-
vided in (UK ICO, 2018) may therefore be integrated
in DPM and should at the same time reduce the am-
biguity of the legislation. Finally, we showed that for
process-heavy activities there is often a lack of de-
tailed information about a single data handling pro-
cess as well as missing contact persons. A stronger
focus on documentation (for future verification pur-
poses) as well as using other information sources (for
current verification purposes), such as the EA, may
therefore be necessary to effectively mitigate these
problems.
RQ3: How are DPOs Collaborating with EAM
and How do they Evaluate the Helpfulness of
EAM?
From the 38 participants, only 12 stated they col-
laborate with EAM. The most often provided reason
for this low collaboration rate is by far the nonexis-
tence of EAM in the organizations. While the group
of non-collaborators is rather indecisive in their at-
titude towards a collaboration with EAM, the DPOs
that do work together with EAM consider the collab-
oration significantly more helpful. Furthermore, in
both groups the (potential) support of EAM for DPM
was rated higher than the support of DPM for EAM.
When looking at the EAM support for the nine activi-
ties, it can in general be described as quite effective. It
proves especially helpful for reporting to management
and verifying already existing data handling processes
as well as for the creation of new data handling pro-
cesses and maintaining RPAs. Only the cooperation
with the supervisory authority is rather ineffectively
supported if it is at all.
Outlook
Our findings validate future research that addresses
one or more DPM activities. With respect to the col-
laboration between EAM and DPM, future research
might detail on organizational factors that influence
this collaboration, e.g. maturity levels of EAM. Our
future work aims at advancing the notions of privacy
and data protection from the perspective of EAM.
ACKNOWLEDGEMENTS
This work was sponsored by the German Federal
Ministry of Education and Research (BMBF) grant
01IS17049 / UMEDA and X-DACE. The responsibil-
ity for the content of this publication lies with the au-
thors.
REFERENCES
Almeida Teixeira, G., Mira da Silva, M., and Pereira, R.
(2019). The critical success factors of gdpr implemen-
tation: a systematic literature review. Digital Policy,
Regulation and Governance, 25(4):1.
Bieker, F., Friedewald, M., Hansen, M., Obersteller, H., and
Rost, M. (2016). A process for data protection impact
assessment under the european general data protec-
tion regulation. In Schiffner, S., Serna, J., Ikonomou,
D., and Rannenberg, K., editors, Privacy Technolo-
gies and Policy, pages 21–37, Cham. Springer Inter-
national Publishing.
Burmeister, F., Drews, P., and Schirmer, I. (2019). A
privacy-driven enterprise architecture meta-model for
supporting compliance with the general data protec-
tion regulation. In Proceedings of the 52nd Hawaii
International Conference on System Sciences.
Christina Tikkinen-Piri, Anna Rohunen, and Jouni
Markkula (2018). Eu general data protection regula-
tion: Changes and implications for personal data col-
lecting companies. Computer Law & Security Review,
34(1):134–153.
CIPL and AvePoint (2018). Organisational Readiness for
the European Union General Data Protection Regula-
tion. Technical report.
CMS Hasche Sigle (2019). GDPR Enforcement Tracker.
http://www.enforcementtracker.com/, accessed 2019-
11-29.
Ernst & Young and International Association of Privacy
Professionals (2019). IAPP-EY annual privacy gov-
ernance report 2019. Technical report.
Farwick, M., Breu, R., Hauder, M., Roth, S., and Matthes, F.
(2013). Enterprise architecture documentation: Em-
pirical analysis of information sources for automation.
In Proceedings of the Annual Hawaii International
Conference on System Sciences, pages 3868–3877.
ICEIS 2020 - 22nd International Conference on Enterprise Information Systems
664
Huth, D., Tanakol, A., and Matthes, F. (2019). Using Enter-
prise Architecture Models for Creating the Record of
Processing Activities (Art . 30 GDPR). In 23rd IEEE
International Distributed Object Computing Confer-
ence (EDOC), Paris.
International Association of Privacy Professionals (2019).
Measuring Privacy Operations. Technical report.
Koc¸, H., Eckert, K., and Flaig, D. (2018). Daten-
schutzgrundverordnung (dsgvo): Bew
¨
altigung der
herausforderungen mit unternehmensarchitekturman-
agement (eam). HMD Praxis der Wirtschaftsinfor-
matik, 55(5):942–963.
Larno, S., Sepp
¨
anen, V., and Nurmi, J. (2019). Method
Framework for Developing Enterprise Architecture
Security Principles. Complex Systems Informatics and
Modeling Quarterly (CSIMQ), 117(20):57–71.
Miniwatts Marketing (2019). Internet World Stats. https:
//internetworldstats.com/stats.htm, accessed 2019-10-
01.
Sirur, S., Nurse, J. R. C., and Webb, H. (2018). Are we there
yet? Understanding the challenges faced in complying
with the General Data Protection Regulation (GDPR).
In Proceedings of the 2nd International Workshop on
Multimedia Privacy and Security, pages 88–95. ACM.
UK ICO (2018). A guide to ICO audits. https:
//ico.org.uk/media/for-organisations/documents/
2787/guide-to-data-protection-audits.pdf, accessed
2019-11-28.
Empirical Task Analysis of Data Protection Management and Its Collaboration with Enterprise Architecture Management
665