Content uploaded by Jil Klünder
Author content
All content in this area was uploaded by Jil Klünder on Nov 20, 2017
Content may be subject to copyright.
Content uploaded by Jil Klünder
Author content
All content in this area was uploaded by Jil Klünder on Sep 11, 2017
Content may be subject to copyright.
HELENA Study: Reasons for Combining Agile
and Traditional Software Development
Approaches in German Companies
Jil Kl¨under1, Philipp Hohl2, Masud Fazal-Baqaie3, Stephan Krusche4, Ste↵en
K¨upper5, Oliver Linssen6, and Christian R. Prause7
1Leibniz Universit¨at Hannover, jil.kluender@inf.uni-hannover.de
2Daimler AG, philipp.hohl@daimler.com
3S&N CQM GmbH, masud.fazal-baqaie@sn-cqm.de
4Technische Universit¨at M¨unchen, krusche@in.tum.de
5Technische Universit¨at Clausthal, steffen.kuepper@tu-clausthal.de
6FOM University of Applied Sciences for Economics and Management,
oliver@linssen-consulting.de
7German Aerospace Center, christian.prause@dlr.de
Abstract. Many software development teams face the problem of se-
lecting a suitable development approach fitting to their specific context.
According to them, the combination of agile and traditional approaches
seems to be the solution to handle this problem. However, the current
state of practice with respect to hybrid approaches is not sufficiently ex-
amined. Most studies focus either on traditional or on agile methods, but
the combination of both is not well investigated yet. The “Hybrid dE-
veLopmENt Approaches in software systems development” (HELENA)
study performs a large-scale international survey in order to gain insights
into the distribution of hybrid approaches. So far, the study indicates sev-
eral reasons why companies combine agile and traditional approaches.
The hybrid approaches aim at improving the frequency of delivery to
customers, the adaptability and the flexibility of the process to react to
change. Furthermore, it is the aim to increase the productivity. In this
publication, we present the current state of the German results and out-
line the next steps.
Keywords: HELENA study, hybrid software development, empirical
study in Germany
1 Introduction
Nowadays, there exist various methods and practices to develop software. The
methods consist of agile and plan-driven processes [6]. However, it seems to be
a best practice to combine both approaches. While the plan-driven process pro-
vides clear process models with an overall project structure, the agile approach
enables more flexibility and individuality by focusing on shorter time-to-market
and customer satisfaction [1]. In order to obtain the advantages of both ap-
proaches, hybrid software development seems to increasingly spread into indus-
try. To investigate this topic in detail, the HELENA study was brought to life.
© SPRINGER. PREPRINT. This is the author's version of the work. It is posted here by permission of Springer for your personal use.
Not for redistribution. The definitive version was published in the conference/workshop proceedings.
The study investigates the combinations of agile, traditional, and other kinds
of software development approaches in use. Furthermore, the study examines
how agile methods and practices are integrated into traditional development
approaches and why they are selected.
Currently, 85 researchers from 26 countries contribute to the study of hybrid
development approaches. This paper presents the current state of the data col-
lection in Germany, shows an overview of preliminary results and outlines the
next steps with respect to data analysis.
2 Related Work
There are only a few publications focusing on the prevalence of hybrid ap-
proaches: Boehm and Turner [3] motivate the combination of agile and plan-
driven approaches. They mention that a changing world needs agile and dis-
ciplined development methods. They characterize “home grounds” where the
approaches are most likely to succeed and identify five critical dimensions. With
a classification within the critical dimensions, it is possible to set up a balanced
strategy for a successful combination of agile and plan-driven approaches. The
presented risk-based method takes advantage of the strengths and mitigates the
weaknesses of both approaches.
Diebold and Zehler [4] examine the process of combining agile and traditional
development methods. They distinguish between the revolutionary and the evo-
lutionary approach, which di↵er in the order of occurrence of the methods. The
authors describe the coexistence of both development methods, but they do not
investigate their distribution.
Kuhrmann and Linssen [5] examine the use of process models in Germany.
They compare the data from 2006 with the data of 2013 and observe the emer-
gence of many di↵erent models and approaches. They point out that the com-
bination of traditional process models and agile development approaches is per-
vasive. However, agile approaches are not as dominating as promoted by the
agile community. Theocharis et al. [7] report of a high popularity of hybrid ap-
proaches. They experience the lack of quantitative data representing the use of
development methods. The HELENA study aims at examining this research gap
in detail.
3 Data Collection in Germany
Since May 2, 2017, the questionnaire of the HELENA survey is available on-
line in German, English, Spanish and Portuguese. The German team of HE-
LENA consists of 14 researchers from 11 di↵erent institutions. The researchers
encouraged practitioners from di↵erent German organisations including SMEs
and companies to participate in the study. Therefore, they sent personalized
emails to contacts within organisations and used mailing lists of software engi-
neering communities. Like teams from other countries, they also distributed the
questionnaire using social media via Twitter, XING and ResearchGate.
The data points collected until this intermediate report seem to indicate selec-
tion and response biases resulting from the invitation method (personal emails).
To mitigate these biases, the researchers started a Google AdWords campaign
in order to find additional participants without a personal relationship to the
researchers. (Note that the data collected during this campaign is not included
in this report.) After 10 days, advertisements to “participate in the scientific
survey” were displayed 40K times and 300 people clicked through the survey.
Until now, five people completed the questionnaire over this campaign (one of
them from Peru as advertisements were initially not restricted to Germany).
4 Overview of Preliminary Results
Based on the data collection until August 15, 2017, the German team collected
95 complete data records from German software developers. Most of the par-
ticipants (33%) are employed in very large organizations with more than 2500
employees. 31% work in large organisations with more than 250 employees (cf.
Table 1). There have been 45 more responses from larger organisations than from
smaller ones. Among the selection bias, a possible reason might be that hybrid
approaches are more interesting for large companies because they are more likely
to use traditional development processes and aim to speed up development. Hy-
brid approaches promise them an improved development process. Small software
companies tend to apply agile methods and practices right from the beginning.
Hence, they often do not think about implementing hybrid approaches so far.
The HELENA study also asks about the size of the developed software prod-
ucts and the project length. In three of four cases, the product size is more than
one person year (76%). Only 2% of the projects do not last longer than two per-
son months. One third of the teams is not distributed (35%). Half of the teams
is distributed either globally (26%) or nationally (same country) (24%). 15% of
the projects are regionally distributed, i.e. distributed on the same continent.
One third of the participants either works in the automotive domain (16%)
or in the financial sector (15%). The automotive sector is strong in Germany.
Hence, it is plausible that there is a high participation from automotive software
developers. 12% of the participants work in the space domain. However, these
12% do not represent the real-world industry distribution of the space domain
in Germany and hence may indicate skewed representation.
Company size # Participants # Companies using hybrid approaches
<10 10 7 (70%)
10-50 5 5 (100%)
51-250 20 14 (70%)
251-2500 29 22 (76%)
>2500 31 25 (71%)
Table 1 . Number of companies using hybrid approaches
26 project or team managers (27%) and 16 developers (17%) from German
companies participated in the survey. Eleven participants are quality managers
(12%). Most participants have more than ten years of working experience (62%).
The findings in Table 1 show that the combination of agile and traditional de-
velopment methods do not depend from the size of the company. In each com-
pany size category, more than 70% of the interviewed participants use hybrid
approaches.
39 participants in our study (41%) stated that each project within their
company can individually decide which process should be used. 20 participants
(21%) report that decisions are made on business unit level. 38% of the projects
are operated according to a in-house standard process. Projects either decide
about specific practices and methods on demand during the project (37%) or a
project manager tailors the process in the beginning (19%). In 15% of the cases,
the customer is taken into account when selecting the practices and methods.
Fig. 1 gives an overview of some goals, companies want to reach by selecting
individual development approaches, such as time-to-market, employee satisfac-
tion and improved delivery pace.
Fig. 1. Reasons for companies to implement hybrid approaches (extract)
Most very large (>2500 employees) companies combine agile and traditional
approaches to improve the frequency of delivery to customers (64%), to im-
prove the adaptability and flexibility of the process to react to change (64%)
and to improve the productivity (64%). Large companies (251-2500) also aim
at improving the productivity (81%), the planning and estimation (67%) and
the adaptability and flexibility of the process to react to change (62%). Micro
companies (<10 employees) also want to increase the productivity (57%) and
the external product quality (57%). The small companies mostly want to satisfy
the employees (80%), which seems to be less important for companies which are
either smaller (29%) or larger (33% resp. 19%). The very large companies also
want to increase employee satisfaction (50%).
5 Future Work
This paper presents ongoing research. So neither the data collection nor the
analysis are complete yet. Next, we present a set of initial research questions for
exploring the distribution of hybrid approaches in Germany and worldwide.
In order to examine the distribution of hybrid development approaches within
di↵erent company sizes, we are interested in analysing domain-specific contexts.
In the future, we want to examine, if there is a correlation between organization
size, the implemented new roles and the way of working in order to gain insights
into advantages and disadvantages, difficulties and experiences with more or less
suitable combinations. Therefore, we aim at answering the following research
questions:
RQ1: Are there any domains working with a purely agile approach?
RQ1.1: Are there context factors that enable the implementation of agile?
RQ1.2: Which agile approaches are in use when implementing agile?
RQ2: Are there any domains working with a purely plan-driven approach?
RQ2.1: Are there context factors that inhibit the integration of agile and lead to the
implementation of plan-driven approaches?
RQ3: Which domains primarily use hybrid approaches?
RQ3.1: Which domain-specific context factors support the implementation of hybrid
approaches?
RQ3.2: Which combinations are widely distributed and which ones are less suitable?
RQ3.3: Are there best practices when implementing hybrid approaches?
RQ3.4: Do common practice and best practice di↵er from each other?
According to Boehm [2], agile and plan-driven software development ap-
proaches have di↵erent home grounds, i.e., agile development is favourable for
fast-paced markets, while domains with high failure costs tend to favour tradi-
tional development models.
RQ4: What is the e↵ect of software criticality on the choice of development
approach?
RQ4.1: Is there a clear relationship between the choice of the development approach
and the criticality of developed software?
RQ4.2: Do domains with expected higher failure costs (e.g., aerospace, automotive,
medicine) favour more traditional development approaches?
6 Conclusion
The results of our study indicate a high popularity of hybrid development ap-
proaches in Germany. Independent of the size of the organization, many project
teams combine individually selected development approaches. Most of the or-
ganisations aim at improving the productivity, the customer’s perceived prod-
uct quality, planning and estimation as well as the frequency of delivery to the
customer. We plan to extend our data collection and analysis in future work.
References
1. Beck, K., Beedle, M., Van Bennekum, A., Cockburn, A., Cunningham, W., Fowler,
M., Grenning, J., Highsmith, J., Hunt, A., Je↵ries, R., et al.: Manifesto for Agile
Software Development (2001)
2. Boehm, B.: Get Ready For Agile Methods, With Care. Computer 35(1), 64–69
(2002)
3. Boehm, B., Turner, R.: Balancing Agility and Discipline: A Guide for the Perplexed,
Portable Documents. Addison-Wesley Professional (2003)
4. Diebold, P., Zehler, T.: The Right Degree of Agility in Processes. In: Managing
Software Process Evolution, pp. 15–37. Springer (2016)
5. Kuhrmann, M., Linssen, O.: Welche Vorgehensmodelle nutzt Deutschland? In: Pro-
jektmanagement und Vorgehensmodelle 2014. pp. 17–32 (2014)
6. Kuhrmann, M., M¨unch, J., Diebold, P., Linssen, O., Prause, C.R.: On the Use of Hy-
brid Development Approaches in Software and Systems Development: Construction
and Test of the HELENA Survey. In: Projektmanagement und Vorgehensmodelle
2016. pp. 59–68 (2016)
7. Theocharis, G., Kuhrmann, M., M¨unch, J., Diebold, P.: Is Water-Scrum-Fall Re-
ality? On the Use of Agile and Traditional Development Practices. In: Interna-
tional Conference on Product-Focused Software Process Improvement. pp. 149–166.
Springer (2015)
