The impact of tailoring criteria on agile practices adoption: A survey with novice agile practitioners in Brazil

Highlights

• Tailoring criteria have a relevant impact on the adoption of agile practices.

• Main criteria are external environment, previous knowledge and internal environment.

• Organizations adopt agile practices to fit their context, culture and objectives.

• Organizations adopt hybrid or custom software methods to meet their needs.

• Novice practitioners in Brazil start using traditional software development practices.

Abstract

The software development industry adopts agile methods in different ways by considering contextual requirements. To fulfill organizational needs, adoption strategies consider agile methods tailoring. However, tailoring according to the context of the organization remains a problem to be solved. Literature on criteria for adopting software development methods exists, but not specifically for agile methods. Given this scenario, the following research question arises: what is the impact of software method tailoring criteria on the adoption of agile practices? To answer this question, we conducted a survey among agile practitioners in Brazil to gather data about importance of the tailoring criteria and agile practices adopted. A model for agile practices adoption based on the tailoring criteria is proposed using the results of the survey with a majority of novice agile practitioners. The proposed model was validated using PLS-SEM (partial least squares structural equation modeling) and the survey data. Results show that adoption of agile practices was influenced by criteria such as external environment, previous knowledge and internal environment. Results also indicate that organizations tend to use hybrid/custom software methods and select agile practices according to their needs.

Introduction

Agile methods and practices have been widely adopted by the software development industry (Qumer, Henderson-Sellers, September 11–14, 2006, Deemer, Benefield, Larman, Vodde, Nishijima, Dos Santos, 2013, VersionOne). However, adoption strategies can vary, from full adoption at once to iterative and incremental approaches (Nerur, Mahapatra, Mangalaraj, 2005, Soundararajan, Balci, Arthur, August 5–9, 2013, Rohunen, Rodriguez, Kuvaja, Krzanik, Markkula, 2010). Organizations use agile practices adoption and practice selection to generate tailored software development methods consistent with their values, culture, reality, needs and strategies (Ahmed, Sidky, 24–28 August 2009, Abbas, Gravell, Wills, August 9–13, 2010, Esfahani, Yu, Annosi, June 22–24, 2011, Kurapati, Manyam, Petersen, May 21–25, 2012, Madi, Dahalin, Baharom, 2011, Qumer, Henderson-Sellers, 2008, Rohunen, Rodriguez, Kuvaja, Krzanik, Markkula, 2010, Lpez-Martnez, Jurez-Ramrez, Huertas, Jimnez, Guerra-Garca, 2016). Software method tailoring research provides ways to tailor the software method and adapt it to the context to which it is being applied (Fitzgerald et al., 2000). Likewise, that research defines tailoring criteria to guide the tailoring process, regardless of the type of method adopted (Kalus and Kuhrmann, 2013). The association of agile practices adoption with software method tailoring criteria raises the research question: what is the impact of software method tailoring criteria on the adoption of agile practices?¹

Not all organizations are ready to fully implement agile, when considering cultural and technical aspects (Qumer, Henderson-Sellers, 2008, Lpez-Martnez, Jurez-Ramrez, Huertas, Jimnez, Guerra-Garca, 2016, Hoda, Noble, 2017). These types of changes demand a considerable financial investment and have a big impact on the daily tasks performed by the teams. Instead, the adoption could be partial, keeping in place practices currently used by the organization, combining them with a new set of agile practices, to enable the organization to achieve its goals.

There is no standard on how agile practices fit into the organizational context (Rohunen et al., 2010). Agile practices adoption can be unique, as each organization takes into account the specific needs of each area of the development process. Often, organizations need help choosing the right combination of agile practices for their needs. This makes the usage of the tailoring criteria important, as it provides a possible solution for the agile practices adoption problem. The tailoring criteria, having already been defined in the literature (Kalus and Kuhrmann, 2013), can create a standard for guiding the selection of agile practices by establishing the importance of certain aspects of the organizational context.

The approaches we found in the literature for agile practices adoption generally focus on reporting previous practical experiences, or proposing new techniques for selecting such practices. The objective of this research is to understand the impact of tailoring criteria on the adoption or selection of agile practices. In our approach, we developed a model for the adoption of agile practices based on software method tailoring criteria defined by Kalus and Kuhrmann (2013) to analyze the impact of tailoring criteria on agile practices adoption. Multivariate statistical analysis techniques allowed the model to be created and validated. In the proposed model, tailoring criteria and agile practices have been represented using PLS-SEM (partial least squares structural equation modeling) (Hair Jr, Hult, Ringle, Sarstedt, 2013, Wong, 2013). We conducted a survey among novice agile practitioners in Brazil and gathered the data among agile practitioners to feed and validate the model. We used the SmartPLS 3.0 tool (Ringle et al., 2015) to validate the model and results show that three out of six tailoring criteria have a relevant impact on the adoption of agile practices.

The remainder of the paper is organized as follows. Section 2 provides technical background while Section 3 presents the related work. Section 4 discusses the research methodology, data gathering and analysis procedures. The results of the study are presented in Section 5, and the findings are discussed in Section 6. The conclusions and future work opportunities are outlined in Section 7.