1 Introduction

This paper implants the foundations for knowing the best practices in knowledge management and investigates the intensity of awareness and implementation of knowledge management principles and practices by the Indian software industry working in Agile Methodologies. Besides large organizations, the success of Small and Medium Enterprises (SMEs) can be related to how well they manage their knowledge. The Indian software industry is growing with every passing day; Nasscom predicted that India’s technology industry is expected to record a positive growth of 2.3% in FY2021 to touch $194 billion and India’s exports of software services (excluding exports through commercial presence) were estimated at US$ 133.7 billion during 2020–21, This corroborates the dominance of the Indian software industry in the global scenario. To attract more clients, Indian industry has started working in agile methodologies [39]. Agile methodologies to support better communication and practice-driven structure that helps in generating closed collaboration among team members and customers. Agile methodologies are based on four basic values called agile manifesto (www.agilemanifesto.org). One key value supports in favor of working software without documentation rather than a non-working software with a huge amount of documentation ([22]; Singh and Singh, 2012). Opponents of the agile methodologies argue, what will happen if a person leaves the company in between software life cycles and it will be almost impossible for a new person to understand the product without documentation or knowledge transfer. The success of organizations across all scales (large, medium, and small) can be linked with how well the organizations have been managed knowledge. Although large organizations are well observed from Knowledge Management (KM) perspective, fewer studies tried to analyze KM in SMEs, moreover, there is limited research linking KM in Multinational Companies (MNCs) and SMEs working in agile methodologies [19, 24, 33, 46].

A framework is being proposed and depicting the determinant role of KM approaches on organizational characteristics. This study presents a more critical classification of KM approaches that have been adopted based on size, geographical dispersion, and type of organization. This paper offers a new context, showing agile organizational characteristics into KM approaches. The study contributes to research by providing a more nuanced classification of KM approaches and types of KM approaches that have been adopted based on the size, geographical dispersion, and task nature of organizations. This will also provide insight for the managers and authorities responsible for implementing KM using agile methodologies.

The organization of rest of this paper presents as section 2 briefly introduce agile knowledge management practices and its related research work. Section 3 discusses motivation behind presenting this research work and objective with which this research has been conducted. Research problem is framed in section 4 and section 5 discusses the research approach i.e. the methodology that has been followed in the present study. Results are presented in Section 6, and Section 7 presents Fuzzy Analytic Hierarchy Process to understand the multi criteria decision making process. In section 8, study has been concluded and directions for future work have been discussed.

2 Agile knowledge management

The most vital resource/asset in an organization is knowledge and the people who create knowledge [16, 21]. These organizations produce new knowledge origins by skilfully transmuting tacit, explicit, individual, and collective knowledge reservoirs with the help of a dynamic spiraling process [30]. Criticism on agile methods is its applicability and scalability in a geographically distributed environment [44]. In such a case, a tailored KM framework is required that can be integrated and synchronized with the collocated and distributed development to generate value for the business. KM can be easily accepted and employed in Agile Software Development (ASD) environment [49]. The culture of agile development comprises communication, cooperation, and knowledge sharing as basic values [25]. KM is about learning, and agile software development establishes an environment that supports learning processes [17, 32]. Many agile practices such as on-site customer, collective code ownership, collaborative workspace, whole team, pair programming, pair rotation, stand-up meetings, and sprint planning meetings foster KS and KM in agile projects. It acts as a key to some of the challenges faced in implementing KM in organizations because KM practices are naturally embodied in agile practices and help in improving the KM process [39].

Some of the experienced reports published by Vanhanen and Korpi [45], Cao et al. [6], Sfetsos et al. [37], Sison and Yang [41], Beecham et al. [3] conclude that XP/pair programming helps in greater learning and knowledge sharing. The researcher claims that agile methodologies are effective in the dissemination and retention of knowledge and have a positive impact on KM practices. Some of the case studies on the agile methodologies from a knowledge perspective also directed towards embeddedness of KM in agile practices: Study of two software development organizations highlights the significance of tacit knowledge in agile methodologies and knowledge management aspects of agile development [18]. Pair programming also seems to benefit teamwork, knowledge transfers, and learning (Vanhanen and Lassenius, 2005; [2, 10]).

A conceptual framework for knowledge sharing of Misra and Amritesh (2014) emphasis on learning of the individual practitioners for the sustainability of an agile team. This learning framework attempts to create and exchange knowledge among the involved entities. Agile methodologies have been practiced by practitioners for more than a decade, but still, there is very little research work has been steered to relate it to the domain of KM research. In late 2000, Bjørnson and Dingsøyra [4] noticed the possible application of cartographic school of KM to ASD because of its low-cost technical infrastructure. This agrees with Chau and Maurer [7] that floats a framework to handle the technological infrastructure for synchronous as well as asynchronous knowledge sharing among co-located as well as distributed agile team members. But the framework fails to explain the integration of KM practices of traditional and agile methods where organizations use traditional and agile methodologies for software development i.e. Partial Agile organizations [12]. Cases have been reported that organizations customize agile processes/practices to fit in their corporate model. Skipping some practices (Hodgkins and Hohmann, 2007, [42]), stripping some and enhancing other agile practices [9], replacing the agile terminology with traditional vocabulary [31] are some of the poorly reported customization implemented in large organizations. Furthermore, modifying the agile practices for distributed sites can negatively affect the development process (Paasivaara, 2008). The distributed organization will impose an additional burden on communication and knowledge sharing, but distribution and agile would still be used together [39]. Applying agile in scale will require deviations from some of the suggested practices [20]. Deficiency of communication in a distributed environment may affect sharing of knowledge and experience among agile teams [26].

3 Motivation and objective

Agile software development and management practices provide a platform where agile teams create and share their knowledge. The exchange of experience and knowledge increase the abilities of teams to cope with uncertain and ambiguous situations. However, contextual factors e.g. unstable teams, distributed teams, traditional culture may hinder knowledge sharing process, thus leading to mistrust, waiting times, and delays in delivery. Organizational forms are heavily based on collaborative relationships among small, medium, and large organizations. SMEs have generated competitive and dynamic environments where knowledge fertilization in SMEs is important in supporting the network of collaboration and the competitiveness of the whole system [14, 15, 35]. Knowledge sharing and learning models developed for the sustainability of ASD practices can be integrated with the existing KM infrastructure that can justify the scale and economic viability of the system. Large organizations are reaping the benefits of knowledge management practices, whereas small and medium enterprises have not exploited the adoption of these practices and show poor performance as compared to large originations [13, 27]. Similarly, few studies show KM practices that are naturally embodied with agile practices run in parallel [40]. Although many researchers have highlighted the processes and dimensions of knowledge and their adoption in large organizations, concerning small and medium organizations, there is hardly any consensus on the framework of KM, and it is still fragmented. Therefore, the level of adoption of KM within SMEs is heterogeneous and in the absence of a general framework for the adoption of KM among SMEs is making hindrance to take advantage of agile software development methodologies [19]. This research paper aims at addressing the importance of knowledge in software engineering and exploring the way agile practices support knowledge management systems within software organizations. Forms of knowledge present in agile processes can help in developing the KM strategy for the development and sustainability of SMEs working in agile methodologies by integrating these strategies with the existing KM framework.

The research is conducted with the objective tobring out KM adoption varies among small, medium, and large organizations working in agile methodologies. Exploration of features of agile methodologies will help in understanding the various forms of knowledge present in these practices and further integration of these practices in the KM framework can give fruitful results to the practitioners. This aspect of agile practices in an evolutionary phase, although research regarding KM in SMEs working in agile methodologies is increasing, more research efforts are still needed especially for analyzing KM from the viewpoint of agile methodologies. So, the following detailed questions have been framed for the analysis:

  1. a)

    Is there any variation in the adoption of KM practices at the organizational level?

  2. b)

    Is there any relationship between the core area of software engineering, team distribution, type of organization?

  3. c)

    Is there any relationship between the number of employees and KM practice level, keeping in mind all the characteristics of the organization?

The following are alternative hypotheses that have been investigated.

HA1:

The size of the organization plays a determinant role in the adoption of KM practices

HA2:

The core area of working of software organizations plays a determinant role in the adoption of KM practice.

HA3:

Team distribution plays a determinant role in the adoption of KM practices

HA4:

Type of organization plays a determinant role in the adoption of KM practices

HA5:

The adoption of KM practices is positively associated with the size of the organizations across different dimensions.

4 Research problem

Researchers have listed three aspects that avert SMEs from adopting KM practices in their routine work [33, 46]: Tacit nature of knowledge; common knowledge that is shared by every employee of the organization; Deficiency of human and financial resources in SMEs. Agile practices hold the key to some of the problems faced by KM (especially in the medium and small-scale industry) as several KM practices are naturally embodied in the agile practices. However, there is a gap in studies that deal specifically with agility and knowledge management [5, 38]. Inter-organizational and intra-organizational knowledge transfer involve the transfer of knowledge between the development team and customers and within a development team, respectively. The need is to comprehend the factors that contribute towards the formation, retention, and dissemination of knowledge in agile processes [1]. Thus, it is of practical significance to examine the acceptance of agile methodologies from a knowledge management perspective.

From literature review it was prominently surfaced that there are for main pillars that may affect the KM practices embedded in Agile Software Development: Size of organization, type of organization, core area of work of the organization and types of team. Most of the studies discussed in previous sections explore the togetherness of agile and KM practices, but fail to consider other organizational characteristics/dimensions that can play a vital role in implementing KM in organizations like the size of the organizations, types of teams working for the development of the software, number of employees in an organization, distribution of the teams, etc. Based on these broadly defied 4 parameters we have further classified each parameter into further sub parameters. These all factors can play a major role in deciding how KM infrastructure can be implemented in organizations [39]. In most of the large organizations, KM would have been formally introduced in organizational structure, but it is not the same case with medium and small-scale organizations. It will be unfair if one says that SMEs do not have a KM strategy in place. If an organization practices agile methodology somehow there is a KM infrastructure in place, as many KM practices are naturally embedded in agile methodologies so it must have to explore what different types of KM practices knowingly and unknowingly being practiced in these kinds of organizations.

As the world is becoming a global village, collaborative relationships among small, medium, and large enterprises working with collocated or distributed teams have gained popularity, exploration of the extent of adoption of KM practices among different dimensions is not much explored. Exploration of the literature leads to a considerable question, faced by many KM researchers that have not paid enough attention to is: whether different dimensions of the organizations (Scale of the Industry, Core area of operations, Team distribution, type of organization) affect the adoption of KM practices (Connelly and Kelloway [11], [36], Davis et al., 2007 and Zanjani et al. (2009))? To answer these questions, a framework is proposed that categorizes different KM practices in small, medium, and large organizations and makes propositions about how the size, core area of work, type of organization, and geographical dispersion affect the adoption of KM practices (Fig. 1). The survey is used as an instrument for measuring the variation in adoptions depending upon the characteristics of the organizations. The survey is directly adopted from the Knowledge Management Assessment Tool (KMART) (Andreson, 1996). A questionnaire was based on five constructs of the KMART tool: knowledge acquisition, knowledge sharing, training and mentoring, KM environment, and KM technology.

Fig. 1
figure 1

KM adoption framework

Full size image

5 Research approach

The survey is used as an instrument for measuring the adoption variation depending upon the functional specifications of organizations like size, the core area of work, type of organizations, and team distribution. Initially, organizations were classified into three categories: largest (having employees more than 5000), medium (having employees between 501 and 5000), small (having employees between 50 and 500) depending upon the strength of their employees. Respondents from organizations are also divided into three levels: A (Strategic Managers), B (Tactical Managers), C (Operational Managers). It was not possible to cover the whole universe of the study. A sample of only those SE organizations has been chosen which use many of the agile practices and have some knowledge Management System (KMS) in place. Top 500 Companies in terms of average business per employee registered with NASSCOM were identified and 100 companies were randomly approached to participate in the survey. Even distribution, 20 organizations in each category were approached, but eventually, 34 companies agreed to participate in the survey. Ten respondents from different levels are taken from each company making 340 responses in total. SPSS toll has been used for analysis of the data. A stratified sampling technique was used for the selection of the respondents. Level A population of the cluster has a lesser number of the population as compared to level C, so 1:6 responses were collected to maintain balance among the population as mentioned in Table 1.

Table 1 Plan and size of the samples for 34 respondents
Full size table

To gain insight of KM practice level, organizations are divided according to different characteristics (Table 2) where fully agile means that all the development operation of the organization are carried out in agile methodologies only and partial agile means some of the projects/teams in the organization are proceeding with these methodologies.

Table 2 Dimensions of the organizations
Full size table

The Shapiro-Wilks test was used to check the normality of the data. A value of 0.120 indicates that the distribution is normal.

6 Results and discussion

6.1 KM adoption and organization size

Levene’s test for homogeneity of variance was also calculated at 0.102 assuring that the population variances by size are approximately equal. Descriptive analysis (Table 3) of the data shows that the medium size organizations show a tendency of least adoption of KM practice (69.0%) with the highest variance of 19%. Large organizations display the highest rate of KM adoption of 77.13% whereas the variation in the adoption was lowest in small organizations (7.5%). Small organizations exhibit the least variation 9.88% in adopting KM practices with minimum KM practices adoption of 60.44% and a maximum of 83% of the small organizations.

Table 3 Descriptive analysis of KM adoption level by organizational size
Full size table

The mean difference in KM adoption based on the size of organizations within each group is calculated with One-way ANOVA as depicted in Table 4.

Table 4 KM adoption by organization size
Full size table

One-way ANOVA is not found significant (p > 0.05), indicates no variation in the adoption of KM Practice level with the size of the organization, Therefore, hypothesis HA1 can be rejected based on a significant level. The discoveries of the research support the argument that regardless of size, knowledge issues apply to all organizations. Results contradict with the results of Connelly and Kelloway [11], Serenko et al. [36], Davis et al. (2007), and Zanjani et al. (2009) that apprehend no relationship between size and knowledge sharing in the organization whereas Xu and Quaddus [47] complements to our results and asserted that size does not affect knowledge systems. Our findings support Moffet and McAdam’s [28] argument that irrespective of organizational size issues related to knowledge does not apply to all organizations. The results confirm that the notion of KM applies to all organizations, with hardly any significant difference in the adoption of KM among the large, medium, and small-scale organizations.

6.2 KM adoption and area of engineering

Software engineering organizations can also be divided by the core area of software engineering: product development; consultancy; or both (product and consultancy). The non-significant value of Levene’s test (0.281) indicates that the population variation for groups is almost equal. Descriptive analysis in Table 5 shows the Mean and standard deviation for each group, where product development (PD) shares the highest adoption mean of 81.2%, whereas organizations providing consultancy and organizations providing software development and consultancy both are sharing almost similar adoption levels of 68%. The mean of KM adoption is the highest for product development organizations (81.23%).

Table 5 Descriptive analysis of KM adoption
Full size table

Results of one-way ANOVA (Table 6) do not show any significant disparity in the mean KM adoption. Although the confidence value is close to 0.05 it can be concluded that hardly any significant relation is found between KM adoption and working area of organization (p = 0.083) as depicted in Tables 7, 8 and 9. So, the H03 hypothesis can be rejected based on the results obtained.

Table 6 KM adoption practice and area modularity
Full size table
Table 7 Summary of KM practice level by team distribution
Full size table
Table 8 KM adoption: Co-located and distributed teams
Full size table
Table 9 Summary of KM practice level by type of organizations
Full size table

Outcomes from the study coincide with Sant (2011) on Indian IT consultancy service organizations where authors comment that the growth of knowledge management practices in India has been limited mainly in the service industries like IT, Consultancy, and some of the electronics and communication industries. There is hardly any difference in KM practices applied in software development and services companies. [8]. A similar study that overlap with the results of the author is on one of the Indian Giant software development and services Industry in comparison to the pure service industry from KM perspective was conducted. Major software companies like Infosys and Unisys have visualized the significance of the KM and imparted the KM system into the organizational structure that suits their working environment. A survey of these organizations does not highlight any significant difference in implementation of KM practices and both organizations have achieved a benchmark in implementing KM in their organizational structure [29].

6.3 KM practice level and team distribution

Globalization has given the advantage to the organizations to build their offshore offices and take advantage of the low-cost development of software from resource-rich countries like India. KM practice level of teams working in co-location and distribution was checked. KM adoption is highest for the team working in co-locations but has a greater variation in the adoption of KM practices as compared to teams working in a distributed environment. Variation in the adoption of KM practice is found high in co-located (23.93%) as compared to distributed agile teams (20.66%). A detailed analysis is given in Table 7.

Variation in KM adoption between distributed and co-located teams is determined by one-way ANOVA, before applying the test homogeneity of variance is calculated and was not significant providing confidence that variation of groups is of equal. Therefore, it is acceptable to use the t-test for equality of means to determine whether differences exist between the KM practices levels of distributed and co-located teams. The T-test is used to determine equality of means did not confirm any statistical difference between means of distributed and co-located teams. So, the null hypothesis (H04) that there is variation in the adoption of KM practices can be accepted. Testes do not confirm any variation in the adoption of KM practices by distributed or co-located teams.

Whereas results of Zenun et al. [48], Kroll et al. [23], Razzak and Mite [34] significantly differ from the present research. These Studies emphasized different approaches/practices for knowledge transfer and knowledge creation in distributed teams. Different strategies were implemented to manage knowledge locally and globally. One of the reasons for disagreement can be both of the research were focused on a single/ some of KM strategies: knowledge transfer [48] and Knowledge creation [23] rather than taking different dimensions of the KM, whereas Lindsjorn et al. [26] have taken other dimensions and a collective KM strategy for comparison between distributed and co-located teams and presents some agreement with the current study as the authors underline that there is hardly any impact of distribution on the management of tacit knowledge in agile teams [43].

6.4 KM practice level and type of organization

Agile software development practices are relatively new in software engineering environments. Term software engineering was coined around 1970, whereas agile alliance came into existence around 2001. The author of this research does not find any research that has focused on KM from the perspective, whether the organization is fully agile or partial agile. Most of the organizations are practicing both methodologies for product development, but when the profile of the organizations was selected significant number of organizations were encountered that explicitly acknowledge being fully agile companies, i.e. software was developed with agile methodologies only. So, it was important to judge KM from this perspective also. Software engineering firms are distinguished based on these technologies: Fully agile organizations mean organizations that are using agile practices in all of their software development projects and partially agile means organizations which are using agile practices to only some of their projects i.e. they are using both traditional and agile practices for different product developments (Table 9).

Fully agile companies show greater adoption of KM practices 80.95% compared to partially agile organizations which adopt 69.05% of KM practices, while partially agile organizations show a greater variation of 25.7% as compared to 13.97% of fully agile organizations. Variation in the adoption of KM by fully or partially agile organizations was conducted using an independent t-test. Value of 0.083 of Levene’s test gives the confidence that variation of groups by team distribution is almost equal (results in Table 10).

Table 10 Independent samples test comparing KM levels of type of organizations
Full size table

Statistical analysis suggests that the null hypothesis can be rejected as value confidence (p = 0.031) less than the threshold value. Therefore, the test confirms that KM adoption does vary between full and partial agile organizations (Table 10). This variation can be directly related to the claims of the researchers that KM practices are naturally embodied in agile practices whereas KM must be artificially induced in traditional development techniques.

6.5 The relationship among KM practice level and number of employees

The association between an Organization’s KM adoption and the number of employees is investigated from different dimensions such as organization size, core area, and type. Pearson product-moment correlation test is used to measure the association between KM adoption and the strength of employees in the organizations. No significant association among all sizes of SE organizations was found.

Statistics show that for the size A organization significant values come out to be 0.65, for size b it is 0.38 and for size cit was 0.33, all values above the threshold value of p (Table 11). Hardly any difference is found between KM adoption patterns of all sizes of organizations and the strength of the employees. Another area to evaluate is, whether there is an association between the type of organization and the strength of employees (Table 12).

Table 11 KM adoption and strength of organizations
Full size table
Table 12 KM adoption and area of working
Full size table

A test was conducted for organizations working fully in agile practices for every product development and organizations use agile practices partially (only for some of the product development). A significant relationship was found between fully agile organizations and employee strength of the organizations whereas hardly any association was found between partial agile organizations and the number of employees (p = 0.037 and 0.931, Table 13).

Table 13 KM adoption and type of organizations
Full size table

7 Fuzzy analytic hierarchy process

It is a combination of fuzzy set theory and Analytic Hierarchy Process (AHP) and considered as one of the most influential approaches for multi-criteria decision making (MCDM). Initially, the authors (Laarhoven and Pedrycz, 1983) have shown the effort to work on fuzzy ratios by using the triangular membership function (TMF). The popular fuzzy approach was given by the author (Chang, 1996) for extent fuzzy analysis, in this, he used the triangular fuzzy numbers for pairwise judgment scale. AHP offers a broad and balanced hierarchical structure for addressing decision problems on a common goal and related criteria. AHP helps quantify the weight of the appraised criteria in the form numeric basis. The criteria weight of each element determines its relative importance with the other elements of the hierarchy. Hence, it facilitates the decision-makers to identify and prioritize significant factors. FAHP has been used in this paper to find out the critical factors in adoption KM in an organization. Sub-criteria that have been identified form literature are globally ranked with the help of FAHP so that organization can understand the hierarchies of the factors that can be kept in the mind before implementing KM in agile process. Stepwise implementation of AHP process have been described below.

  • Step 1: The fuzzy pairwise assessment matrix is made. The matrix is shown in eq. 1.

$$ \overset{\sim }{A}=\left[\begin{array}{cccccc}1,1,1& \tilde{a}_{12}& \tilde{a}_{13}& \tilde{a}_{14}& \cdots & \tilde{a}_{1n}\\ {}\tilde{a}_{21}& 1,1,1& \tilde{a}_{23}& \tilde{a}_{24}& \cdots & \tilde{a}_{2n}\\ {}\tilde{a}_{31}& \tilde{a}_{32}& 1,1,1& \tilde{a}_{34}& \cdots & \tilde{a}_{3n}\\ {}\vdots & \vdots & \vdots & \vdots & \ddots & \vdots \\ {}\tilde{a}_{n1}& \tilde{a}_{n2}& \tilde{a}_{n3}& \tilde{a}_{n4}& \cdots & 1,1,1\end{array}\right] $$
(1)

where ã = (lij,mij,uij) and i = 1,2,3,4,5……n and j = 1,2,3,4,5……n are triangular fuzzy numbers.

  • Step 2: The fuzzy synthetic extent value (FSi) related to criteria i is calculated as:

$$ F{S}_{i=}\left(\ {\sum}_{j=1}^n{l}_{j\kern0.5em },{\sum}_{j=1}^n{m}_{j\kern0.5em },{\sum}_{j=1}^n{\mu}_{j\kern0.5em }\right)\ast \left(\frac{1}{\sum_{i=1}^n{\mu}_{i\kern0.5em }},\frac{1}{\sum_{i=1}^n{m}_{i\kern0.5em }},\frac{1}{\sum_{i=1}^n{l}_{i\kern0.5em }}\right) $$
(2)

  • Step 3: To calculate the degree of possibility of P1(l1, m1, u1) and P2(l2, m2, u2) triangular fuzzy numbers of P1 > = P2 is defined in eq. 3, eq. 4, and eq. 5 as:

$$ \mathrm{V}\left({\mathrm{P}}_1>={\mathrm{P}}_2\right)=1\ \mathrm{iff}\ {\mathrm{m}}_1>={\mathrm{m}}_2 $$
(3)
$$ \mathrm{V}\left({\mathrm{P}}_1>={\mathrm{P}}_2\right)=0\ \mathrm{iff}\ {\mathrm{l}}_1>={\mathrm{u}}_2 $$
(4)
$$ V\left({P}_2\ge {P}_1\right)=\kern0.5em \frac{{\boldsymbol{l}}_{\mathbf{1}-}{\boldsymbol{\mu}}_{\mathbf{2}}}{\left({\boldsymbol{m}}_{\mathbf{2}-}{\boldsymbol{\mu}}_{\mathbf{2}}\right)-\left({\boldsymbol{m}}_{\mathbf{1}-}{\boldsymbol{l}}_{\mathbf{1}}\right)} $$
(5)

The fuzzy weight is calculated as shown in eq. 6.

$$ {\mathrm{FW}}^{`}={\left({\mathrm{d}}^{`}\left({\mathrm{A}}_1\right),{\mathrm{d}}^{`}\left({\mathrm{A}}_2\right),{\mathrm{d}}^{`}\left({\mathrm{A}}_3\right),{\mathrm{d}}^{`}\left({\mathrm{A}}_4\right)\dots \dots .{\mathrm{d}}^{`}\left({\mathrm{A}}_{\mathrm{n}}\right)\right)}^{\mathrm{T}}\ \mathrm{where}\ {\mathrm{d}}^{`}\left({\mathrm{A}}_{\mathrm{i}}\right)=\min\ \mathrm{V}\left({\mathrm{FS}}_{\mathrm{i}}>={\mathrm{FS}}_{\mathrm{r}}\right)\ \mathrm{and}\ \mathrm{i},\mathrm{r}=1,2,3\dots \mathrm{n}\ \mathrm{and}\ \mathrm{i}\ne \mathrm{r} $$
(6)

The non-fuzzy weight or normalized weight is calculated as shown in eq. 7:

$$ \mathrm{FW}={\left(\mathrm{d}\left({\mathrm{A}}_1\right),\mathrm{d}\left({\mathrm{A}}_2\right),\mathrm{d}\left({\mathrm{A}}_3\right),\mathrm{d}\left({\mathrm{A}}_4\right)\dots \dots .\mathrm{d}\left({\mathrm{A}}_{\mathrm{n}}\right)\right)}^{\mathrm{T}} $$
(7)

8 Fuzzy analytical hierarchical process

AHP is used to calculate the relative ranks of the criteria that may help organization in implementing KM in it. Enablers are criteria used to achieve the goal i.e., implementing KM in an organization. AHP is used to demonstrate the relative and global ranking of all the criteria’s that can help in achieving the goal (Table 14). KM Technology is the most important criteria for enabling KM in agile organizations, followed by training and mentoring and knowledge sharing. Whereas the KM environment is found to be the least important factor in implementing KM in agile methodologies. It is surfaced by some authors that [39, 40] agile practices by default create an environment of knowledge management in the organizations so it not necessary to do something extra to feel like an inclusive KM environment.

Table 14 Ranking of criteria and sub-criteria
Full size table

It is worth mentioning here that the sub-criteria technological solution for transferring knowledge is the most important factor for successful implementation of KM in agile process among others, furthermore, the criteria KM Technology is ranked 1 among other criteria for implementation of KM in organizations working in Agile methodologies. Globally first and second sub-criteria for implying KM belongs to the KM technology group. Thus, we can say that ‘technology for implementing KM’ plays the most vital role in agile software development. After technological support, Human-centric IT is 2nd most important factor in the implementation of KM. This seems to be in concessions with the agile manifesto where the human-centric approach is given preference over tools and techniques. The third factor comes out to be knowledge acquisition through peers. Self-organizing team structure gives a boost to this factor as this is the best possible way to acquire knowledge through peers of self-organized and cross-functional teams. In KM environment criteria knowledge sharing culture is the most important sub-criteria whereas open spaces are the least important criteria. Similarly, in knowledge sharing, discussion forms come out to be the best practice for applying KS, and the Rotation of people among projects is the least important factor. It’s important to note here that this sub-criterion is also the least important criteria among all sub-criteria for implementing KM and is ranked at 17th place.

AHC helps understand the important criteria that can help in achieving a set goal. Ranking help in understanding the weightage that one can give while implementing KM in the organization working in agile methodologies.

9 Conclusion

The conceptual framework presented in this study has both theoretical and practical contributions, whereas on one side it provides researchers more insights on the embeddedness of KM in agile practices on the other side, it gives practitioners practical aspects of KM and agility. So, from the perspective of theoretical contributions, this paper provides a fresh viewpoint for viewing and synthesizing aspects that can influence the acceptance of KM and agile practices depending upon different characteristics of the organizations. From a practical viewpoint, this paper highlights the current knowledge of agile practices forms a KM perspective and adoption patterns of KM in agile practices among different organizational demographics. This can guide practitioners regarding adoption patterns of agility and KM, and they can reap the benefits by combining the best of both worlds. The study found a variation in the overall adoption of KM practice by organizations. Responding organizations have adopted, on average, 73% of the KM practices put forth in the survey. The standard deviation of 16% and 21.83% of the coefficient of variance indicates variation across organizations in the adoption of KM practices (total 29). More than 11% of organizations are using less than half of the practices, 35% of organizations are using KM practices ranging from 50% to 75%. The organization with minimum adoption has implemented only 30% of practices and the organization with the highest adoption rate has adopted more than 97% of the KM practices. To a surprise, most of the organizational characteristics are not found in the significantly associated with KM adoption except KM adoption level in full and partial agile organizations and the relationship between the organization KM adoption and the number of employees in the organization for product development and fully agile companies are found significant. Opposite to the claims of many researchers, this study does not find any significant difference between KM adoptions between distributed and co-located agile teams. At the end, fuzzy AHP has been introduced to prioritize all KM criteria as well as sub-criteria in implementing KM in agile organizations. The ranking of criteria in decreasing order is KM technology, Training and Mentoring, Knowledge sharing, Knowledge Acquisition and KM environment respectively. The most important and least important sub-criteria are Technical solution for transferring knowledge and Rotation of people among projects respectively.