Knowledge is a key competitive element for businesses. Yet, the ability for a company to effectively manage knowledge becomes even more important considering the strategy formulation. Thus, knowledge management (henceforth, KM) may be viewed as a way of understanding how the organization searches for alternatives that optimize business processes and achieve competitive advantage. Useful knowledge comes from business experience, analysis, research, innovation and creativity. In that sense, market knowledge, competition, customers, business processes, technology-and everything else that revolves around these elements-confer competitive advantage to the firm. As companies state their knowledge management requirements (by communicating what they look for and expect), a demand arises to “know what they know” and to effectively use knowledge in an effective manner (Davenport and Prusak, 2003).

There are three main streams of research in the Knowledge Management field. The first considers KM a process, whereas the second sees it in terms of the creation of knowledge for the organization. The third stream considers it an intangible asset (Boff, 2000; Davenport and Prusak, 1998; Nonaka and Takeuchi, 1997; Weggeman, 1997).

With respect to knowledge transference and sharing, Sveiby (2003) posits that some intangible assets do not only bring revenues to organizations: they also provide training to the employees, encourage the development of new skills, and improve the company’s image. In managing knowledge, one must see the business from a different perspective, since it produces effects that are not simply of a financial nature.

In the same vein, knowledge retention becomes a key concern for the organization; particularly when there are processes in place for structuring storage and retrieval for future use of such knowledge. Often, the retention of tacit knowledge is carried out through artifacts such as documents, files, papers, conversations, images, thoughts, software, databases, e-mail messages and retention policies. Knowledge retrieval is performed by individuals; thus, processes must stimulate certain behaviors and interactions. Capturing knowledge and business experiences is no simple task: it demands time and effort (Jamil, 2006; Lapa, 2004; Newman and Conrad, 1999).

The definition of memory is related to an individual’s ability to remember or retain acquired knowledge (Ferreira, 2004). Nonetheless, scholars point out that the term can also be applied collectively, by applying it to organizations under the various denominations of organizational memory (OM), collective memory or corporate memory. Much research work has studied the organizational memory, stating that it encompasses the processes of managing, storing and retrieving knowledge, which are deemed to be essential activities in contemporary business administration (Dieng et al., 1999).

The definition of OM involves complex questions regarding its multiple dimensions. Still, the concept is considered a viable approach, in line with the proposition of improving business processes (Bannon and Kuutti, 1996). Accordingly, OM should not be regarded as a form of storage and accumulation of information that works in isolation; rather, it is a process of instrumentalizing the retention of organizational knowledge, and it contributes to a continuous learning process that is supported by organizational experiences and decision-making efficiency (Choo, 2003; Nonaka and Takeuchi, 1997; Menezes, 2006).

According to Ackerman and Malone (1990), OM relates to the ability of the organization to benefit from its past experiences, in order to act effectively in the present. The authors highlight the use of technological tools for storage and retrieval of knowledge, as an integral part of OM. In addition, they point out that some of the key features of OM are repositories of internal information (such as corporate manuals, databases, filing systems and people’s own stories, i.e., forms of maintaining organizational knowledge in physical or digital form).

Therefore, the OM theory hinges on the mechanisms of storage and retrieval, which render it possible to reuse corporate and individual knowledge. This knowledge may be maintained in various repositories and is critical for the improvement of effectiveness in the organization (Abecker et al., 1998; Argote, 2013; Steil and Santos; Santos, 2012).

Extant research shows a practically unanimous consensus with respect to the manner that OM empowers the organization, in its quest to increase competitiveness. Its main concepts are founded on functions for preserving, recovering and using the previous business experiences. Putting it in other words, OM learns from history, and helps make new decisions.

According to Miranda (2008), knowledge ought to be understood as one of the key assets of the organization. Due to that, its integrity must be ensured through constant vigilance and concern. The organization must constantly develop mechanisms to facilitate the retention of consensual and integrated knowledge.

Table 1 summarizes the approaches found in literature, along with key elements and assumptions. It also presents the convergence of concepts found in literature, which will base our proposition for the model for software development companies.

Table 1
Overview of OM Approaches

Source: Prepared by the authors.

Considering the extant theory, despite the conceptual differences in the OM’s dimensions, we conclude that there is a conceptual convergence of the basic processes that encompass OM; namely, it pertains to knowledge storage, retention, use and access to information.

Based on this assumption, we selected a theoretical model for the development of the study, using Walsh and Ungson’s (1991) model, depicted in figure 1.

Figure 1.
Structure of Organizational Memory, defined by Walsh and Ungson (1991).
Source: Adapted from Walsh and Ungson (1991).

The model presents the internal components of the OM retention function: individuals (which refers to the members of the organization and their own experiences); the organizational culture (or forms of perception, symbols, sagas, stories, rumors that are transmitted to its members); transformations (which are built from past experiences, influence work practices and are constantly changing); organizational structures (which influence the behavior of individuals, and are related to the roles of individuals within the organization); ecology or physical structure (which helps shape and reinforce the behavior of individuals within the organization); last, the external component (related to the issue of the organization’s memory reaching beyond its limits of scope). Though not directly related to the organization’s internal memory, external sources could retain past information of the organization, which may be recovered.

The following session discusses the aspects related to innovation. We examine the main conceptual definitions used to explore the workings of the Organizational Memory (particularly the retention function), and the relationship with the OI dimensions in software development companies.

The term innovation was coined to describe the degree of novelty in a specific innovation. Alternatively, it may denote the degree of speed that an individual or organization implements something new (Garcia and Calantone, 2002). From an organizational perspective, innovation is defined as the ability to introduce or develop new processes, products or ideas.

Our initial approach to Organizational Innovation (OI) research considers the structural characteristics of an innovative organization (Mintzberg, 1979; Teece, 1998).

Conversely, innovativity is defined as the ability to create something new, or to initiate a movement towards change. It may be defined in several ways; one definition is the ability to introduce a new product in the market, or to open a new market through various arrangements and strategic orientations. It could also be defined as the capacity with which the company engages its innovations. A third and broader definition posits that it is the capability of introducing new products to the market, or opening new markets through a combination of strategic orientation and innovative behavior (Mohd, Bukhari and Hilmi, 2012).

Alternatively, innovativity is related to a dimension of entrepreneurial orientation. The concept plays a prominent role in the context of entrepreneurship, which relates it to the notion that firms must constantly seek to differentiate themselves from the competition. This concept promotes an entrepreneur outlook, insofar as strategy formulation and attitudes will support the development of new ideas, novelties, experiments and creative processes, leading to new products, services and processes (Lunpkin and Dess, 1996).

To understand the identification of the dimensions of innovation, we need to appreciate the processes that support the organizations’ capacity to innovate. Such factors may be seen from a resource perspective, with behavior and activities that are dynamically mobilized for the development of new products, processes and systems (Quandt, 2009). On the other hand, Tellis et al. (2009) posits that macroeconomic indicators related to labor force, capital, local culture and government regulations do not consistently define the manner that the radical innovation in firms take place.

Numerous studies have focused on the measures of process to evaluate the degree of innovation in firms (Beyhan et al., 2009). In any event, Quandt, Bezerra, and Ferraresi (2015) postulate that these measures are insufficient in gauging the elements that lead to innovation. Accordingly, there is a need to investigate the conditions that enable the organization to be innovative. In this sense, the authors proposed a theoretical model comprised of ten dimensions to evaluate and identify organizational innovation.

Their model comprises an integrated and dynamic set of capacities, behaviors, processes and activities which promote the innovation. The model derives from the extant research, and uses these dimensions in a recurrent manner, as shown in Table 2 (Quandt, Bezerra and Ferraresi, 2015).

Table 2
Measurement dimensions for the Organizational Innovation.

Source: prepared by the authors, based on Quandt, Bezerra and Ferraresi (2015).

Based on the model proposed by Quandt, Bezerra and Ferraresi (2015), we will establish conceptual associations between the ten OI dimensions: Strategy; Leadership; Culture; Organizational Structure; Processes; People; Relationship; Technological Infrastructure; Measurement and Learning. We will also assess the extent to which OM influences each of these dimensions.

This session presents the main research results. During the data collection period, a link to the online questionnaire was sent. An invitation to participate in the survey was reiterated in three instances. Of the 391 online questionnaires sent, 64 questionnaires were effectively answered. The model was subjected to a validation analysis of the OM and OI variables.

With regards to the characterization of the companies, twelve questions were built with the purpose of contextualizing the time that the company has been in business, the area of ​​coverage and business segment; these questions were the part of the data collected in the first block of the questionnaire, dealing with the respondent’s profile and firm performance, as characterized by the company size (i.e., the number of employees and gross annual sales turnover). Regarding the respondents, the research sought to verify the position, age, gender, education and tenure in the firm.

We note the predominance of male respondents who occupy managerial positions, conditions that encompass 86.2% of the cases. Another relevant data that the research identified refers to the participant’s education: most were graduates, and many held a postgraduate degree. Table 4 presents a summary of the main characteristics of the respondents, and their respective companies.

Table 4
Synthesis of the characterization of the companies surveyed.

Source: Prepared by the authors, primary research data.

Regarding the characteristics of the firms, a significant number of companies has up to 49 employees (83.1%). Interestingly, when considering the time the company has been in business, we note that the percentage of companies that have been operating at least five years is 78.5%. This suggests a scenario of stabilization in terms of the quantitative growth of its employees. As for the business market, the survey reveals that there is an average performance of companies in more than one simultaneous market. It should be noted that the most representative sector is services and which demands the companies surveyed, and they represent 59.7% of the respondents.

In general, the characteristics of the software developer companies in the present study are predominantly micro and small companies. This fact is justified by the profile of this segment, that is: of the companies that work in the software development and production, approximately 93 % are categorized as micro and small enterprises (Abes, 2015).

Another result to be highlighted is the performance in the service sector, which represents 59.7% of the companies that direct their solutions to this target segment. The time of operation is also significant, since 78.5% of the companies have been active for more than seven years in the market. This can be considered an indicator of stability of these organizations.

In the next section, we present the analysis of the OM and OI results dimensions.

A statistical analysis involving the six components of OM and the ten dimensions of OI was carried out. Operational variables were defined for each of the six components of the OM retention. For the “Individual” component, ten variables were defined; for “Culture”, seven variables; for “Transformation / Process”, four variables; for the “Structure”, three variables; for “Work Environment”, three variables; and for the “External Files” component, three variables were evaluated.

In the OI dimension, the questions were grouped into variables of the ten dimensions of OI. For the “Strategy” dimension, seven variables were analyzed ; for “Leadership”, there were six variables; for “Culture”, six variables; for the “Organizational Structure”, five variables; for “Processes”, four variables; for “People”, eight variables; for “Relationship”, four variables; for “Technological Infrastructure”, five variables; for “Measurement”, five variables and, for the “Learning” dimension, five variables.

In order to perform the confirmatory factor analysis procedure, we analyzed the measurement model and used the SmartPLS software to calculate the Convergent Validity and Discriminatory Validity of the data. Three rounds of statistical treatment were required. For each round, we adjusted the indicators that presented discrepancy with respect to their results, based on the stipulations found in literature.

Table 5 shows the values ​​after the second round. We note that the OM indicator related to the Individual presents an AVE value slightly below 0.5-accordingly, the third round sought to adjust that variable.

Table 5
Adjusted data quality scores from the SEM of the second round of the research tests.

Source: Prepared by the authors.

Thus, from the results of the Confirmatory Analysis, we reach validation for the Measurement Model. Table 6 shows the AVE and CC values ​​for the second-order latent variables for the Retention Function and OI. We note that, for both variables, the AVE value is higher than 0.5. According to Fornell and Larcker’s (1981) criteria, we conclude that the model converges in a satisfactory manner.

Table 6
AVE and CC of second-order latent variables.

Source: Prepared by the authors.

Next, we evaluate the Composite Reliability (CC), which also assesses the reliability of the model. To obtain satisfactory validity, the values should be above 0.7, according to Hair et al. (2014). Inspection of Table 6 shows that the values ​​of 0.8956 for the second-order latent variable of the Retention Function, as well as 0.9518 for OI, ​​are well above the requirements, and are therefore deemed adequate and significant for the measurement model.

Subsequent to the validation of the Measurement Model, we analyzed the structural model, using Pearson coefficients (R²) to evaluate the portion of the variance of the endogenous variables that are explained by the structural model (Ringle et al., 2014), shown in figure 2. Our results considered the configuration of the SmartPLS software to run the bootstrapping function, with 1,000 samples.

Figure 2
Structural Model without the use of control variables.
Source: Prepared by the authors.

Inspection of Table 7 shows that the model’s R² is equivalent to 49.6%. We conclude that the model has a great predictive effect, since the model does not include control variables. Another important result is the P Value: to be statistically significant, this value should not exceed 0.05. Our result is 0.000, and this characterizes a highly statistically significant model. Now looking at the structural coefficient (0.704), its interpretation means that, for each standard deviation that increases in the retention function, the OI will tend to increase by 0.704 standard deviations.

Table 7
Impact of OM on OI without control variables.

Source: Prepared by the authors.

In order to establish greater consistency of the proposed model, some tests with control variables ensued, first individually, and later running the global model. Table 8 depicts the results by simultaneously testing three variables: OM impacting in OI, companies in business for 10 years or more, and billings over R$ 3,600,000.00. We note that, for this test, the structural coefficient reduces from 0.704 to 0.632. This result indicates a limited effect in the relation between the control variables and the conceptual model that was tested.

Table 8
Impact of OI from the pooled analysis with the Retention Function and the two control variables

Fonte: Prepared by the authors (2016).

According to the results in table 8, the values ​​of (f²) had a high effect size for the OM (Retention Function) towards OI. The value obtained is 0.794, well above the minimum required to characterize a relevant effect for the model. However, for the control variables tested, the effect size values ​​of (f²) were not significant.

From this study, we can identify an important relation between the Organizational Memory and the organizational aspects related to Organizational Innovation. Overall, the analyzed model accounts for 49.6%. In other words, OM variables are likely to produce an effect in the dimensions related to OI. The results are considered statistically significant in ​​applied social sciences areas, and characterizes a high effect index.

In the proposed model, we notice that there is a significant relationship between OM and OI, with a structural coefficient of approximately 0.704. This means that, for each increment in the OM retention function, the organization will improve its performance in terms of innovation according to this proportion.

We verify that companies that maintain and foster organizational environments that stimulate practices and actions for the transposition of knowledge understand the importance of the elements of OM. They will develop and strengthen mechanisms for the acquisition, storage and retention of information, while they seek socialization and exchange of organizational knowledge. This chain of events ends up being highlighted in the scope of innovation, since it stimulates the creation of an innovative organizational environment, which is, in turn, synonymous to better organizational performance and competitiveness.

The purpose of this research study was to establish the relationship effects between OM and OI in software development firms. In practice, the investigation shows that there is a significant relationship between OM and OI. This finding is supported by the validation of the measurement model and based on the structural model. The analysis of the data collected confirms the robust statistical results, and the validity of the proposed model in this research.

We also note that, as a company reaches its market consolidation (i.e. a longer time in business), the retention function plays an instrumental role in increasing the organizational innovation capacity. This fact can be observed from the statistical model tested, when we see a difference in the effect of the company’s performance after 10 years in business.

Another issue that deserves to be highlighted is the set of practices evaluated through the data collection instrument itself. These practices can, on their own, help managers to adopt best practices leading to environments that are more conducive to fostering a culture of innovation. This can be verified from the actions directed to the practices of the OM retention function, which are directly related to the variables that influence and promote posture of innovation in the Software Development organizations.