A growing number of research studies confirm the superior role of communication management in influencing the results of information technology (IT) projects. Oz and Sosik (2000) identified five major factors that lead to cancellation of IT projects, with one of them being poor communication. Similarly the Standish Group (1998) compares project success with a three-legged stool, with one of the three legs being communication. Research on the influence of project management knowledge areas, such as those defined in the Project Management Institute’s (PMI®) A Guide to the Project Management Body of Knowledge (PMBOK® Guide), on Earned Value results showed that communication management has the strongest impact of all knowledge areas on cost variances in IT projects (Müller and Turner 2001). Hutlinski et al (2001) identified a possible reason for this by showing that complexity levels of communication are highest during requirements specification stages and the implementation stages of IT projects. In their research more than 50 percent of the respondents indicated moderate to great communication difficulties during these stages. This is complemented by Keller’s (2001) observation that a project team’s external communication has a larger impact on financial results than project internal communication. Effectiveness in a project team’s external communication is achieved through alignment of the communication needs with the task at hand. Tushman (1978) showed that successful organizations are flexible in adjusting their communication practices to the needs of the projects or tasks. This avoids misunderstandings, which lead to rework and increased project costs. The criticality of balancing social and procedural aspects of communication at the interface between organizations was identified by Allen et al (1980). His research showed that performance in technical service projects is higher when the project manager assumes primary responsibility for coupling the project to other parts of the organization. However, empirical studies showed that projects often lack good communication beyond the boundaries of the project team (Partington 1997). Research by Ibbs and Kwak (1997) identified IT firms as being especially low in their communication maturity when compared with firms from engineering and construction, information management and movement, as well as high tech manufacturing industries.
The study described in this paper builds on this prior research and investigates differences in choices for communication media, contents, and frequency between buyer and seller firms’ communication in IT projects. The unit of analysis is the formal communication that takes place between a project sponsor of a company buying an IT project from an external vendor, and the project manager responsible for the implementation stage of the IT project sold by this vendor.
In these project related buyer-seller relationships the IT vendors establish project teams as temporary agencies to achieve project specific objectives, and appoint project managers to manage the projects on their behalf (Turner and Müller 2002). Within these interorganizational settings the project manager’s organizational position relative to the employing organization, the customer and project, is one of in-between home organization and customer, but close to the project. For the project manager this creates a perspective of “customer-focused delivery of what is contracted” towards the project, which is different from the buying firm’s project sponsor perspective who perceives the same project from an intracompany change management perspective (Partington 1997). Being the agency for change the two managers have to integrate different requirements, stemming from their different perspectives, by balancing their respective organization’s internal and external, as well as project specific requirements. This balance requires communication between the parties.
The research described in this paper shows the differences in choices for communication by project manager and sponsor, and analyses how these choices are influenced by the relational norms between the organizations, their respective organizational structure, and the project specific risks. Through that the external, internal, and project specific influences on the project manager’s and sponsor’s communication choices in IT projects are analyzed. The research question can be stated as follows:
Exhibit 1. Research Model
Are there differences between project sponsors and project managers’ choices of appropriate communication media, frequency and contents during IT project implementation, and how do organizational structures, relational norms and project risks influence these choices?
Communication, such as in IT buyer-seller relationships, has been addressed from many different research perspectives. This study here uses established models from organizational communication theory and from organization theory to investigate their applicability in the transient context of customer delivery projects, where teams are formed for the period of time necessary to develop and deliver an IT project to a seller firm’s customer.
Definitions of communication are manifold and range from the interchange of information to that of a social discourse (Oxford 1996). This is also reflected in the changing nature of communication in the project management literature. Here earlier definitions of communication management referred to mechanistic sender-receiver structures (Project Management Institute 1987). More recent definitions refer to the dynamic processes for planning, information distribution, performance reporting, and administrative closure, together with the social and integrative characteristics of communication, which link people, ideas, and information necessary for project success (Project Management Institute 2000, 117).
The following section introduces the research model, followed by identification of research variables and the development of research hypotheses.
Research Model
The research model is build of two related constructs to measure:
1. The difference between project manager’s and sponsor’s choices in communication media, frequency, and contents, used in formal buyer-seller communication in IT projects
2. The influence of external, internal, and project specific factors on these choices.
Measures were taken independently by project manager and sponsor to allow for a subsequent gap analysis. The items measured and their classifications derived from a review of project management related literature and its suggested communication practices, where applicable existing research models from communication theory were used (e.g., Rice and Shook 1990; Russ, Daft, and Lengel 1990).
The second construct builds on prior research, which identified organization structure and coordination mechanisms as the two dimensions effecting information-processing capacity in organizations (Tushman and Nadler 1978; Huber and Daft 1987), and adds risk as a project specific dimension.
Following that three independent variables are used, namely organization structure (of the project team, and buyer firm’s organizations respectively), Relational Norms from Transaction Cost Economics (TCE), and project specific risk. The appropriateness of TCE for research in project based buyer-seller relationships was recently demonstrated by Turner and Keegan (2001) through their research on governance structures in project driven organizations and in functional organizations. The scope and perspective of TCE matches the unit of analysis within this research by focusing on the individual transaction, i.e., the project, the associated governance structure, and the project manager-sponsor interaction. Project specific risk, the third independent variable, was identified in prior research as an influential factor on communication in project management because of the need to mitigate risk through increased communication between project managers and sponsors (Turner and Cochrane 1993).
The underlying axiom of the research is that organization structure, relational norms, and project risk are determinants for structural operation mechanisms, which are expressed as project manager and sponsor’s choice in appropriate communication media, frequency and contents, (see Exhibit 1).
On the basis of this model the following research questions are investigated:
Q 1.0 Is there a gap between project manager’s and sponsor’s choices for communication in IT projects?
Q 2.0 Are project manager’s and sponsor’s choices of communication media, frequency, and contents impacted by the situational factors of organization structure, relational norms, and project risks, i.e.:
Q 2.1 Is there a correlation between organization structure, at either seller or buyer organization, and their respective choice of communication media, frequency, and contents?
Q 2.2 Is there a correlation between relational norms in the project manager-sponsor relationship and the choice of communication media, frequency, and contents?
Q 2.3 Is there a correlation between project risk, when measured as unclearness of goals and methods, and choice of communication media, frequency, and contents?
Q 2.4 What is the relative impact of the three independent variables of organization structure, relational norms, and project risk on the choice of communication media, frequency, and contents?
Q 3.0 What are the dominating patterns in communication between project managers and project sponsors?
Communication Elements
Based on a review of project management literature and its suggestions on formal communication between project manager and sponsor the following classifications were made.
Communication Frequency
This refers to the number of formal communication events between project manager and sponsor. It includes all media and possible communication contents. The literature generally recommends communication frequencies at either variable intervals, such as at milestone achievement, or at fixed intervals, like biweekly, or continuous communication, such as daily emails or verbal updates.
Communication Contents
This refers to the information content exchanged between project manager and sponsor at each formal communication event. The following content is suggested by the project management literature:
• Status and achievements, i.e., project plan adherence in terms of schedule, budget, and implemented functionality
• Changes to the project, i.e., in scope, plan, risks, quality requirements, and so on
• Issues and open items list
• Definition of next steps in the project
• Analysis and communication of trends
• Quality and progress measures, i.e., earned value analysis and quality metrics.
Communication Media
Communication media refers to the way a message is conveyed from project manager to sponsor and vice versa. Project management methodologies (e.g., CCTA 2000) suggest use of written media, whereas some authors also recommend face-to-face meetings for formal communication events.
This study uses classifications predominantly used in communication theory, i.e., face-to-face, verbal (or oral), and written communication (e.g., Russ, Daft, and Lengel 1990; Rice and Shook 1990) to identify prevalent media choices.
Differences between Communication Expectations and Capabilities
Openness or congruence in interpersonal communication is critical for openness between the communicating parties. Interpersonal relationships between individuals are impacted by the level of congruence of experience, awareness, and communication, to the extend that congruence leads to more accurate mutual understanding, improved psychological adjustment, and mutual satisfaction in the relationship (Dance 1967). Individuals in organizations have difficulties anticipating the communication needs of their communication partners. Communication patterns in organizations are based on the organization’s “sensable” representations of their external environment, i.e., a perception of the environment and not the environment itself (Huber and Daft 1987). A finding supported by the theory of Autopoeisis and Cognition (Maturana and Varela 1980) which states that externally triggered activities within a living system are not triggered directly by the outside world, but by the living system’s internal representation of the outside world. In case of project manager-sponsor interactions it means that the contents of communication is based on an idea one party has about the other party’s expectations, but not on the real, firmly stated communication expectations of the partner. Meanings and opinions of individuals in organizations are unique and shaped by an individual’s position in an organization, therefore meanings of individuals cannot be identical and “shared” understandings are at best transitory (Shulman 1996). Given these findings from organizational communication research it is unlikely that a high level of congruence in communication between project manager and sponsor can be found.
Relational Norms, Organization Structure, and Project Risk as Antecedents for Choices in Communication Media, Frequency, and Contents
Relational Norms
Transaction Costs Economics (TCE) provides a general perspective to buyer-seller relationships, which involves not only the exchange of products and services for money, but also social interaction and other interactive processes. It investigates the costs of drafting and negotiating contracts, as well as monitoring and enforcing these agreements (Rindfleisch and Heide 1997). A specific relational TCE dimension was developed by Heide and John (1992) through their concept of Relational Norms. Prior research by Müller (2002) on the impact of relational norms on the communication preferences of mainly European IT project managers showed a strong positive correlation between relational norms and communication frequencies, preferences for written status reports and importance of communication media. This research here extends these findings both in terms of geography and project manager-sponsor perspective.
Organization Structure
Organization structure is the way the organization is built up, and the way the relations and relationships between people in an organization are regulated (Warner 1996). Most of the existing models of organization structures in permanent organizations define a continuum from bureaucratic to organic structures. Related to the increase in organic structures is an increase in communication between and within the organizations (Mintzberg 1982; Morgan 1996; Burns and Stalker 1994). Organic organization structures, when compared with mechanistic structures, show more ambiguous and equivocal messages due to the need for increased horizontal communication. While bureaucratic structures rely on written reports with predefined contents, organic structures require more interactive communication media. Members of bureaucratic organizations are expected to have a clearer understanding of the specialized issue they are communicating. Therefore communication contents are reduced to numbers or other measures (Burns and Stalker 1994). Given that, it is expected that the emphasis on quantitative measures like quality metrics and earned value results, will be higher in bureaucratic structures.
Project Risk
The link between communication and project risk was identified by Turner and Cochrane (1993). They classified project risks by the level of unclearness of project goals and unclearness of methods to achieve these goals, combined with the need to reduce these risks through communication. The level of equivocality and ambiguity in stating project objectives, together with the uncertainty caused by the difference between the information needed and the information available for project related decisions, will impact the media choice in communication between project manager and sponsor (Russ, Daft, and Lengel 1990). According to Media Richness Theory it can be expected that projects with higher risk levels use more frequent and more personal communication, whereas lesser risk levels will require more “routine” communication, using written reports at fixed intervals like biweekly or monthly reports.
Project management literature, especially project management methodologies, suggest written media for communication between project manager and customer. Therefore it is expected that methodology clearness has a significant impact on use of written media.
Based on this literature review the following hypotheses are established:
H 1: There will be a significant difference between communication contents expected by project sponsors in buyer firms and the content chosen by project managers in IT seller firms.
H 2: There will be a positive relationship between the relational norms in the project manager-sponsor relationship and routine communication patterns, expressed as a preference for written status reports and meetings at fixed intervals.
H 3: There will be a positive relationship between the extent of organic organization structure and communication frequency. The more organic the organization, the higher the frequency in communication.
H 4: Bureaucratic organization structures will favor written reports, while organic structures favor more interactive media like oral communication or face-to-face meetings.
H 5: There will be a negative relationship between the extent of organic organization structure and the use of quantitative measures, like quality metrics and earned value numbers.
H 6: There will be a positive relationship between message equivocality and media richness; rich media (e.g., face-to-face meetings) will be selected for communication in projects with unclear goals, and lean media (i.e., written communication) for projects with clearly defined goals.
H 7: There will be a positive relationship between clearness of methodology and use of written media for communication.
Research Design
The research comprises confirmatory and empirical parts. The former part tests the hypotheses stated above in the transient environment of IT projects, and the latter part identifies dominant communication patterns between project managers and sponsors.
Focus groups with representatives from buyer and seller firms validated the research model and questionnaire. Data were independently collected from project managers and project sponsors through an Internet-based survey. Multivariate data analysis techniques were applied to test the hypotheses and to develop a model for project manager-sponsor communication in IT projects.
Questionnaire Design
Established instruments were used whenever possible. Organization structure was measured along the bureaucratic-organic scale developed by Burns and Stalker (1961) and operationalized by Khandwalla (1977). Data on relational norms were collected using the set of questions from Heide and John (1992). The third independent variable, project risk, was measured through three questions on methodology clearness and three on objective clearness.
The three dependent variables media selection, communication frequency, and contents were assessed using the suggestions from the literature review. The development of questions and scales followed Churchill’s (1999, 392) process and the suggestions from the focus groups. The classification into written, verbal, and face-to-face communication was adapted from prior research by Rice and Shook (1990). All measures were done using seven point Completely Disagree to Completely Agree Likert scales.
Communication frequencies were individually measured for written, verbal, and face-to-face communication using items for: daily, weekly, biweekly, monthly, at milestone achievement, at project phase or project end, no communication.
Communication contents was also assessed individually for written, verbal, and face-to-face communication using the items: project status and achievements, earned value and quality measures, issues or open items list, changes to project, trends, next steps, other contents.
The design allowed for all possible combinations of dyadic communication between project manager and sponsor, e.g., daily verbal updates with monthly written reports and face-to-face meeting at milestone achievement.
The eighty-four items questionnaire was pretested using sixty-eight project managers and sponsors. A response rate of 26 percent was achieved.
Sampling Frame and Sample
The final survey took place from 10 September 2001 to 2 January 2002. The web-link to the buyer and seller survey was distributed to 267 PMI Chapter Presidents and Special Interest Groups, 30 Managers of IPMA country organizations, as well as the online project management newspaper PMNet. All respondents were asked to distribute the web-link among their members. Approximately 600 project managers were addressed within NCR Corporation. Five PMI/IPMA organizations responded that they posted the survey on their website. PMNet published the link in their November and December issue. Due to the snowball approach of distribution a response rate cannot be calculated. However, the provider of the Internet survey tool tracked a “hit rate.” It showed that 29 percent of those who accessed the survey also answered it.
The number of responses totaled to 209, of which 200 were used for analysis. The sample used for analysis consisted of 62 project sponsors and 138 project managers. Respondents came from 34 countries, 47 percent of the respondents were from Europe, 34 percent from USA and Canada, and 19 percent from other regions. The average age of respondents was 42 years, with 20 years of business experience, and 10 years of project management experience. Project sponsors had in average 6 years of project sponsorship experience.
Method of Analysis
Content validity for the measures was established through careful selection of validated measures, based on extensive literature review. Latent dimensions represented by the items were extracted using backward factor analysis with Varimax rotation for the multi-item constructs of communication frequency and contents. This process followed the recommendations of Hair et al (1998). Three frequency factors were identified, namely:
• Variable interval communication, i.e., at project milestone or project phase achievement
• Fixed interval communication, i.e., at biweekly or monthly intervals
• Continuous communication, i.e., daily or weekly communication.
Four content factors were identified, which also reflected the media usage for the different content types:
• Personal review, i.e., face-to-face meeting covering all contents variables
• Project analysis, i.e., earned value measures, quality metrics and trends using all media types
• Written status reports with possible follow up, i.e., written communication about status, achievements, issues, changes, next steps, and other items, including a verbal or personal follow-up on other items if necessary
• Verbal update, i.e., an oral update on status, issues, changes, and next steps.
Constructs with less than five items were averaged, as recommended by Hair et al (1998). Relational norms and organization structure constructs were each averaged as in earlier studies.
Differences between project manager and sponsor responses on the communication factors were identified using t-tests and One-Way ANOVA techniques. The impact of organization structure, relational norms, and project risk on the contents and frequency factors was analyzed using regression techniques on the basis of total sample, as well as independent for buyer, seller, and the three different geographies of Europe, North America, and other regions. Communication patterns were analyzed using canonical correlation analysis, which allows taking into account the interdependencies of the dependent variables when they are impacted by a set of independent variables. Significance levels were set at 0.1.
Results
Total Sample Analysis
The main difference in communication choices between project manager and sponsor was found in their preference for project analysis. Project sponsors from buyer firms showed a significantly higher preference (p < 0.005) for project analysis, i.e., use of earned value measures, quality metrics, and the development of trends. Hypothesis 1 can be accepted; there is a significant difference in one of the seven factors measured. Hypotheses 4 and 6 can also be accepted because the level of organic organization structure is negatively correlated with written status reports, and the clearness of objectives is positively correlated with the use of written media.
Selection of appropriate communication media, frequency, and contents is mainly impacted by project risk factors and to a lesser degree by organization structure. Hypotheses 2, 3, 5, and 7 can be rejected. On a total sample basis there is neither indication of relational norms impacting the choices for communication, nor organization structures impacting communication frequencies or choice for project metrics.
Analysis by Buyer and Seller
Individual analysis of the buyer and the seller samples revealed the different impact of demographic and situational factors on the perceived appropriateness of communication media, frequency, and contents.
Project sponsor’s choices are mainly influenced by demographic variables like work experience and project management experience, and only to a lesser extend through relational norms and objective clearness. An increase in work experience of sponsors reduces both communication frequency and the choice for personal reviews. However, years of project management experience are directly related to preferences for written status reports. Higher relational norms increase the choice for project analysis. Levels of risk are positively related to personal reviews and negatively to the use of written status reports.
From a buyer perspective only Hypothesis 6 can be accepted, because objective clearness is negatively related with use of personal reviews and positively with written status reports. All other hypotheses can be rejected.
Project managers choices are mainly impacted by project risk variables, and to a smaller extend by organization structure.
Clearer objectives lead to more written status reports, less analysis, and personal reviews. Continuous communication is increased through methodology clearness, but decreased through objective clearness and organic organization structures. Only Hypothesis 6 can be accepted, clear objectives lead to use of less rich media, e.g., written reports. All other hypotheses can be rejected from a seller perspective.
Analysis by Geography
Individual analyses for North America, Europe, and the sum of other countries showed that communication choices differ by geography. Communication between European project sponsors and managers is mainly influenced by relational norms, prior experience, and buyer seller perspective. In North America it is mainly impacted by project risk and to a smaller degree by project management experience. Within other countries a mix of experience, risk, relational norms, and buyer-seller perspectives impacts the dyadic communication.
High relational norms in Europe are related to fixed communication intervals, personal reviews, and written status reports. Therefore Hypothesis 2 can be accepted. Significant differences prevail between sponsor and manger in the appropriateness of project analysis. Sponsors prefer analysis, whereas managers don’t. From a European perspective Hypothesis 1 can be accepted and Hypotheses 3 to 7 rejected.
From a North American perspective choices for communication are set by project risks. Clear objectives reduce the need for continuous communication and increase use of written status reports. Methodology clearness increases continuous communication and decreases personal reviews. Project management experience is positively related with project analysis. From a North American perspective Hypothesis 6 can be accepted, there is evidence that clearer objectives lead to selection of leaner media. All other hypotheses can be rejected.
In the other countries communication choices are different between buyer and seller when it comes to project analysis. As within Europe, buyers prefer analysis. Increasing relational norms, objective clearness, and experiences reduce the communication contents and frequency. However, increased work and sponsor experience lead to increased choice for analysis and variable interval meetings respectively. Clearer methodologies increase the choice for continuous communication. Hypothesis 1 can be accepted. There is evidence of significant differences between buyer and seller, while all other hypotheses are to be rejected.
Patterns of Communication Media, Frequency, and Contents
On total sample basis methodology clearness, relational norms, and objective clearness are associated with fixed communication intervals and written reports. These associations are measured as positive loadings of canonical variates. The dependent variables of personal review, project analysis, and continuous communication are negatively associated with the former set of independent variables. In summary, the relationship between the situational variables of organization structure, relational norms, project risk with the choices for communication media, frequency and contents can conceptually be displayed on a continuum ranging from “routine delivery projects” to “joint implementation projects.” Exhibit 2 shows the model of association. “Routine” projects, are associated with low risks (i.e., clear objectives and methodology) and higher relational norms, using written status reports at biweekly or monthly intervals. Projects with higher risk levels and weaker relationships between buyers and sellers are associated with more detailed analysis, personal reviews, and continuous communication, thus a higher engagement level of customers.
Exhibit 2. Descriptive Model of Buyer-Seller Communication
Conclusion
There is no single answer to the question of how project managers from IT seller firms communicate with project sponsors from buyer firms. A survey of 200 project managers and sponsors showed that choices in communication media, frequency, and contents vary by project manager-sponsor perspective, and geographically between North America, Europe, and other regions. The research questions posed at the beginning can now be answered as follows:
Q 1.0 There is a significant difference in project manager and sponsor’s choices for project analysis, i.e., in using earned value techniques and quality metrics for project tracking and development of trends. Even if project objectives are clearly stated buyer firms expect the use of earned value methods and quality measures to track progress and identify trends. Seller firms put low priorities on these activities.
Q 2.0 Are project manager and sponsor’s choices of communication media, frequency, and contents impacted by the situational factors of organization structure, relational norms, and project risks?
Q 2.1 There is evidence on the level of total sample that bureaucratic organization structures prefer written communication. However, this result did not show up in analysis of subsamples. No evidence was found that organization structures impact communication frequencies or more “narrow” communication contents like earned value and quality measures.
Q 2.2 Several statistically significant measures provided evidence that relational norms impact the choice of communication frequency, contents, and media in Europe, but not in other parts of the world. European project manager-sponsor communications are much more based on mutual relationships than in other countries.
Q 2.3 From a total sample perspective risk is clearly the dominating influence on communication choices of project managers and sponsors. Four of the seven communication factors are directly impacted by project risk. However, the impact on project managers is much higher than on project sponsors. From a geographical view risk does not impact the communication choices in Europe, but in all other parts of the world.
Q 2.4 From a global perspective project risk showed the largest relative impact of all situational variables. Buyers, however, are mostly driven by demographic variables, such as project management experience and work experience. The impact of organization structure and relational norms on a global basis is minor. From a geographical perspective organization structure has no impact in any part of the world. Relational norms impact communication in Europe only, and project risk determines the communication in all non-European countries.
Q 3.0 Patterns were identified through association of written reports at biweekly or monthly intervals with the situational variables of objective clearness, relational norms, and methodology clearness. These projects represent relatively stable communication relationships. The choice for continuous communication, deep project analysis, and personal reviews is negatively associated with the situational variables, thus representing a higher engagement level of the buyer, which possibly indicates joint implementation activities or increased control by buyers due to troubled projects.
This study addressed an often-cited weakness of project management research, i.e., the inward looking perspective of the existing normative project theories, and a general incoherence of project organization theory (Ekstedt et al 1999). For that the study was based on established theories and methodologies originally developed in the context of permanent organizations. These theories were investigated for their appropriateness in the context of temporary project organizations. The results show that organization theory is not a good predictor for communication in IT project based buyer-seller relationships. Relational norms, as defined within Transaction Cost Analysis are a good predictor for communication between project managers and sponsors in Europe, but not in other parts of the world. Media Richness Theory and the project management based theory of risk, measured by clearness of objectives and methodologies, are most appropriate to predict communication behavior.
The empirical part of the research showed that communication between buyers and sellers in IT projects can be seen in a continuum, with one end representing “routine delivery projects” using biweekly or monthly written reports in projects associated with low risk levels and good buyer-seller relationships. Whereas at the other end “joint implementation projects” with higher risks and more formal relationships are associated with continuous communication, personal meetings and detailed analysis.
The findings are equally important for project managers and sponsors, because it allows both to identify the expectations of the communication partner, and the drivers thereof, e.g.:
• Communication frequency increase caused by higher relational norms in Europe, versus an increase caused by higher project risks in North America.
• Communication contents expectations of buyers driven by business and project management experience together with a desire for analysis and trends, versus seller firms’ choices driven by project risk.
• The indication that stable, routine communications at fixed intervals are associated with low risks and good relationships, versus increasingly high communication frequencies, reviews, and analyses being indicative of deviation from stability in the project.
This knowledge allows anticipating communication practices of business partners and to improve efficiency in project related buyer-seller communication. The importance of this efficiency for project results was identified in the beginning of this paper.
Researchers in project management can benefit from this study through its identification of appropriate theories for project based buyer-seller relationships. Future studies should relate the findings to project results, as well as cultural factors that affect communication. This will further broaden the understanding of interorganizational communication in IT projects and add to the congruency in sponsor’s and manager’s choices for formal communication during IT project implementation.
