Ralph Katz.

An investigation into the managerial roles and career paths of gatekeepers and project supervisors in a major R&D facility online

. (page 1 of 2)
Online LibraryRalph KatzAn investigation into the managerial roles and career paths of gatekeepers and project supervisors in a major R&D facility → online text (page 1 of 2)
Font size
QR-code for this ebook


An Investigation Into the Managerial Roles and Career
Paths of Gatekeepers .md Project Supe3~visors in a Major

RSD Facility*

Ralph Katz
Michael Tushman

January, 1981





An Investigation Into the Managerial Roles and Career
Paths of Gatekeepers and Project Supervisors in a Major

R5D Facility*

Ralph Katz January, 1981

Michael Tushman WPl 187-81

Accepted for publication in Industrial R^D Strategy and Management ,

edited by Alan Pearson, Basil Blackwell Press, 1981.

An Investigation Into the Managerial Roles and
Career Paths of Gatekeepers and Project
Supervisors in a Major R5D Facility

R§D project teams must process information from outside sources in
order to keep informed about relevant external developments and new
technological innovations (Myers and Marquis, 1969). Furthermore, empirical
studies over the past 15 years have demonstrated that oral communications,
rather than written technical reports or publications, are the primary means
by which engineering professionals collect and disseminate important new
ideas and information into their project groups (yXllen, 1977). While such
personal contacts may be essential, there are alternative communication
structures by which RSD groups can effectively draw upon information outside
their organizations (Katz and Tushman, 1979) . In particular, the research
reported here focuses explicitly on the role played by gatekeepers in the
effective transfer and utilization of external technology and information.
Since most gatekeepers are also project supervisors, this study also contrasts
the managerial roles and subsequent career paths of gatekeeping
supervisors against project supervisors not functioning as gatekeepers.

Communication and Performance

Generally speaking, previous research has shown that project performance
is strongly associated with high levels of technical communication by all
project members to information sources within the organization (i.e., high
levels of internal communication). The positive findings of Allen (1977),
Pelz and Andrews (1966), and Farris (1972) strongly argue that direct contacts
between project members and other internal colleagues can enhance project


While direct conununication by all project members may be effective for
internal communications, the particular method for effectively keeping
up-to-date with technical advances outside the organization are
probably very different. Numerous studies, for example, have shown that
project performance is not positively associated with direct project member
communication to external information areas. In fact, most studies have
found them to be inversely related (e.g., Allen, 1977; Katz and Tushman, 1979;
Baker, Siegmann and Rubenstein, 1967). It seems that most engineers are
simply unable to communicate effectively with extraorganizational information


One explanation for these significant differences stems from the idea
that technological activities are strongly local in nature in that their
problems, strategies, and solutions are defined and operationalized in terms
of particular strengths and interests of the organizational subculture in
which they are being addressed (Katz and Kahn, 1978; Allen, 1977). Such
localized definitions and shared language schemes gradually unfold from the
constant interactions among organizational members, the tasks' overall objec-
tives and requirements, and the common social and task related experiences of
organizational members. These idiosyncratic developments are a basic
determinant of attitudes and behaviors in that they strongly influence the
ways in which project members think about and define their various problems
and solution strategies.

Such localized perspectives eventually become a double-edged sword.
As long as individuals share the same common language and awareness, communication
is rather easy and efficient. Conversely, when individuals do not share a
common coding scheme and technical language, their work-related communications
are less efficient, often resulting in severe misperceptions and misinterpre-
tations (Dearborn and Simon, 1958). Thus, the evolution of more localized


languages and technological approaches enables project members to deal
effectively with their more local information processing activities within
the organization; yet at the same time, it hinders the acquisition and
interpretation of information from areas outside the organization. This lack
of commonality across organizational boundaries serves as a strong communi-
cation impedance causing considerable difficulty in the communications of most
engineers with external consultants and professionals (Allen, 1977; Price,

Given this burden in communicating across differentiated organizational
boundaries, how can project groups be effectively linked to external infor-
mation areas? One way is through the role of project gatekeeper; that is,
certain project members who are strongly connected to outside information
domains but who are also capable of translating technical developments and
ideas across contrasting coding schemes (Allen and Cohen, 1969) . Through
these key members, external information can be channelled into project groups
by means of a two-step communication process (Coleman, Katz, and Menzel, 1966).
First, gatekeepers gather and understand outside information, and subsequently
they translate it into terms that are more meaningful to their locally con-
strained colleagues. Gatekeepers, as a result, perform an extremely valuable
function, for they may be the principal means by which external ideas and
information can be effectively transferred into R&D project groups.

While substantial literature applauds this gatekeeper concept, there is
virtually no direct evidence that gatekeepers enhance project performance.
Support has to be inferred indirectly either from the empirical findings of
Katz and Tushman (1979) and Allen, Tushman, and Lee (1979) or from the
case studies in project SAPPHO (Achilladeles, Jervis, and Robertson, 1971).


Our initial research question, then, concerns the association between project
gatekeepers and technical performance. Is this relationship positive across
all forms of R§D activity or are some project areas more effectively linked
to external technology through direct contact by all project members rather
than through a gatekeeper? Moreover, if gatekeepers are necessary for effec-
tive technology transfer, then must they be the primary source for collecting
outside information, or can they also serve to facilitate the external
communication of their more locally constrained colleagues?

Gateke epers, Perform ance, and the Nature of the Task

The need for a two-step process of information flow depends on a strong
communication impedance between the project group and its external information
areas. To the extent that different technical languages and coding schemes
exist between project members and their external technical environments,
communication across organizational boundaries will be difficult and
inefficient. In particular, most technological activities (unlike the
sciences) are strongly local in nature. The coupling of bureaucratic inter-
ests and demands with localized technical tasks and coding schemes produces
a communication boundary that differentiates these project groups from their
outside areas. Product development groups in different organizations, for
example, may face similar problems yet may define their solution approaches
and parameters very differently (Katz and Tushman, 1979; Allen, 1977). As
a result, it becomes increasingly difficult for most technologists to
integrate external ideas, suggestions and solutions with internal technology
that has become locally defined and constrained. It is hypothesized, there-
fore, that locally oriented projects (i.e., development and technical service
projects) will require gatekeepers to provide the necessary linkages to


external information areas - without j^atekeepers, direct external contacts

by members of local projects will be ineffective.

Contrastingly, if external information sources do not have
different language and coding schemes from members of the project group,
then a significant communication impedance will not exist. Work that is
more universally defined (scientific or research work, for example) is
probably less influenced and constrained by local organizational factors,
resulting in less difficulty vis-a-vis external communications. Under these
conditions, project members are more likely to share similar norms, values,
and language schemes with outside professional colleagues, thereby, permitting
effective communication across organizational and even national boundaries.
They are simply more capable of understanding the nature of the problems and
corresponding solution approaches employed by their relevant external
colleagues. Hagstrom (1965), for instance, found a strong positive corre-
lation between the productivity of scientists and their levels of contact
with colleagues from other universities. For universally defined tasks,
therefore, it is hypothesized that gatekeepers are not required to link projects
with their relevant external information areas; instead, direct outside inter-
action by all project members is more advantageous. The nature of a project's
work, therefore, should be a critical factor affecting the development of localized
languages and orientations and consequently will moderate significantly the
relationship between project performance and the usefulness of gatekeepers.

Role of Gatekeepers

If gatekeepers enhance the performance of project groups working on
locally defined tasks, then what specific information processing activi-
ties of gatekeepers contribute to higher project performance? There

are at least two alternatives. The more traditional explanation

is that gatekeepers function as the primary link to external

sources of information and technology - information flows through these

key individuals to the more local members of the project team (Allen and

Cohen, 1969J . Relevant external information is transferred effectively

into a project group because of the capable boundary spanning activities

of the project's gatekeeper.

Another possibility is that gatekeepers also assume an active training,
development, and socialization role within their work groups. From this
perspective, gatekeepers not only gather, translate, and encode external
information, but they also facilitate the external contacts of their
project colleagues. By helping to direct, coach, and interpret the external
communications of their fellow project members, gatekeepers act to reduce
the communication boundary separating their projects from outside informa-
tion areas.

If gatekeeping permits other project members to communicate effectively
with external areas, then for localized projects with gatekeepers, there
should be a positive association between a project's external communication
and its performance. On the other hand, if gatekeepers do not play this
more active role, then an inverse relation is more likely to exist between
the external communications of locally oriented group members and project
performance. Because gatekeepers work and interact so closely with other
project members about technically related problems, it is hypothesized that
gatekeepers fulfill this larger role of both gathering outside information
and facilitating the external communications of their project colleagues.

Gatekeepers and Project Supervisors

If most gatekeepers are also fiist-level project supervisors (Allen,
1977), then to what extent can any project supervisor substitute as a
gatekeeper and play this linking role to external areas? Supervisors
of locally oriented projects who are not gatekeepers face the same
communication impedance as their project subordinates when communicating
externally. As a result, without the benefit of a gatekeeper, the communi-
cations of non- gatekeeping supervisors outside the organization will
be inversely related to project performance. In contrast, supervisory
gatekeepers are capable of communicating effectively across organization-
al boundaries and consequently will show a positive association between
external communication and project performance.

Finally, if there is a significant distinction between the informa-
tion processing activities and capabilities of project supervisors who
are gatekeepers and those who are not, then to what extent will they
also have different career paths within the technical organization. Are
gatekeeping supervisors, for example, more likely to be promoted to
particular laboratory positions than non-gatekeeping supervisors? From
an exploratory point of view, this research describes the career paths
of these different kinds of project supervisors over a 5-year period.
The key issues are whether gatekeeping and non-gatekeeping supervisors
were promoted and utilized differently within the organization over
this time period, and whether they are currently effective in their
respective career positions.

This study was conducted at the R§D facility of a large American
Corporation. Employing a total of 345 professionals, the laboratory was

organized into 7 departments, each containing its ov^-n set of projects. At
the time of our study, 61 separate project groups existed across the 7
departments. These groups remained stable over the data collection
period, and each professional was a member of only one project group.


To measure actual communications, each professional kept
track of all other professionals with whom he or she had work-related oral
communication on a randomly chosen day each week for 15 weeks. The
sampling of days was arranged to allow for equal numbers of weekdays.
Respondents reported all contacts both within and outside the laboratory's
facility, including whom they talked to and how many times they talked with
that person. Social and written communications were not reported. An
overall response rate of 93% was achieved over the 15 weeks. In addition,
68% of all communication episodes reported within the laboratory were
reciprocally mentioned by both parties. Given these high rates of response
and mutual agreement, these sociometric methods provide a rather accurate
picture of the verbal interactions for all laboratory professionals.

For each project member, internal comir.uni cat ions was measured by
summing the number of work-related contacts reported over the 15 weeks between
that member and all other professionals within the organization. External
or outside communication was measured by suroning the member's reported
communications to other professional individuals outside the organization,
including R5D consultants, professors, vendors, customers, and the like.


As discussed by Katz and Tushman (1979), these individual scores
were also aggregated to obtain project measures of internal and
external communication.

Conceptually, project gatekeepers are defined as those members who
are high internal communicators and who also maintain a high degree of
outside communication. In line with previous studies (see Allen, 1977).
this study operationalized gatekeepers as those project members who were
in the top fifth of both the internal and external communication distri-
butions. Gatekeepers were identified in 20 project groups while 40
projects had no gatekeepers within their memberships.

Project Type

R§D tasks differ along several dimensions, including time span of
feedback, specific vs. general problem-solving orientation, and the
.generation of new knowledge vs. utilization of existing knowledge and
experience (Rosenbloom and Wolek, 1970). Based on these dimensions,
distinct project categories were defined ranging from research to development
to technical service. Such a categorization also forms a universal (research)
to local (technical service) project continuum. As discussed by Katz and
Tushman (1979), respondents were asked to use these specific project
definitions and indicate how well each category represented the objectives
of their task activities. A second question asked respondents to indicate
what percentage of their project work fell into each of the project categories.
A weighted average of these two answers was calculated for each respondent
(Spearman- Brown reliability = .91)

To categorize projects, however, the homogeneity of members' perceptions
of their task characteristics had to be examined to check for the appropriateness


of pooling across individual project members (see Tushman, 1977 for details).
As pooling was appropriate, individual responses were averaged to get final
project scores, yielding 14 Research, 23 Development, and 23 Technical
Service projects. Research projects carried out more universally oriented
scientific work (discovering new knowledge in glass physics, for instance)
while development and technical service projects were more locally oriented in
that they worked on organizationally defined problems and products.

Project Performance

Since comparable measures of project performance have yet to be developed
across different technologies, a subjective measure was employed. Each
Department and Laboratory Manager (N = 9) was separately interviewed and asked
to evaluate the overall technical performance of all projects with which he
was technically familiar. Whenever an informed judgement could not be made,
they were asked not to rate the project. From these interviews, each project
was independently rated by an average of about 5 managers using a seven-point
scale ranging from very low to very high. These individual ratings were
averaged to yield overall project performance scores (Spearman-Brown
reliability = .81).

Follow-up Study

Approximately 5 years after these previously described data were
collected, we returned to this R§D facility to locate the current laboratory
positions of the original set of project supervisors. During this time
interval, a dual ladder promotional system had been installed.
According to the company, the technical ladder was introduced to reward
individual professionals whose "technical competency and contributions are


well-recognized." All technical ladder positions were above the original
project supervisory level. As a result, we were able to determine from
our follow-up analysis whether a project supervisor had either (1) been
promoted up the managerial ladder, (2) been promoted up the technical
ladder, (3) had not been promoted above the project level, or [4) had left
the R5D facility.

Finally, a very high-level manager currently investigating problems asso-
ciated with the dual ladder system was asked to evaluate the particular
project supervisors who had been promoted up the technical ladder (N = 12).
Based on his knowledge of the current technical contributions of these
individuals, each was rated on a 4-point scale ranging from low to high. Unfort-
unately, similar performance ratings for project supervisors promoted up the
managerial ladder could not be obtained.


Gatekeeper Presence and Proj ect Performance

The performance means reported in the first row of Table 1 clearly
indicate that, in general, the performances of projects with gatekeepers
were not significantly different from the performances of projects without
gatekeepers. As previously discussed, however, locally oriented projects
(i.e., development and technical service] should display a positive asso-
ciation between gatekeeper presence and project performance. Universal-
type or research projects, on the other hand, should show an inverse
relation between gatekeeper presence and project performance.

The breakdown of performance means by project type strongly supports
these differences in the appropriateness of the gatekeeping function. As
shown by Table 1, research projects without gatekeepers were significantly
higher performing than research projects with gatekeepers. It may be that
research projects are more effectively linked to external information areas


through direct member contacts

Insert Table 1 About Here

In sharp contrast, development projects with gatekeepers were signi-
ficantly more effective than development projects without gatekeepers. Unlike
research groups, then, development projects are linked to outside information
areas more effectively through the use of gatekeepers. No significant differ-
ences Ln project performance, however, were discovered between technical
service groups with and without gatekeepers. As a result, the mechanisms
used by technical service projects to import external information effectively
remain unclear.

Role o f Gatekeepers

It was suggested that on locally oriented tasks, gatekeepers may do
much more that simply channel outside information into their project groups.
They may also act to reduce communication impedance, facilitating the
external communications of their fellow project colleagues. In contrast,
locally oriented projects without gatekeepers will have no clearly effective
link to external areas.

Results reported in Table 2 support these ideas. For local projects
without gatekeepers, there was a consistent inverse association between
members' external communication and project performance. For projects with
gatekeepers, however, a significantly different pattern emerged - external
communication was positively associated with project performance. Further-
more, these correlational differences were strong even after the direct
communication effects of gatekeepers were removed! For both development


and technical service groups, gatekeepers and their project colleagues
were able to communicate effectively Kith outside professionals.

Insert Table 2 About Here

The significant correlational differences between projects with and
without gatekeepers strongly support the argument that gatekeepers influence
the ability of local project members to communicate effectively with external
sources of technical information. Members of research projects, on the
other hand, do not seem to face a communication impedance when communicating
externally, for Table 2 shows that the level of outside interaction by all
research project members was positively associated with performance inde-
pendent of a gatekeeper's presence within the group. Gatekeepers as a
result, may not play an important information processing role in the more
universally oriented research projects, but they appear to play a vital
role in the more locally defined development and technical service projects.

Gatekeepers and Project Supervisors

Can project supervisors substitute for gatekeepers in linking their
projects to external information areas? The correlations reported in Table
2 do not support this position. For development and technical service
projects, the greater the external communication of project supervisors
who were not gatekeepers, the lower their project's performance. Generally
speaking, therefore, supervisors are not necessarily an effective link
to external domains. Contrastingly, the association between outside contact
and project performance was very positive for supervisors who were also
gatekeepers. Such significant correlational differences strongly imply
that ;upervisory status alone cannot effectively deal with the demands for


keeping in touch with outside information sources.

In light of these significant role differences, were gatekeeping and
non-gatekeeping supervisors likely to receive» the same kinds of
promotions? The results of Table 3 suggest they did not. The follow-
upstudy of the facility some 5 years later reveals that almost all of the gate-
keeping supervisors had been promoted up the managerial ladder. Of the 12
gatekeeping supervisors remaining with the company, 11 are in higher-level
managerial positions. Although non-gatekeeping supervisors were almost as
likely to be promoted, they were not as likely to receive managerial
promotions. Almost as many non-gatekeeping supervisors were promoted up the
technical ladder as were promoted up the managerial ladder. In fact, of the

13 project supervisors who had made it up the technical ladder, only one


had been a technical gatekeeper!

Insert Table 3 About Here

While gatekeeping supervisors were essentially promoted up the managerial


Online LibraryRalph KatzAn investigation into the managerial roles and career paths of gatekeepers and project supervisors in a major R&D facility → online text (page 1 of 2)