development, implementation, and diffusion of new or improved solutions and
work methods. Furthermore, the project
outcomes were briefly discussed in retrospect. All three rounds of interviews
were recorded digitally.
Data Analysis
The empirical data were analyzed using
an exploratory thematic analysis procedure inspired by Braun and Clarke
(2006) and Spiggle (1994). Data analysis
followed a four-stage iterative approach
involving repeated rounds of reading
and categorizing of the data. Stage 1
involved line-by-line open coding while
listening to the recordings and reading
the transcriptions and field notes from
each interview. We, thereby, identified
a long list of initial codes related to the
three aggregate dimensions of exploration, exploitation, and co-creation from
the data within each case. In Stage 2, we
searched for links among the first-order
codes within each case, which facilitated
grouping them together into second-level categories and third-level themes
(Braun & Clarke, 2006; Spiggle, 1994).
A core aspect of this stage was that
we allowed concepts and relationships
to emerge from the data, rather than
being guided by an explicit and detailed
theoretical framework. In Stage 3, we
conducted cross-case analysis, looking
for similar concepts and relationships
across cases, comparing the categories
and themes produced in the second
stage. We conceptualized and labeled
these themes by capturing the content
on a higher level of abstraction and
by referring to existing literature that
described similar concepts (Spiggle,
1994). In the final stage, we drew on existing studies on organizational learning
and co-creation to refine our labels. The
themes were refined through repeated
investigations both of patterns of commonality and of atypical examples, and
were repeatedly compared with the literature in order to achieve congruence
in terminology.
The thematic analysis identified five
key learning themes, each describing a
different type of explorative or exploit-
ative learning process that the project
actors engaged in, either individually or
in co-creation. Exploration occurred in
terms of: ( 1) adaptation, which involved
dealing with changes derived from dif-
ferent sources of uncertainty related to
ground conditions, poor tendering doc-
uments, client requirements in the early
stages, and late end-user involvement;
and ( 2) radical development, which
involved either process or product
development. Exploitation occurred in
terms of: ( 3) incremental development,
which involved continual improvement
of existing knowledge and technologies;
( 4) knowledge sharing across projects
through post-project review meetings
and lessons-learned sessions; and
( 5) innovation diffusion across projects
and organizations. These five learning
themes form the overarching frame of
the empirical findings presented in the
next section. In addition, the thematic
analysis identified four themes related
to the co-creation dimension: ( 1) co-
creation during the design stage, ( 2) co-
creation during the production stage,
( 3) barriers to co-creation, and ( 4) driv-
ers for co-creation. See the Appendix
at the end of the article for additional
details on codes and themes.
Findings and Analysis
Adaptation: Dealing With Uncertainty
In all five projects, substantial changes
of existing plans and routines in terms
of adaptations to dealing with unpre-
dictable or changing circumstances
were major parts of daily work. Adapta-
tion is related to explorative learning in
the sense that it involves challenging
change efforts that require new solu-
tions to either production processes or
the end product. Project participants
were proud of being able to resolve any
problems that occurred, and adapta-
tion processes were deemed to have
saved some of the projects from fail-
ure. The empirical findings identify
four main categories of adaptation
processes. These categories are related
to dealing with challenges stemming
from uncertain ground conditions, poor
tendering documents, uncertain client
requirements in the early stages, and
late end-user involvement.
Dealing With Uncertain Ground
Conditions
All projects that included extensive civil
engineering work (i.e., Projects 1, 2, 3,
and 5) faced challenges stemming from
uncertain ground conditions, in terms
of issues with the ground water and geo-technical challenges of working in rock
and clay. These difficult ground conditions, which were especially apparent
in work involving tunneling, piling, and
excavation, led to significant adaptation
during production, because work plans
and routines had to be changed. For
example, Project 5 experienced major
problems caused by groundwater when
excavating the bottom of the shaft for a
fuel bunker 15 meters (approximately
50 feet) below ground level. The original
design using a ‘cut and cover’ procedure
caused a severe inflow of contaminated
groundwater, which exceeded the capacity of the pumping equipment. The escalating and chaotic situation was managed
by the client and the contractor who
combined their competences to jointly
devise a new solution: “During the excavation, we encountered an unforeseen
problem with the groundwater flow. We
were afraid that we would drain the whole
city. In true partnering spirit the contractor suggested that we would investigate
whether a diaphragm wall could be an
option, which is relatively new as a permanent structure in Sweden. At that point
we had great collaboration” (Client 5B).
Dealing With Poor Tendering
Documents
Another critical antecedent of adaptation processes in the projects was the
poor quality of the clients’ original
designs, plans, and tendering documents. The low-level quality of these initial documents and plans meant that the
parties routinely needed to develop and
re-negotiate new plans, solutions, and
drawings because errors were revealed
