Miner WP 2005 - The Dynamics of Routines and Capabilities in New Firms

8/3/2019 Miner WP 2005 - The Dynamics of Routines and Capabilities in New Firms

http://slidepdf.com/reader/full/miner-wp-2005-the-dynamics-of-routines-and-capabilities-in-new-firms 1/39

THE DYNAMICS OF ROUTINES AND CAPABILITIES

IN NEW FIRMSYan Gong

Department of Management and Human ResourcesSchool of Business

University of Wisconsin – Madison975 University Avenue

Madison, Wisconsin 53706(608) 265-8695

[email protected]

Ted Baker

Department of ManagementSchool of Business, U-1041 MG

University of ConnecticutStorrs, CT 06269(860) 408-1567

[email protected]

Anne S. MinerFord Motor Company Distinguished Professor

Of Management and Human ResourcesSchool of Business

University of Wisconsin – Madison975 University Avenue

Madison, Wisconsin 53706(608) 263-4143

[email protected]

Draft Version

January 10, 2005



1

ABSTRACT

Organizational routines and capabilities play a key role in organizational survival and

prosperity. This inductive paper explores the interplay between the development of routines and

capabilities in new firms. We draw on data from transcripts of in-depth interviews in sixty

young knowledge based firms; founding teams included scientists, business people and mixtures

of both.

The data revealed a surprising variety of processes through which new ventures develop

routines and these routines intertwine with capabilities. Our findings show several distinct

trajectories between routines and capabilities in new firms. For example, in some cases

capabilities preceded supporting routines, rather than the other way around. Further, non-routine

behavior including improvisational actions could provide foundations for capabilities.

Our propositions advance theories of organizational learning and entrepreneurship. First,

we uncovered a rich set of patterns through firms succeeded or failed to learn from their own

experience in developing new routines and capabilities. Learning from failure in new ventures is

aided by declarative knowledge but inhibited by procedural knowledge within founding team.

Second, we also found a pattern of absorptive inertia – some new firms developed the capacity to

absorb knowledge from outside the firm, but at the same time developed an unwillingness to

absorb external knowledge.



2

Scholars have long used the concept of routines to describe coordinated, repetitive sets of

activities in organizations, and a major stream of work on organizational adaptation puts a

spotlight on the key role of organizational routines (Cyert and March 1963; Feldman 2000;

Feldman and Petland 2003; Miner 1991; Nelson and Winter 1982). Organizational routines

guide organizational activity, creating stability and boosting efficiency under commonplace

conditions of scarcity of attention (Cyert and March 1963; Feldman 2000; Feldman and Pentland

2003; Ocasio 1997). Research also suggests that routines are often the central components from

which firms build organizational capabilities (Nelson and Winter 1982; Teece, Pisano and Shuen

1997; Winter 2000). Most of the rich empirical literature on routines and capabilities seeks to

understand large and established firms, in which it is appropriate to take the existence of routines

as and empirical and theoretical premise. But such assumptions are misplaced in studying new

firms, which have had little opportunity to develop routines or build them into capabilities. If we

are to understand the development of capabilities in new firms, it is essential, therefore, that we

understand where routines come from in new ventures, and the dynamics of the relationship

between capabilities, routines and other organizational behaviors.

We therefore began our inductive study with two simple questions. First, we wondered,

where do routines and capabilities come from in new venture? As one respondent reported in

our interview when asked about advertising procedures, “Well, I knew that we don’t have a,

we’re new. How do we have a routine in anything? (BraChem 11:29)”

Second, we asked, what is the relationship between capabilities, routines and other

patterns of behavior in new ventures? We noted that although the literature tends to make

assumptions about links between routines and capabilities, their core definitions do not have to

predefine such links. For example, a soccer player may have a routine of putting on her right



3

shoe first. No capability would be lost were this routine forgotten. She may combine routines

and capabilities in a set of distinctive sideline moves that provide her with repeated goal-scoring

opportunities. Finally, she may occasionally improvise new moves that provide new capabilities

and make her a better player.

Although it is not hard to see these distinctions when pondering individual level routines

and capabilities, we wanted to find out how they are linked in new firms and also how and if

these links change over time. We therefore left open the notion that a specific bundle of

organizational activities could be a routine, a capability, neither or both. We began by

examining prior definitions of routines and capabilities in order to adopt working definitions that

specified the necessary conditions for something to be a routine and for something to be a

capability. Table 1 illustrates organization scholars’ prior definitions of routines and capabilities.

--- Insert Table 1 About Here ---

Organization scholars have built on the intuition behind standard operating procedures

(Cyert and March, 1963) to develop multiple definitions of organizational routines. Upon close

examination, coordination and repetition of behavior have been at the core of most definitions

used by evolutionary and learning theorists (e.g., Nelson and Winter 1982; Miner, 1991;

Feldman 2000). In our study, we therefore defined organizational routines as a “coordinated ,

repetitive set of organizational activities (Miner, 1991).”

Prior models of capability development have often suggested that organizations build

capabilities from existing routines or enhance capabilities through replacement of obsolete

routines (e.g. Nelson and Winter, 1982; Teece, Pisano and Shuen, 1997). That is, routines have

been seen as the primary building blocks of capabilities. Some prior definitions even include

routines as defining elements of capabilities. Based on our broad review of the literature (see



4

Table 1 for representative examples of prior use), and seeking to leave the relationship between

routines and capabilities an empirical rather than a definitional question we adopt the following

definition of organizational capability for this study: “An organization has a capability when it

can reliably perform particular activities or reliably accomplish particular classes of intended

outcomes at or above a given performance level.”

We emphasize that such a capability may or may not be deployed at any given time.

Consistent with both dictionary definitions of capability (e.g. The American Heritage Dictionary

of The English Language (2000)) and prior use in psychology, capabilities may exist in ‘latent’

form or as ‘potential.’ For example, Paulhus and Martin (1997) operationalized ‘interpersonal

capabilities’ in terms of ‘how likely is it that you could be dominant if the situation requires it,’

focusing on the likelihood of a specific capability being deployed in a specific context. We also

emphasize that organizational capabilities underlie the likelihood, but not the guarantee of

positive organization performance in any given situation.

METHODS

Sample. We drew our sample of 60 young knowledge-based firms from three archival

sources: the Dun & Bradstreet database (D&B), Creating High-Tech Business Growth, a list of

firms published by a university group , and the 2001 Directory of High Technology Companies

published by the local utility company. We also reviewed our list with local experts who had

information about new, knowledge-based firms at their earliest stages.

Starting with a list of firms aggregated from the four sources, we used a stratified random

sampling approach with several filters. The goal was to generate a sample of knowledge-based

young firms operating in the same geographic area, with a mixture of firms with and without

direct roots to the nearby research university. Our sample contains firms that were in the focal



5

County, began operating no earlier than January 1, 1995, and employed at least three people as

of November, 2001, and operating within the drug, biotech or IT industries (SICs 283, 737, 873).

These filters generated a sample that included 147 firms.

Among the 147 firms, we have interviewed founders of 60 companies. No firm has

refused our request that they participate in the study, and every firm has allowed us to tape our

interviews. During the interview process, we found that two firms actually started before

January 1, 1995, and two other firms operated in line of business different from their published

SIC. Table 2 provides a descriptive summary of our sample.

--- Insert Table 2 About Here ---

Data and Analysis. Our data collection and analysis followed standard grounded theory

building techniques (Denzin and Lincoln, 1998; Ragin, 1987; Strauss and Corbin, 1998). A

project team member contacted each founder by phone and introduced us as university

researchers investigating management practices in knowledge-based firms, and scheduled a time

to visit the firm and conduct interviews.

We conducted pilot interviews, which permitted us to improve our protocol and study

how the interview materials and interview style affected respondent’s reporting behavior. We

developed refined protocols for formal semi-structured interviews that began with open-ended

questions but then moved toward standardized probes. These included written instruments

completed during and after the interviews.

At least two and sometimes more members of the project team conducted each interview.

The typical interview lasted 2.5-3 hours, with some lasting much longer. All interviews were

taped and transcribed by a professional transcription service with extensive experience with

similar projects. An advantage of studying firms of this type is that the founders are typically



6

involved in all key aspects of the business and consequently have firsthand knowledge of the

firm’s day-to-day activities. Our respondents were generally able to offer very detailed

responses to our questions, and to provide detailed timelines and histories for their firms.

We collected detailed information on employee hiring, and on specific non-employees

that the founders reported as having substantially influenced the firm. The interview and

documentation process generated over 1,725 pages of transcripts, plus detailed field notes.

The possibility of retrospective bias by informants is a potential threat to the quality of

our data. However, because we studied young firms and because we asked founders to describe

specific events without providing them a framework with which to evaluate and interpret their

answers, we believe that this threat is minimized.

Although none of our questions asked directly about the processes through which the

firms acquired and altered routines, our initial reviews of transcripts revealed a wealth of data on

a variety of processes and patterns that appeared to strongly influence the presence of routines

and capabilities. These data were ancillary to the primary research questions of the original

study but are central to our current focus on patterns and processes of routine and capability

development in young firms.

We studied the transcripts, using our two basic questions about the dynamics of routines

and capabilities in young firms. Through discussions among members of the project team and

with colleagues outside the team, we developed an initial framework for organizing the relevant

data and characterizing the events and processes. Through multiple iterations between our

developing theoretical framework and the data, we first generated a large number of themes and

apparent patterns. We then subjected these apparent themes and patterns to stringent scrutiny,

attempting to challenge the level of support to be found in the data. Eventually, a limited



7

number of themes and patterns withstood this scrutiny and were demonstrated to have support in

our data. We describe these our findings in the section below, drawing on transcript materials as

illustrations.

SOURCES OF ROUTINES IN NEW VENTURES

Because the prior literature strongly suggests that capabilities are composed from

routines, we first examined sources of routines in the new ventures we studied. In later sections,

we describe the dynamic relationships we observed between routines and capabilities, and

develop propositions regarding some of the processes that affect these dynamics.

We observed three primary sources of routines in our sample: importation from prior

employers, integration of routines from firm networks, and creating routines from real time

experience.

Importing Routines from Prior Organizations

Founders differed in the sorts of knowledge of prior routines they brought to their new

ventures. In some cases, founders had actually been the people responsible for performing

routines at prior workplaces, rather than the people responsible for developing routines or

deciding how or when to implement them. In other cases, founders had participated primarily in

decision-making or oversight of specific routines at prior firms, but had little actual “hands-on”

knowledge of the performance of the routines. This pattern led us to distinguish procedural

knowledge, defined as “the ability to execute sequences to solve problems” from declarative

knowledge, defined as “implicit or explicit understanding of the principles that govern a domain

and of the interrelations between units of knowledge in a domain (Rittle-Johnson, Siegler and

Alibali, 2001).”



8

The patterns of founder knowledge regarding specific prior routines appeared to be

associated with the patterns of importing routines from where founders used to work (see Figure

1). Procedural knowledge is closely associated with skills and habits, and it may become

automatic or accessible unconsciously over time (Cohen and Bacdayan 1994; Moorman and

Miner 1998). Such knowledge was frequently associated with what we labeled “replication,” or

the attempt to directly copy a prior routine into the new organization. In contrast, when founders

had declarative knowledge (sometimes combined with procedural knowledge), we often

observed what we labeled “transplantation,” a more deliberate, selective and perhaps slower

process of importing routines from an old to a new firm.

--- Insert Figure 1 About Here ---

In many cases, replication was a quick, simple and apparently effective solution to the

desire to develop routines in the new ventures. However, replication of routines between very

different source and target operating environments sometimes triggered interesting and

unexpected organizational behavior. For example, members of one founding team had worked

as employees of a music business entrepreneur and later founded a software consulting firm.

Following a routine that they had learned and taken for granted in the music business, they used

a potential partner’s intellectual property on their website. This resulted, to their great surprise, in

a devastating lawsuit.

In contrast, some founding teams made use of declarative knowledge about prior routines

to engage in a very deliberate process of selection during which some routines and bundles of

practices were imported while others were explicitly rejected.

In one case, two of the three founders of an IT firm became unhappy with the evolution

of human resource routines – the original vision had been “we're going to have fantastic



9

engineers and we're going to make sure that they are growing and have a great place to thrive.

(IstNet 3:25)” – and they left with three employees to found another firm offering competing

services. They brought with them deep understanding of the history of the prior firm’s routines,

and a desire to avoid replicating the prior firm’s failures. Before embarking on the new venture,

the five future founders met on a weekly basis and carefully planned which routines they would

and would not transplant. As one founder put it, “And then (we) got together and talked about,

you know, what are we going to do with our company? And we looked at a lot of, at what G (the

previous company) had done right in the beginning and then we used a lot of examples of what G

had started doing wrong. And so we had a great, it’s like we’d all been involved since almost the

beginning at G and seeing them go from, you know, a few people to 120 people. And so we got

to ride in that cycle and be part of it. And we got to see it then go wrong. So it was a great lab

for us. (IstNet 2:35)”

The founder believed that they had been able to learn from the failures of their prior firm

to build a business that not only fit their values better, but also was also able to outcompete the

old firm. The old firm eventually entered bankruptcy, and the new firm was able to take over

and retain most of its accounts.

Combining Routines from Founding Teams and Networks

We also observed cases in which new firms generated routines by combining elements of

routines brought to the firm by different founders or by members of its network, including for

example, advisors, bankers, suppliers and customers. We label this process “combination.”

Figure 2 illustrates the combination of elements of routines from multiple sources. This process

often seemed to work quite smoothly, but in some cases – especially those involving scientist

and non-scientist founders – the combination process generated confusion and even struggle.



10


For example, one scientist had a single patent for which he perceived multiple potential

product applications in the pharmaceutical industry, including a promising cancer drug. His

three business partners proposed a routine they had followed before, which involved focusing

their resources on exploiting one relatively easy application and using this to generate the equity

financing the firm would require if it was to develop even one of the applications to the point of

commercial viability.

The scientist continued to insist on following his previously successful routine of

pursuing multiple simultaneous research paths, though learning to couch his preferences in

business terms. In the scientist’s words, “So I thought well, okay, so maybe, I’m not a table

slammer or anything like that, but I continued to be fairly strongly, we need to develop these.

This product right here, you’re telling me this is a billion dollar market. We can’t just let that sit

(ChemPro 7:23).” For four years, the scientist continued expending resources on multiple

projects. None of the projects reached the stage where it might be commercialized, but the

product that showed most promise was not one that any of the founders would have chosen to

focus upon initially.

Around this time, the scientist began to appreciate the potential value of focused business

routines while the business partners started to see some value in keeping multiple scientific

options open. The firm then decided upon and developed a routine in which the most promising

short-term projects would become the primary scientific focus, while other applications would be

given less immediate attention. Thus, instead of developing multiple potential drug product lines,

the firm began to concentrate on applying their patent to treating a specific medical condition

resulting from chemotherapy. Overall, the difficult and lengthy process of evaluating and



11

integrating founders’ knowledge of very different routines resulted in a new “combination”

routine that the founders viewed as contributing to their firm’s success.

Creating Routines from Real-Time Experiences

New firms in our sample created new routines based on real-time experience. We focus

here on one particularly intriguing pattern – creating routines by extending existing routines to

new domains, which we labeled “extension” (Figure 3 illustrates routine extension).

In some cases, founders started their businesses with technical or scientific expertise but

little or no business experience or and few useful business contacts. In such cases, the founders

sometimes relied on the crude extension of existing routines with which they were familiar as a

way to deal with business challenges. While this often seemed to fail, in other cases such

extension seemed to create idiosyncratic and useful routines.


For example, one firm, founded by software engineers with little business experience,

faced the new challenge of how to bill customers efficiently in cases where they had no

substantial relationship with a customer, but the customer called up and requested small or

simple services. The founders had no experience with billing systems or common routines for

billing. Instead of asking someone to help them develop a billing routine, and instead of buying

a packaged billing and accounts receivable package, they used their experience in web design to

develop – from scratch – a web-based system for billing minor customers for minor requests.

The founders believe that their approach allowed them to optimize customer management

during a time when they really couldn’t afford to say “no” to any customer. One founder

explained it this way, “But on the profit side of things, we would never have said that if a client’s

going to call you once a month for $200, that’s not worth your time. Instead we developed a



12

system to make it very efficient to capture that $200 . . . right, they call you . . . on their web

page. You know, you talk to them. You do what they need, hit a button on their web page.

They get billed, period, end of story (CloNet 7:23).” They also believe that the information

generated by their system has allowed them to make better strategic and market-segmentation

decisions than their local competitors.

Having closely examined the patterns of activities through which the firms we studied

were able to develop routines, we next examined the trajectories that the firms followed in their

attempts to develop capabilities.

TRAJECTORIES BETWEEN ROUTINES AND CAPABILITIES

Much of the prior literature assumes – sometimes by definition (See Table 1) – that

routines precede and provide the primary building blocks that firms combine to create

capabilities. In prior studies, important organizational capabilities – especially when studied

after they are well established – often appear to rest on a foundation of routines. We observed

this pattern as well. Our study also provided a window into the critical role that improvisation,

which is “present to the degree to which design and execution converge temporally and

substantively, with intention and an element of novelty (Miner, Bassoff and Moorman, 2001),”

sometimes played in allowing firms to build new capabilities.

Strikingly, in the firms we studied, capabilities sometimes also preceded their supporting

routines, thereby reversing the typically assumed order. More generally, the relationship

between routines, capabilities and improvisation was far more fluid than our reading of the prior

literature had led us to expect. In this section, we examine two trajectories: one in which

routines precede the capabilities they support and one in which capabilities precede their

supporting routines. We also describe the role of improvisation in these migrations.



13

Trajectory I: The Deployment of Routines into New Capabilities

The trajectory we describe here is represented by Figure 4, which shows improvisation

producing elements of routines that are then combined into new capabilities.


ChemPro, an environmentally progressive five-person chemical products company,

attempted to build the capability to do business nationally. To do so, it adopted routines that one

of its founders had used for managing the national distribution of similar products for his prior

employer. Over two years, ChemPro established more than thirty independent distributors,

selling to them at discounts of up to forty-five percent off of retail prices. Many of the

distributors were new and small, and faced frequent problems requiring ChemPro’s support.

For much of this unexpected support, ChemPro improvised solutions as specific problems

occurred. For example ChemPro’s distributors were initially required to carry their own

inventory, but many of them managed this poorly as “They started getting lazy on that.”

Distributor stockouts could completely stop production for a small end user business and lose the

customer for ChemPro. Although not set up to provide direct-to-customer shipping, ChemPro

would scramble and do whatever it could to get the product to the end customer on time, filling

in for the distributors’ failures. After a few improvisations, in which they solved a variety of end

user customers’ problems in real time, ChemPro decided to incorporate some of the improvised

solutions as routines. For example, it began providing shipping directly to end-users and thereby

eliminating the requirement that distributors to maintain substantial inventories.

After it “spent the first two years really putting together a nice distributor program,”

ChemPro began to realize that their distribution routines did not provide them with the capability

to do business profitably on a national basis. In contrast, they noticed that as they lost



14

distributors and began providing service directly to customers, their overall revenues declined,

but profits improved. The routines established for developing and working through distributors

did not provide the capabilities ChemPro required, but the solutions it had improvised to

troubleshoot problems and later supported with routines provided a more effective capability of

doing business profitably on a national basis. ChemPro rapidly dismantled its distributor

network and built what the founders describe as an increasingly profitable national business,

based on their direct-to-customer support routines. Referring to Figure 4, ChemPro improvised in

order to deal with some limitations of their distribution routines, and then developed the

improvisations into routines that supported the new national capability.

Trajectory II: Capabilities that Precede Supporting Routines

As in the trajectory described above, firms following the trajectory described in this

section end up with capabilities that are supported by organizational routines. However, the

process we describe here is exactly opposite the process of building capabilities upon routines.

Instead, we describe firms that deal with a problem or opportunity by improvising a solution,

turn the solution into a capability through sustained improvisational effort and later backfill the

capability with supporting routines that supplanted the improvisation. This process is

represented schematically in Figure 5.


Shortly after founding, FastSoft contracted to complete a software development project in

five weeks, assuming that they could apply their capabilities in using a commercial “grid

generation” package in a straightforward way. The commercial software turned out not to work

for them. One of their founders described their response as “Try not to panic I guess is the first

thing,” while a second noted, “There is a little bit of panic after we realized that the commercial



15

package wasn’t going to do it for us because we knew we only had … five weeks, and we spent

about a week trying to get that to work and when we realized it wasn’t going to work…” The

contract was one of the firm’s “first big projects” and “so we knew our reputation was kind of on

the line here.”

The firm responded to this crisis by improvising a solution completely separate from the

commercial package in a striking case of design converging with execution: “So what we did at

that point is we decided, okay, we’re going to have to write our own software here … So I spent

the next 24 hours straight writing a software package that was completely hard coded for one

specific case. I had sort of an idea how we could do something. And it was not going to solve

our big problem for us, but I take the small problem and said, okay, we need to know if this is

going to work. And coding through the whole night to get this thing, and then the next day we

tried it out and it worked.”

After the improvised solution worked on the first project, they tried it on another, and “It

worked on a very similar small-scale problem. So that’s sort of when we knew that that’s what

we needed to do to make that work.” They recognized that the software they had improvised

represented a new capability they could use to win more business, and later claimed that “a large

part of the work that we get now is because we have that software.” FastSoft continued refining

the improvised tool, and this capability is now supported by taken-for-granted routines: “We’ve

developed that over a period of about a year to be our, that is what we now use for making grids

… We use it every week, at least once a week, every week. And some projects are simply using

that code.” The new capability was built through improvisation and preceded the routines that

developed over time to support the capability.



16

Overall, the multiple trajectories through which we observed firms construct capabilities

suggested to us the importance of maintaining a clear distinction between the concepts of

routines and capabilities. Although routines appear to often serve as the building blocks

capabilities, other types of organizational behavior, such as improvisation, also sometimes

generate organizational capabilities. Moreover, capabilities may exist in the absence of

supporting routines and may serve as the precursor to the development of such routines. Figure 6

seven possible combinations of elements in the trajectories we have described. It also suggests

that behaviors other than improvisation – for example planning and the implementation of prior

plans – may, like improvisation, play a key role in new firm capability development.

ORGANIZATIONAL LEARNING

Absorptive Inertia

In previous sections we noted that the new firm’s network could be an important source

of routines. When founders had little or no prior experience related to a routine, however, as in

the case of the firm that developed the web-based billing system, they sometimes had trouble

integrating such routines. As the founding teams gradually accumulated experience they become

more appeared to become more capable of making good use of routines provided by their

networks. For example, CloneRight, founded by a group of scientists, attempted to be the first

firm to clone a particular kind of farm animal. Initially they did not know anything about

breeding, so they attempted to import breeding routines a scientist at a distant research

university. The process was plagued by errors and misunderstandings and even some public

embarrassments as the CloneRight scientists did not know enough about the topic to understand

or exercise any judgment regarding the routines they were attempting to adopt. CloneRight

eventually decided to improvise, they ‘rolled up their sleeves’ and ‘made it up as it went along’



17

and figured out the breeding routines on their own. As one founder described, “It wasn’t

working. There had been some miscommunication of facts and basically we had to become, we

knew when we entered the arrangement we were not specialists in the area of (type of animal)

reproduction, so we sought one out, and what we thought was supposed to be. Turned out we had

to become reproduction specialists. And so when we had to we rolled up our sleeves and we

became such.” (CloneRight 8:24)

Gradually CloneRight created their own routines and knowledge eventually the firm got

the point that it was able to effectively draw upon and learn from external expertise and routines.

These and similar observations echo the construct of ‘absorptive capacity’ (Cohen and Levinthal,

1990; Zahra and George, 2002), which illustrates that organizations need the right types and

levels of prior knowledge to effectively assimilate external knowledge. The literature on

knowledge transfer provides further support for the absorptive capacity argument: because the

inter-firm knowledge transfer process involves tacit and sticky information (Hansen, 1999), the

level of absorptive capacity of the focal firm is critical for effective interorganizational learning.

Based on our observations and relevant literature, we propose:

Proposition 1a: Related founding team experience eases the process of integrating

routines from network sources.

We also observed, however, that the accumulation of experience and knowledge can have

very different – and perhaps contrary effects to what research on absorptive capacity leads us to

expect. In contrast to the way that increasing topical knowledge increased firms’ ability to draw

on external routines, we observed that the same increases in knowledge and local routines

appeared to reduce firms’ desire to learn from other organizations and their willingness to import

or combine elements of external routines.



18

In the above farm animal cloning firm, for example, precisely when the firm have

developed related routines and expertise that would allow them to effectively draw upon routines

suggested by outside experts, CloneRight became unwilling to do so. As one founder described,

“Lost confidence in experts. Realized that really the only expert when you are faced with a

problem is you’ve got to just roll your sleeves up and dig in because you’re the only person

committed to finding the answer. I’ve lost my faith in consultants. Consultants are good for

gathering information from, but they’re usually noncommitted and certainly not accountable for

any results. (CloneRight 10:23)”

We define such unwillingness to draw from external expertise and routines as ‘absorptive

inertia.’ In contrast to absorptive capacity, absorptive inertia inhibits the interorganizational

learning that can be essential for the growth and survival of knowledge-based young firms. This

overall pattern we observed with the animal cloning case suggests a self-limiting dynamic in

firm’ likelihood of drawing new routines from their networks. When the cattle breeding firm had

no founder experience regarding cattle reproduction, it was unable to draw successfully on

external routines. But as it developed some experience and was able to make use of external

sources, it very rapidly became unwilling to integrate routines available through its network.

Proposition 1b: Related founding team experience may reduce willingness to draw

routines from network sources.

Learning from Failure

A recurring theme in our data and in this paper is that there is no lock-step series of

behavioral steps through which organizational capabilities are developed and changed. The

young organizations we studied attempted to accomplish their goals using whatever behavior in

their repertoires appeared most appropriate at the time.



19

Although our respondents seldom framed their behavior in terms of “experimentation,”

their catholic approach to creating and adapting capabilities resulted in dynamics of capabilities

that could often be well understood in terms of principles of experimental learning. Facing a

new challenge, the firms would try something new and observe what happened. When

something appeared to work, the organization did more of it. Thus organizations that solved a

problem through improvisation often subsequently engaged in repeated improvisation as a way

to build capabilities. Similarly, when an organization tried something and it appeared not to

work, it often learned from the failure and did less of it.

We were surprised, however, by cases in which experimental learning did not follow this

simple pattern. In particular, we observed a variety of situations in which a firm appeared to fail

to learn from the failure of their attempt to address a new challenge. In examining these

situations, two main themes emerged from our data regarding failure to learn from failure (FLF).

First, it is difficult for firms to learn from failures if they do not recognize them as failures. We

explore this pattern below in terms of what firms learn from the process of stopping an

improvisation. Second, even when a firm recognizes that the outcome of an attempt to meet a

challenge is a failure, the knowledge structure supporting the firm’s behavior appears to

constrain the likelihood of learning from failure (LF). We explore this pattern below in terms of

how declarative and procedural knowledge shape firms’ learning from the failure.

Stopping Improvisation

In this section, we examine LF and FLF when improvisation leads to failure. Prior

studies of improvisation have often assumed that it is necessary to explain why and how episodes

of improvisation start , but to our knowledge, no prior research has examined conditions under

which organizations stop improvising.



20

Improvisation is often called upon to play a role in dealing with unexpected events. In

many of the cases described to us during this study, improvisation appeared to succeed in solving

some problem or exploiting some opportunity, occasionally – as described earlier – creating a

new capability in the process. Successful improvisation often led to the development of routines

that supplanted the improvisation, and improvisation stopped primarily as a result of its own

success.

In contrast, we focus here on cases in which improvisation resulted in failure and the

improvisation was stopped as a result. We paid particular attention to evidence regarding what

firms learned from improvisational failure. We describe two scenarios of improvisational

failure, the first leading to LF and the second leading to FLF. A primary difference between

these two patterns hinges on whether the firm recognizes its own lack of capability.

Information technology consulting firm CloNet sold a project that required integrating

and synchronizing “a whole bunch” of existing audio and video files for a new client. When

they negotiated the contract, they expected – incorrectly as it turned out – that the project would

be straightforward: “And it was, you know, it was just … we looked at the job, and we thought,

okay, there’s some production work, but … not technical hurdles. It’s pretty easy. And then it

just turned out that on the target platform that happened to be the one installed in their offices, it

didn’t work right.” CloNet’s response was to improvise, immediately adding additional staff to

the project and moving in rapid sequence from one technological approach to another. After

continuing with several different approaches, including trying to get the customer to switch to a

different platform technology, the firm concluded that it simply did not have the capability to

solve the problem, as the founder described: “So what we ended up doing was saying, ‘okay,

here’s all (your) money back. If anybody else can solve this problem, we’ll cooperate with them



21

and support them, as needed, for free. Sorry we couldn’t do it….’ We punted … That was the

verb we used at the time, we’re going to have to punt.”

This is an episode of failed improvisation: the firm spent its resources on repeated

improvised experiments in support of developing a solution that did not work in the end.

However from this experience and especially the need to “punt” – which our respondents

associated with the recognition that the firm simply lacked the capability to solve the problem –

the firm was able to infer a relatively subtle strategic lesson. Based on the improvisational

failures, it began to establish routines that required project managers to create an initial

contingency framework within which, for any substantial project, improvisation would be

structured and evaluated. From the perspective of the two CloNet founders with whom we

discussed the issue, the lessons learned from recognizing the failure and stopping the

improvisation have been valuable, “We’ve only had to actually punt just, like, once. This is our

only unsolved, you know.”

A second case was quite similar in some respects, but led to very different outcomes.

Under contract with a major new customer, DeaSoft created a software system that worked on

their own test computers and environment, but when they installed it on the customer’s

computers, it simply wouldn’t work. Wanting to solve the problem and avoid failure, the firm

improvised, trying to do whatever they could to get the system to work, including flying

additional staff to the customer’s site, and even an attempt to make unplanned changes to the

customer’s computer systems.

DeaSoft did not find a solution. Much like CloNet, DeaSoft did not appear to have the

capability to improvise a solution. Unlike CloNet, however, DeaSoft continued to improvise

despite repeated improvisational experiments and failures, and was committed to continued



22

attempts to get the system to work. But, the decision to label the improvisation a failure and put

a stop to the continuing activity was made by the customer, rather than by DeaSoft itself, “The

customer very generously said, okay, we can see that it works and that your configuration of this

box is the same as ours is, as far as we know. And yours works and ours doesn’t, and our people

can’t figure it out. And they paid for it.”

While CloNet recognized that they lacked the capability to solve the problem they faced

either through continued improvisation or otherwise, DeaSoft did not appear to learn any such

lesson. Although the firm labeled the event a “big disaster,” it did not infer the need for any

change in its approach. Instead, the firm understood the event as largely the customer’s fault or

problem, emphasizing that in the end, the customer paid for the technology even though it was

useless to them. Clarifying this point, we asked, “If you faced that same event again, would you

handle it the same way?” The founder responded, “Well, I don’t think I had much choice, you

know. I pleaded with them as much as I could. They essentially said, well, yes, you could be

right, we’re not sure, but we can’t do anything about it. But they paid, so . . .”

Proposition 2a: Learning from failure is more likely when then the organization

understands its own lack of capabilities as a source of failure

The Role of Declarative and Procedural Knowledge

Earlier, we distinguished procedural from declarative knowledge of prior routines, and

elaborated their roles on importing and combining routines. In our sample, these two forms of

knowledge also appear to have very different effects on the likelihood of learning from failure.

The founder of EntCon had no practical experience making or implementing HR policies

regarding non-compete agreements. She had intensive declarative knowledge regarding why and

how they were typically used, but considered their use by her prior employers being a failure and



23

a signal of mistrust. In her new firm, she explicitly decided not to add any such legalistic

routines that might send such signals. As she noted, “I didn’t have non-competes because every

other company I’ve ever worked with had non-competes, and I didn’t understand why a

handshake wasn’t good enough.”

When an early employee subsequently left to start a competing business, she described

herself as “shocked” by the event, but decided that it was a “fluke,” rather than a failure induced

by her employment routines: “On one level I understood why he did things the way he had done

them, but I couldn’t believe he did it. I thought his value system, as I said, was so similar to

mine that I couldn’t believe he did it.” Then, within three months, it happened again, and within

three months after that, it happened a third time, and the founder decided that the ongoing trials

of her trust-based routines had to be stopped. She drew on her declarative knowledge in order to

make the change: “I hired a corporate attorney that I still work with today and had a non-

compete drawn up. So any new hires have to sign a non-compete when they join… Yes. Yes, it

was basically anyone that joins from this point on has to have a non-compete.” She asked both

existing and new employees to sign the non-compete agreement and has not had anyone leave to

compete with her since.

Despite the lack of prior procedural knowledge regarding how to implement routines

imposing non-compete agreements on employees, her intensive declarative knowledge helped

her both to see that her hiring routines were generating failures, and also to take quick steps to

address the problem.

In contrast, the founder of another information technology development and consulting

firm called CR brought a great deal of procedural knowledge of recruitment and hiring routines –

especially as they applied to salespeople – but very little declarative knowledge of principals or



24

features that might usefully guide such routines. He was convinced his procedural knowledge –

hard won during the course of hiring and managing, by his estimate, 60 salespeople –

represented a strong source of routines and an important capability for the new firm, “It just was

part of my job, to find sales reps that could go out and represent (Prior employer). So I thought I

was really good at it, you know?”

The new firm, following routines this founder helped to replicate, hired a series of seven

salespeople. Although each hire was subsequently recognized and acknowledged to be a failure,

because they brought in no new revenue, the firm did not make any substantial changes in its

routines for recruiting, selecting or managing them. At the time of our interviews, the firm had,

in the founder’s estimate, “never had a salesperson bring in any business. So that'll give you

some background. Seven people never brought in a nickel.” In the absence of declarative

knowledge that might help to diagnose and improve the firm’s routines, the founder largely

viewed the failures as out of the firm’s control, effectively throwing up his hands and explaining

that the firm tried very hard to avoid hiring any more unsuccessful salespeople, “But people

believe in themselves when they come in. And that's great. And they get you excited about what

they can do.” So the firm maintained its same failed hiring routines and was consequently

unable to develop capabilities in an area believed central to its future success.

Proposition 2b: Learning from failure is aided by declarative knowledge but

inhibited by procedural knowledge.

DISCUSSION AND CONCLUSIONS

We grounded our inductive study in prior research suggesting that routines are a principal

component of organizational capabilities and that organizations build capabilities by combining

routines. The prior literature suggested two puzzles that we investigated in this paper. First,



25

how do new firms – which by definition begin without their own routines – end up with routines

and capabilities at all? Second, what is the relationship between capabilities, routine and other

patterns of behavior in new firms? We expected that addressing these questions would shed

some light on issues of the dynamics of routines and capabilities that underlie organizational

learning and change.

Our results provide rich descriptions of the sources and processes through which new

organizations struggle to develop routines and capabilities. At the broadest level, the results

suggest that the dynamics of routines and capabilities display a much more varied and

theoretically interesting set of patterns and processes in new organizations than prior research

had led us to expect. The patterns and processes we uncovered have several implications for

theories of organizational learning, improvisation and entrepreneurship.

Organizational Learning

Young firms call upon a broad repertoire of behaviors in their attempts to meet new

challenges. Our results show that they respond by replicating, transplanting and combining

routines from founder’s prior experience, by developing and executing plans, by extending

existing routines to new purposes, by improvising and sometimes by flailing around almost

randomly. Precisely because new firms do not have routines in place for handling many of the

problems and opportunities they face, they take whatever actions seem promising, appropriate

and available at the time.

The result is a series of experiments: some planned, some unplanned, others accidental.

For one-time challenges, what matters most is whether the firm’s action deals adequately with

the challenge. What the firm learns from the experience matters less. But when the problems or

opportunities are continuing and important elements of the firm’s task environment – when they



26

represent strategic gaps in the new firm’s capabilities – questions of learning from these

experiments become paramount. An important contribution of our study is to uncover several

important dimensions of organizational learning that appear to play a key role in the dynamics of

routines and capabilities in new firms, and which contribute to theories of absorptive capacity

and learning from failure.

Self-Limiting Learning Cycles

The idea of absorptive capacity has become an important concept for understanding the

differential ability of firms to learn and adapt to changing environments (Cohen and Levinthal,

1990; Zahra and George, 2002). Consistent with, though extending prior research, our data

suggest that when members of founding teams have experiences that are related to potentially

useful routines, the process of integrating these routines into the firm is made easier.

But our results also suggest a dynamic not uncovered in prior research. We found that

exactly the same experiences that improve absorptive capacity simultaneously increase what we

labeled absorptive inertia, which is a decreased tendency and willingness to draw on external

routines. The simultaneity of increases in absorptive capacity and absorptive inertia suggests the

existence of an important temporal dimension to organizations’ likelihood of making productive

use of external knowledge and routines.

These dynamics highlight an underlying learning design issue for young firms. If during

early phases of learning the benefits of absorptive capacity dominate the limitations of absorptive

inertia, and if in later phases absorptive inertia becomes dominant, then there is likely to be a

limited window of learning during which firms can best take advantage of external knowledge

and routines. For example, what is often referred to as “not invented here syndrome,” may be

better understand as symptomatic of a learning window having closed instead of as an atemporal



27

organizational bias. More generally, the juxtaposition of absorptive capacity and absorptive

inertia raises important research questions regarding the timing of investments in knowledge

acquisition, and suggests that temporal issues may deserve more attention than they have often

received in theories of organizational learning.

Learning from Failure

Despite calls for more emphasis on organizational learning from failure (LF), and a

handful of studies that suggest rich possibilities for additional research (Haunschild & Miner,

1997; Kim & Miner, 2000; Sitkin, 1992), few studies have attempted detailed examination of

how organizations learn from their own or other organizations experience of failure.

LF represents a very valuable element of organizational learning. Failure often provides

clearer and more useful information than does success (Lounama & March, 1987; Sitkin, 1992).

Arguments from evolutionary and systems perspectives (Boulding, 1956; Campbell, 1969)

suggest that some minimum amount of failure is essential to maintain requisite variety and

adaptability in adaptive systems (Sitkin, 1992). Such arguments suggest that learning from

failure play an essential role in the dynamics of routines and capabilities.

In contrast, the failure to learn from failure (FLF) represents a serious limitation to

organizational learning. Prior research on FLF has suggested that organizations may fail to learn

from failure because they deny or reinterpret bad news (e.g. as due to randomness and luck) and

thus ignore the information value of failure (Sitkin, 1992). Sitkin argued that LF is most likely

when thoughtful pre-planning precedes a failed action and when the domain in which failure

occurs is “familiar enough to permit effective learning (Sitkin, 1992: 554).” The preplanning

and prior knowledge establish a context in which the firm can appropriately evaluate and

diagnose sources of failure.



28

Our results suggest that the type of prior knowledge a firm brings to its attempts to build

new capabilities strongly shape its ability to learn from failure. In the case of failed

improvisation, we saw that firms that knew something about the limitations of their own

capabilities appeared better able to learn from such failure in a manner that enhanced firm

capabilities. Similarly, in the case of failed routines, firms that had declarative knowledge

regarding their routines appeared better prepared to learn from the failures in ways that

contributed to capability development. In contrast, firms that had only procedural knowledge

took the basis of their capabilities for granted and appeared much less likely to learn from

failure. These results suggest that the accumulation of procedural knowledge dampens firms’

ability to learn from failure and that declarative knowledge provides some shield against this

form of learning inertia. Such dynamics invite what we consider to be fascinating research

questions regarding the role of procedural and declarative knowledge in new firms’ acquisition

of routines from external sources.

Multiple Paths to Capabilities

Finally, our results suggest that organizational routines did not play a privileged role in

the process of developing capabilities. The prior research suggests that organizations develop

and learn new capabilities by combining useful routines. We found, in contrast, that routines

were just one of several different tools that new organizations used in their experimental attempts

to meet new challenges and develop capabilities. Our discussion of “trajectories” showed, in

fact, that in our sample capabilities sometimes preceded routines.

We suspect that a primary reason our results downplay the importance of routines is our

focus on new firms. It is apparent from the strong evidence of overlap between routines and

mature capabilities in established firms that even when capabilities are initially built without the



29

benefit of routines – for example through repeated improvisation – that organizations generally

eventually undergird capabilities with supporting routines.

Improvisation, Entrepreneurship and Dynamic Capabilities

Our study makes two distinct contributions to the growing stream of research on

organizational improvisation (Miner, Bassoff, & Moorman, 2001; Weick, 1998), and also

suggests some directions for research in entrepreneurship and dynamic capabilities.

The Role of Improvisation in Capability Development

In a prior study of knowledge-intensive startup firms, Baker, Miner and Eesley (2003)

showed that firms sometimes solve classes of repeated problems by persistently improvising

solutions. That is, they develop capabilities – initially in the absence of supporting routines –

through continuing reliance on improvisation. Baker et al. (2003) called firms’ ability to create

such capabilities based on improvisation “improvisational competencies.” They observed that

firms that developed improvisational competencies in a particular set of activities were less

likely to become competent at creating and subsequently executing new designs in the same

activities, and suggested that this could cause problems for the firms. For example, they

described a firm that became so good at improvising solutions to technical software problems

that it never developed the ability to investigate and design a root solution to fix the fundamental

sources of the problems.

We also found cases in which firms improvise capabilities and sometimes sustain these

capabilities through repeated episodes of improvisation. However, extending Baker et al. (2003),

we observed cases in which firms were able to move from reliance on improvisation as the basis

for capabilities to the development of routines that subsequently supplanted the improvisation.

While the earlier study stands as a warning about the dangers of over reliance on improvisation,



30

our results suggest that improvisation can be an important source of capabilities, in that it can

also lead smoothly to later dynamics that eliminate the need to rely on improvisation.

The Importance of “Stopping Improvisation”

A substantial body of empirical research examines the causes and effects of

organizational improvisation. In most of prior research, improvisation is assumed to be a

temporary response to some emergent organizational challenge and improvisation is implicitly

assumed to end when the challenge ends or is resolved. However, our results suggest that

whether and when organizations stop improvisational activities in favor of other patterns of

behavior – e.g. preplanned activities or routines – can have important consequences for patterns

of capability development.

Entrepreneurship and Dynamic Capabilities

Part of the explanation for why we found that improvisation played such an important

role in capability development undoubtedly derives from our focus on new firms. While much

of the behavior in established firms is structured by routines and by plans and accompanying

budgets (Nelson & Winter, 1982), new organizations struggle – as one of our respondents noted

– to develop “routines in anything.” Much of the time, new firms figure out what to do by

improvising. The results of the improvisation provide an opportunity for experimental learning.

It may be this reliance on improvisation that gives many entrepreneurial firms the sense of

flirting with the edge of chaos but also a degree of fluidity and freedom from excessive inertia.

Our study makes a substantial contribution to the understanding of how new firms develop and

change their capabilities during a period before much behavior has become subject to the

constraints of routinization. Our results also suggest that there may be great value in continued

study of improvisation within entrepreneurial firms.



31

In recent years, scholars interested in organizational renewal and responsiveness in larger

and older firms have increasingly extended the realm of entrepreneurial studies into the domain

of large corporate entities (e.g., Zahra & George, 2002; Zahra, 1996). We suspect that an

important element of what distinguishes “entrepreneurial” large organizations from others

reflects the processes by which these firms develop, maintain and change their capabilities.

Although recent research suggests that routines can be changed on the fly (Feldman &

Pentand, 2003), routines are nonetheless fundamentally stable in many ways (Winter, 2003), and

an organization that relies solely on routines as the foundation of its capabilities seems unlikely

to adapt effectively to rapidly changing problems and opportunities. Indeed, what are now

commonly called “dynamic capabilities” within those established firms that maintain some kind

of entrepreneurial edge may rest in important ways on the ability of these firms to continue

utilizing a broad repertoire of responses – plans, borrowing, stretching and combining routines,

improvisation and even the occasional random flailing around – as the tools by which their

capabilities remain dynamic.



32

Table 1: Definitions of Routines and Capabilities

Definition of Routines

Source Text

Nelson and Winter (1982,p. 97)

‘We use ‘routine’ in a highly flexible way, much as ‘program’ (or, indeed, ‘routine’)is used in discussion of computer programming. It may refer to a repetitive pattern of activity in an entire organization , to an individual skill, or as an adjective, to the

smooth uneventful effectiveness of such an organizational or individualperformance.’

Levitt and March(1988, p. 517)

‘The generic term ‘routines’ includes the forms, rules, procedures, conventions,strategies, and technologies around which organizations are constructed andthrough which they operate.

Feldman(1989, p. 136)

Organizational routines are ‘complex sets of interlocking behaviors held in placethrough common agreement on the relevant roles and expectations.’

Miner(1991, p. 378)

‘I define an organizational routine as a coordinated , repetitive set of organizationalactivities.’

Pentland and Rueter(1994, p. 484)

‘A set of functionally similar patterns.’

Cohen and Bacdayan(1994, p. 406)

‘By ‘organizational routines’ we mean patterned sequences of learned behaviorinvolving multiple actors who are linked by relations of communication and/or

authority.’Feldman (2000) ‘…routines are repeated patterns of behavior that are bound by rules and customs

and that do not change very much from one interaction to another.’

Definition of Capabilities

Source TextBarney(1992, p. 44)

Capabilities as those organizational characteristics that ‘enable an organization toconceive, choose and implement strategies.’

Stalk, Evans and Shulman(1992, p. 62)

A capability is a set of business processes strategically understood.

Nelson and Winter(1982, p. 103)

Capability: ‘the repertoires of organizations members’ that are ‘associated with thepossession of particular collections’ of resources.

Szulanski (1996, p. 28) Organizational capability as best practice.

Winter(2000, p. 983) An organizational capability is a high-level routine (or collections of routine) that,together with its implementing input flows, confers upon an organization’smanagement a set of decision options for producing significant outputs of aparticular type.

Amit and Schoemaker(1993, p. 35)

Capabilities refer to a firm’s capacity to deploy resources, usually in combination,using organizational processes, to effect a desired end.

Leonard-Barton(1992, p. 113)

Core capabilities: the knowledge set that distinguishes and provides a competitiveadvantage. Four dimensions: knowledge and skills; technical systems; managerialsystems; values and norms.

Grant(1996, p.377)

Organizational capability: a firm’s ability to perform repeatedly a productive task which relates either directly or indirectly to a firm’s capacity for creating valuethrough affecting the transformation of inputs into outputs.

Eisenhardt and Martin(2000, p. 1107)

Dynamic capabilities are the antecedent organizational and strategic routines bywhich managers alter their resource base – acquire and shed resources, integratethem together, and recombine them – to generate new value-creating strategies.

Teece et al.(1997, p. 516)

Dynamic capabilities: firm’s ability to integrate, build, and reconfigure internal andexternal competences to address rapidly changing environments.

Zollo and Winter(2002, p. 340)

A dynamic capability is a learned and stable pattern of collective activity throughwhich the organization systematically generates and modifies its operating routinesin pursuit of improved effectiveness.

Levinthal and Myatt(1994, p. 46)

Capability is evolving as a result of the coupling between industry-level forces andfirm-level dynamics.



33

Table 1: Descriptive Summary of SampleType Year of

FoundingNumber of Employees

Location # of Founders

# of Interviews

Length of Transcripts

UniversityStartup

Software 1999 5 Dane 1 1 45 NInternet 1998 5 Dane 2 1 43 NElectronics 1996 7 Dane 1 1 29 NSoftware 1999 7 Dane 1 1 23 Y

Instrument 1980 3 Dane 1 1 16 YBiotech 1997 2 Dane 2 1 12 YBiotech 1984 7 Dane 4 1 16 YBiotech 2000 5 Dane 2 1 41 YSoftware 1996 5 Dane 4 1 37 YRecycling 1999 9 Dane 2 1 24 YCoating 1996 5 Dane 1 1 34 NInternet 1996 14 Dane 4 1 31 NSoftware 1996 4 Dane 1 1 18 YBiotech 2001 8 Dane 2 1 16 YEngineeringConsulting

1997 4 Dane 6 1 26 Y

Internet 1999 7 Dane 1 1 28 YInternet 1995 6 Dane 3 1 30 NEngineeringConsulting

1997 13 Dane 1 1 27 Y

Software 1997 3 Dane 1 1 30 YIT Consulting 1996 22 Dane 1 1 36 N

EngineeringConsulting

1998 2 Dane 2 1 16 N

EngineeringConsulting

1996 6 Dane 6 1 11 Y

Software 1999 5 Dane 1 1 33 NInternet 1999 3 Dane 3 1 36 NInternet 1999 10 Dane 2 1 19 NBiotech 1998 35 Dane 4 1 31 NInternet 1998 6 Dane 2 1 23 NInternet 1998 18 Dane 3 1 42 NBiotech 1997 30 Dane 3 1 20 NInternet 1995 6 Dane 1 1 41 NInternet 2000 22 Dane 5 1 44 NSoftware 1996 7 Dane 3 1 52 NIT Consulting 1998 3 Dane 3 1 27 NIT Consulting 1999 1 Dane 1 1 33 NBiotech 1998 6 Dane 1 1 26 YIT Consulting 2001 4 Dane 3 1 27 N

IT Service 1998 3 Dane 2 1 26 NBiotech 1990 1 Dane 1 1 16 YIT Consulting 2000 4 Dane 4 1 37 NBiotech 1998 4 Dane 4 1 21 YBiotech 1997 4 Dane 4 1 20 YBiotech 1995 36 Dane 3 1 17 YIT Consulting 1999 52 Dane 1 1 27 NOptical Fiber 1999 3 Dane 1 1 27 NInternet 1996 8 Dane 1 1 39 NInternet 2000 3 Dane 3 1 31 NIT Consulting 1995 2 Dane 2 1 20 NIT Consulting 1999 3 Dane 3 1 42 NBiotech 1997 11 Dane 3 1 31 YSoftware 1996 2 Dane 3 1 42 NBiotech 1997 3 Dane 3 1 18 YSoftware 1996 12 Dane 2 1 31 YBiotech 1996 11 Dane 3 1 40 YIT Consulting 1996 54 Dane 2 1 26 NElectronics 1995 55 Dane 1 1 12 YEngineeringConsulting

2000 10 Dane 1 1 36 N

Internet 1997 17 Dane 5 1 36 NIT Consulting 1997 49 Dane 1 1 42 NBiomedicalEquipment

1991 135 Dane 2 1 21 Y

Software 1996 5 Dane 2 1 24 N

TOTAL 60 1725 (pages)



34

Founding team

Firm A New Firm

Replication/Transplantation

Figure 1: Importing routines from prior organizations

Network

Routine A

Combination

New Firm

Founder 1 Founder 2Routine RoutineA1 A2

Founder 3 Routine

A3

Figure 2: Integrating routines from founders and networks

New Firm at t

Org. Routine A

New Firm at t+1 Org. Routine A

Org. Routine B

Extension

Figure 3: Creating routines from real-time experience



35

Improvisation Routine Capability

Onl im rovisationOnly routines

Only capability

Figure 4: Trajectory I – From Routine to Capabilities

Only improvisationOnly routine

Only capability

ImprovisationRoutine

Figure 5: Trajectory II – From Capabilities to Routines

Capabilities

Other behavior

Routine Ca abilities

Only routine Onl ca abilit

PlanningPlanned behaviorImprovisation

Random

Figure 6: Relationship between Routines, Capabilities and other Behaviors



36

REFERENCES

Argote, L. 1999. Organizational Learning: Creating, Retaining, and Transferring Knowledge.Norwell, MA: Kluwer.

Amit, R. & Schoemaker, P. J. H. 1993. Strategic assets and organizational rent. StrategicManagement Journal, 14(1): 33-46.

Baker, T., Miner, A., & Eesley, D. 2003. Improvising Firms: Bricolage, Retrospective Interpretation andImprovisational Competencies in the Founding Process. Research Policy, 32: 255-276.

Baker, T. & Nelson, R. E. 2004. Making Do with What's At Hand: Entrepreneurial Bricolage. WorkingPaper.

Barney, J. 1991. Firm resources and sustained competitive advantage. Journal of Management, 17: 99-120.

Barney, J. B. 1992. Integrating organizational behavior and strategy formulation research: Aresource based analysis. In P. Shrivastava & A. Huff & J. M. Dutton (Eds.), Advances inStrategic Management, Vol. 8: 39-62. Greenwich, CT: JAI Press.

Baum, J. A. C. & Singh, J. V. 1994. Evolutionary dynamics of organizations. New York: OxfordUniversity Press.

Boulding, K. E. 1956. General systems theory: The skeleton of science. Management Science, 2:

197-208.Burgelman, R. A. 1994. Fading memories: A process theory of strategic business exit in dynamic

environments. Administrative Science Quarterly, 39: 24-56.Campbell, D. T. 1969. Variation and selective retention in socio-cultural evolution. General Systems, 16:

69-85.Cohen, M. D. & Bacdayan, P. 1994. Organizational routines are stored as procedural memory: Evidence

from a laboratory study. Organization Science, 5: 554-68.Cohen, W. M. & Levinthal, D. A. 1990. Absorptive capacity: A new perspective on learning and

innovation. Administrative Science Quarterly, 35: 128-152.Cohen, W. M. & Levinthal, D. A. 1994. Fortune Favors the Prepared Firm. Management Science, 40(2):

227-251.Cyert, R. M. & March, J. G. 1963. A Behavioral Theory of the Firm. Englewood Cliffs, NJ: Prentice-

Hall.Denzin, N. K. & Lincoln, Y. S. 1998. Collecting and Interpreting Qualitative Materials. Thousand Oaks,

CA: Sage.Eisenhardt, K. M. & Martin, J. A. 2000. Dynamic capabilities: The evolution of resources in

dynamic markets. Strategic Management Journal, 21: 1105-1121.Feldman, M. S. 1989. Order without Design. Information Production and Policy Making, Stanford, CA,

USA: Stanford University Press.Feldman, M. S. 2000. Organizational routines as a source of continuous change. Organization Science,

11: 611-629.Feldman, M. S. & Pentand, B. T. 2003. Reconceptualizing Organizational Routines as a Source of

Flexibility and Change. Administrative Science Quarterly, 48: 98-118.Garud, R. & Karnoe, P. 2003. Bricolage versus breakthrough: distributed and embedded agency in

technology entrepreneurship. Research Policy, 32(2): 277-300.Grant, R. M. 1996. Prospering in dynamically-competitive environments: Organizational

capability as knowledge integration. Organization Science, 7(4): 375-387.Hansen, M. 1999. The search-transfer problem: The role of weak ties in sharing knowledge across

organization subunits. Administrative Science Quarterly, 44: 82-111Haunschild, P. R. & Miner, A. S. 1997. Modes of interorganizational imitation: The effects of outcome

salience and uncertainty. Administrative Science Quarterly, 42(3): 472-500.Huber, G. P. 1991. Organizational learning: The contributing processes and the literatures. Organization

Science, 1: 88-115.



37

Hutchins, E. 1990. The technology of team navigation. In J. Galegher, R. Kraut, & E. Carmen (Eds.),Intellectual Teamwork: Social and Technological Foundations of Cooperative Work: 221-242.Hillsdale, NJ: Lawrence Erlbaum.

Kim, J.-Y. J. & Miner, A. S. 2000. Crash test without dummies: A longitudinal study of interorganizational learning from failure experience in the U.S. commercial banking industry:1984-1998. Academy of Management Proceedings, OMT: G1.

Leonard-Barton, D. 1995. Wellsprings of Knowledge: Building and Sustaining the Sources of Innovation.Boston, MA: Harvard Business School Press.

Levitt, B. & March, J. G. 1988. Organizational learning. Annual Review of Sociology, 14: 319-40.Lounama, P. H. & March, J. G. 1987. Adaptive coordination of a learning team. Management

Science, 33: 107-123.Miner, A. S. 1991. Organizational Evolution and the Social Ecology of Jobs. American Sociological

Review, 56(6): 772-785.Miner, A. S., Bassoff, P., & Moorman, C. 2001. Organizational Improvisation and Learning: A Field

Study. Administrative Science Quarterly, 46: 304-337.Miner, A. S. & Haunschild, P. R. 1995. Population-Level Learning, Research in Organizational Behavior:

An Annual Series of Analytical Essays and Critical Reviews, Vol 17, 1995, Vol. 17: 115-166.Moorman, C. & Miner, A. S. 1998. Organizational improvisation and organizational memory. Academy

of Management Review. 23: 698-723.Nelson, R. R. & Winter, S. G. 1982. An Evolutionary Theory of Economic Change. Cambridge, MA:

Harvard University Press.Ocasio, W. 1997. Towards an attention-based view of the firm. Strategic Management Journal, 18: 187-

201. Paulhus, D. L. & Martin, C. L. 1997. The structure of personality capabilities. Journal of Personality and

Social Psychology. 52: 354-365.Ragin, C. C. 1987. The Comparative Method: Moving Beyond Qualitative and Quantitative Strategies.

Berkeley, CA: University of California Press.Rittle-Johnson, B., Siegler, R. S., & Alibali, M. W. 2001. Developing conceptual understanding and

procedural skills in mathematics: An interactive process. Journal of Educational Psychology, 93:346-362.

Sobocinski, P. Z. 1999. Creating High-Tech Business Growth in Wisconsin: UW-Madison TechnologyTransfer & Entrepreneurship. Madison, WI: University-Industry Relations.

Sitkin, S. B. 1992. Learning through failure: The strategy of small Losses, Research in OrganizationalBehavior, 14: 231-266.

Stalk, G., Evans, P., & Shulman, L. E. 1992. Competing on capabilities: The new rules of corporate strategy. Harvard Business Review(March-April): 57-69.

Stinchcombe, A. L. 1990. Information and Organizations. Berkeley, CA: University of California Press.Strauss, A. & Corbin, J. 1998. Basics of Qualitative Research. Thousand Oaks, CA: Sage.Szulanski, G. 1996. Exploring internal stickiness: Impediments to the transfer of best practice

within the firm. Strategic Management Journal, 17(Special Issue: Knowledge and theFirm): 27-43.

Teece, D. J., Pisano, G., & Shuen, A. 1997. Dynamic capabilities and strategic management. Strategic

Management Journal, 18: 509-533.Walsh, J. P. 1995. Managerial and organizational cognition: Notes from a trip down memory lane.

Organization Science, 6(3): 280-321.Weick, K. E. 1993. The collapse of sensemaking in organizations: The Mann Gulch disaster.

Administrative Science Quarterly, 38: 628-652.Winter, S. G. 2000. The satisfying principle in capability learning. Strategic Management Journal, 21:

981-996.Winter, S. G. 2003. Understanding dynamic capabilities. Strategic Management Journal, 24: 991-995.Zahra, S. A. 1996. Governance, ownership, and corporate entrepreneurship: The moderating



impact of industry technological opportunities. Academy of Management Journal, 39(6):1713-1735.

Zahra, S. A. & George, G. 2002. Absorptive capacity: A review, reconceptualization, and extension.Academy of Management Review, 27(2): 185-203.

Zollo, M. & Winter, S. G. 2001. Deliberate learning and the evolution of dynamic capabilities.Organization Science, 13(3): 339-351.

Miner WP 2005 - The Dynamics of Routines and Capabilities in New Firms

Documents

Transcript of Miner WP 2005 - The Dynamics of Routines and Capabilities in New Firms