Supporting Material is available for this work. For more information, follow the link from
the Table of Contents to "Accessing Supporting Material".
Table of Contents
The Strategic Impact of Balancing C apabilities
Abhijit Mandal
University of Warwick
Warwick Business School
Coventry, CV4 7AL
Phone: 024 7652 4504
Abhijit. Mandal@ wbs.ac.uk
ABSTRACT:
In strategic management, changes in strategic positioning and dynamic capabilities have
been recognized as rational and deliberate responses by the top management team to a felt
need for attaining improved organizational performance. However, there may be a delay in
responding to the real-world challenges in this manner. We propose the notion that middle
level managers use their executive skills to balance different responsibilities (which we
call “dynamic balancing capabilities’, DBC) to respond to challenges from the
environment in a relatively shorter time-frame. Through system dynamics modeling we
show that variations in DBC can differentially shape the overall context and thereby
influence the flow and accumulation of resources, leading to differential performance over
time potentially resulting in explicit changes of strategic positioning without the
involvement of top management. Also, concepts like “key success factors”, “best practice”
and “critical resources” are usually employed in a static sense. However, our modeling
tesults reveal that so called critical resources derive their potency from the particular
dynamics of the existing situation. At different times the same resources will play different
toles and will therefore not always be of critical importance.
THE STRATEGIC IMPACT OF BALANCING CAPABILITIES
Conventional wisdom in strategic management suggests that the strategic position of a firm
(Porter, 1996) is decided by the decisions taken by senior managers as part of its top
management team. From this it follows that the firm’s dynamic capabilities (Teece et al,
1997) would be decided to a significant extent by the firm's desired strategic position’. Taken
together, the firm’s strategic position and its dynamic capabilities determine the pattem of
allocation of scarce resources which has then to be executed by middle management /
functional heads of the firm at the operational level. Animplication of this arrangement is
that while senior management will take its time to deliberate over whether a firm should
change its strategic position or not due to a change in circumstances. In contrast, middle
management has to take environmental change in its stride and is required to minimize
disruptions to the essential operations of the firm and to its intended strategic position
In light of these typical middle management responsibilities, we ask whether it is possible for
them to bring about a shift in the intended strategic position of the firm. The objective of this
paper is to explore the mechanism of such a change. Of particular importance is to determine
the key for middle management which enables them either to maintain the same strategic
position as desired by the top management or deviate from it. For this we need to examine a
situation where two firms have initially the same strategic positioning and dynamic
capabilities while being very closely matched to each other in other respects; yet show a
divergence in performance with the passage of time.
Middle management implements strategies through operations. From the resource-based
approach (Wemerfelt, 1984; Bamey, 1986, 1991; Peteraf, 1993) it can be stated that they
execute various processes that transform resources in a cyclic manner (from cash to inputs to
output to cash). Thus, in order to model the resource-transformation processes we draw upon
a well-accepted scheme of conceptualizing resources (Dierickx & Cool, 1989). Adopting this
scheme implies that resources whether they are intangible or tangible are seen as stocks. If
the collection of resource-stocks that exist within a firm differs from another in name or stock
level, then we can say that the resources display heterogeneity. Further, a difference in the
1 Porter identifies the generic strategic positions related cost, focus and differentiation. Conceptually, the
strategic position of a firm expresses the distinguishing features of products/services the firm wants to sell and
levels of critical resource-stocks indicates a difference in performance. Further, the manner in
which the resource-stocks link to each other at any given point in time is designated as the
Tesource structure while the reciprocal influence that a resources has on another resource
directly or indirectly with the passage of time is known as resource interactions. The
transformation of various resources as seen over time results from the policies followed by
the management of the firm.
This paper models the evolution a situation where two very similar units of a fim commence
with identical performance but later diverge. In the next section we present empirical
evidence of the phenomenon. The third section presents details about how the firms were
modeled while the fourth section presents the simulation experiments and its analysis. The
fifth section conc ludes.
EMPIRICAL CONTEXT AND EVIDENCE
Industry Events
In the early ‘90s, the life-insurance industry in the United Kingdom witnessed a number of
financial scandals resulting from customer and client complaints. The common theme in
these complaints were that the insurance firm was either overcharging premium for policies
or selling the customer / client policies that were not suited to their requirements. The
goverment responded in the mid-‘90s by imposing new regulations on all life insurance that
was being sold in the United Kingdom through the industry regulator that is currently known
as the Financial Services Authority. These new regulations imposed greater disclosure of the
various charges that were present in the policies. This increased transparency had two direct
consequences: first, it meant more paper work for the agents and managers and second, it
increased compliance costs for the life insurance selling organizations to remain on the right
side of the law. These direct consequences implied that they had to allocate time exclusively
for this purpose, which detracted from the time they could allocate to their traditional
responsibilities.
In addition to the direct consequences, there were indirect consequences too. Greater
transparency imposed through regulations led to increased transparency in the cost structure.
how it organizes its activities within the firm to make a value proposition to its customers. Dynamic capabilities
The unforeseen (and undesired from the insurance provider's point of view) consequence was
that the increased information led to an intensification of competition on costs that put prof it
margins at every step of the process under pressure. Although this development required a
decision by the senior-most levels of management, in most firms this level remained inactive.
The whittling away of margins hit the middle segments of the market the hardest; the lower
segment (which became increasingly mechanized) and the upper segments (which already
involved customization through specialists) were relatively less affected. Given the relatively
lower returns from agents in the middle segment of the market, almost all the top players
eventually abandoned this segment of the market and effectively disbanded their sales forces
that catered to this section of the market.
The sales forces that survived did so by moving up market - in terms of selling more
customized products through better trained salespersons. An ideal example is Life Assurance
Holding Corporation (managed by its major shareholder, St. James’s Place Capital). Here is
an extract from an interview with Sir Mark Weinberg, Executive Director of St. James's
Place Capital plc and Chairman of Life Assurance Holding Corporation Limited:
"The way to look at our Edinburgh office is as a home to 36 people who are running their
practice we choose to use the word partner to give everyone the feeling they are working with
each other. They all have specialties that they can help each other with to form a long-term
relationship with clients. We are different to the likes of Standard Life and the Prudential,
because they are manufacturers of products, they don't give advice. We are first and foremost
an advisory organization. St James's Place Capital is in the business of increasing the size of
the partnership by 5% to 10% per annum," he says.
But, at a time when life companies are slashing their sales forces, he warns that SJ PC is not
about to take on just any old refugees. "A traditional sales force is in the business of
expanding for expansions sake. We do it the other way round; yes, we'd like to attract more
people but they must not depreciate the quality or the standards of the branch. About 80% of
another company's sales force wouldn't be good enough to work for us. Recruiting for us isa
one-by-one process."
refer to the meta capabilities of the firm used to build new capabilities and modify existing capabilities.
An important result from these events is the trend in the insurance industry towards
commoditization and specialization of service required by policyholders. Many large
insurance firms now sell certain types of insurance (e.g. travel insurance, basic life insurance)
as a commodity directly to prospective policyholders. This is executed in a manner equivalent
to direct marketing through mail-order, remote terminals, the Internet, at supermarkets, at
travel agencies, etc. - thereby avoiding the use of the traditional agent. On the other side,
agents now seek to develop customized insurance packages to meet the special needs of
sophisticated customers - requiring agents to have a well-developed problem solving ability
in the context of their clients’ business. On the whole there is a trend towards commission-
less agents or at least a significant reduction in the commission component of an agent's
emoluments.
The implication of this trend is important for management of insurance firms and agents.
More resources are now diverted towards creating a new type of sales-force - one that is
quick at learning how to tackle issues specific to their clients rather than that goes all out for
increasing the volume of sales. It is agency management that continues to play the critical
tole in delivering a competitive advantage to the insurance firm.
Just as there was a bifurcation in the industry, the specific phenomenon that we are going to
investigate in the firm of interest is also about divergence. Before we present the relevant
details about this firm, we present some general characteristics of the insurance industry and
its constituent firms.
General Characteristics of the Industry and Firms
The insurance industry is an international social institution which has accelerated the overall
growth of business by bringing security, credit and efficiency benefits to individuals,
organizations (including firms) and communities. The function of insurance is to combine a
large number of risks and thus reduce the degree of uncertainty. Insurance may be defined as
a combination of individuals who agree to make small contributions in order to reimburse
those who suffer losses from events that may be foreseen and estimated but whose occurrence
may not be accurately predicted. Thus, the scope of the industry is to enable persons avoid
the financial consequences of risks or uncertain events (Riegel and Miller, 1966). The
industry goes about its task by selling “policies” to those who want to insure themselves.
Policyholders pay relatively small “premiums” (in most cases annually) and make claims
with the insurance company if and when they suffer losses covered by their policies. In the
meantime, the insurance firm is supposed to invest and grow the funds that accrue to them
with a view to keeping them secure to meet pay- out needs and return funds to mature policy-
holders as and when necessary.
Given this scope, major players in the insurance industry have a broad set of tasks, which are
common to them. Figure 1 gives an overview of the typical insurance firm. Their tasks
consist of selling insurance, selecting risks, fixing premiums, writing policies, investing
money, keeping accounts, collecting, researching and analyzing statistics, processing claims
and dealing with legal issues and cases. To execute these tasks they need either to build these
required skills as individual firms, or share them from a common pool - depending upon the
quantum of required investment and the scope for differentiation. Selecting risks and fixing
premiums are not totally within the control of the individual firm, as they have to adhere to
industry standards and regulations. Together with writing policies, this is the responsibility of
the underwriting department, a cost center. Given that new types of policies can easily be
copied, it is really difficult to establish a sustained differentiation with respect to competitors.
Keeping accounts confidential is of course important to each firm; it is obviously not a shared
activity and is charged to the accounting cost center. However, collecting, researching and
analyzing statistics is an activity that gains value with increase of scale; it is therefore a
pooled activity. The claims department (a cost center) is responsible for processing claims
and the legal issues involved therein. The scope to differentiate here is limited, as no firm
would want to either establish a reputation for compromising on payments or take a hit on its
profitability by relaxing payment standards. In contrast, the investment department (a profit
center), which invests the incoming premiums, performs an activity that is hardly unique to
the insurance industry - this activity being similar to that of mutual funds and other
intermediary players in the financial market. In some countries, the financial performance of
the investment department is legally kept apart from the rest of the organization. This brings
us to the remaining activity - the selling of insurance policies.
As in other sales organizations that sell to the masses, there is scope for generating demand
through push (sales) and pull (marketing). The scope for differentiation through marketing is
limited, as it is difficult if not outright impossible to compete on prices and very difficult to
sell differentiated products (policies) on a sustained basis. This is reflected in the relatively
small amount of funds put aside for marketing campaigns when compared to the funds put
aside for the agency department, which handles the sales agents. There exist different kinds
of agency systems in the insurance industry - e.g. the general agency system and the branch
office system in life insurance, the independent agency system and the exclusive agency
system in property and casualty insurance. Common to these is the practice of selling policies
through agents who receive a commission for the sale of these policies. The difference in
these systems lies in the degree of control that the management of the firm exercises over its
agents and the structure of their compensation. It is also the responsibility of the agents to
minimize “lapses” in the policies they sell. A “lapse” occurs when the client discontinues
payment of premiums towards an insurance contract before the contract permits.
It is the agency department that affords the greatest flexibility to firm management when it
seeks to establish a competitive advantage through differentiation and productivity’,
Management decides what kind of agents to hire, how much training they should get, how to
train them, where to spread its agents and how to identify and retain / promote its star
performers. The agency department is responsible for the flow of new money streams to the
organization as well as for maintaining the “going concem” status. It is also the largest cost
item that can actively be managed in the business plan of the insurance company. It is the
quality of agents and their performance that is the most influential in determining a firm’s
profitability in the insurance industry.
Details of the Phenomenon
Our story is about one of the largest insurance providers in the United Kingdom. This
particular insurance provider, like others of similar sizes, consisted of a head-office with
numerous branch offices spread across the country. This structure made it necessary that a
strategic review of the firm examine the performance of the branches at the branch level as a
preliminary step This examination of the different branches yielded detailed data about
performance of the branches in getting new business including characteristics about the sales
personnel, with the data about the sales personnel aggregated to the branch level. Analysis
revealed thatthere were many branches whose performance was significantly below the
average of all the branches while there were a few that were excellent and above average .
? another way in which firms in the industry differentiate itself is through the patter of ownership of the equity
structure of the firm. However, this makes an impact only during exceptional events in the history of the firm
rather than in the day-to-day activities and competitive advantage or competitive dynamics of the firm.
We prepared a histogram of the annual average sales productivity performance of each
branch. From this distribution, we aggregated the branch level data into ten deciles. We found
it useful to focus our attention on three of these groups that covered the entire range of
branches. These groups represent the top 20%, the middle 20% and the bottom 20% of the
frequency distribution, based on the annual average sales productivity performance of the
various branches. This information is portrayed in Figure 2. All three groups showed an
increase in productivity with respect to time, but the productivity of the group representing
the top 20% of the branches seems to have increased significantly more than the productivity
of the group representing the bottom 20%. To get a more accurate perspective, we decided to
filter out the background growth by subtracting the growth rate of the average performers.
This led us to normalize the above graph, and the resultant graph is shown below.
From our knowledge of the industry we know that management would have had a lot of
freedom in the selection, training and retention of agents (sales persons) in order to generate
new revenue. However, the reality was that these agency departments, belonging to the same
insurance firm, were subject to the policies of the general managers at the headquarters which
were the same for all branches. This leads to a paradox: though managers had very limited
freedom to pursue policies different from those specified by the headquarters, yet there was
dispersion in the performance of the different branches around the average performance level.
With the objective of explaining this paradox, we will take a closer look at the agency
department of firms in the insurance industry. We next describe the structure and activities of
an agency department in a typical major firm in the insurance industry. This will give us a
more comprehensive idea about the nature of resources that are employed by the agency
departments of firms in the insurance industry, and how these resources are typically
deployed and managed. Subsequently, we will represent the structure and the activities in a
model of a stylized firm. Figure 4 points out which portions of the insurance firm would be
the subjects of our study. The model of the insurance firm will include only the portions that
fall within the marked boundary. By simulating the behavior of this stylized firm, we will
attempt to explain why a dispersion occurs in the performance across the various branches,
and how did this dispersion develop.
MODELLING THE FIRM
The role of the Agency
As mentioned above, we focus on the agency department of large firms in the insurance
industry. Unless specifically mentioned otherwise, from now on firms will refer to the agency
department of insurance industry firms. Firms to sell policies to those who want to buy
insurance. These policies (also called products) are of varying durationand provide the firm
with premiums for the length of the life of the product, if they do not "lapse". The larger the
product base (i.e. the inventory of live policies sold by the firm), the larger the cash flow and
revenue to the insurance firm. The primary driver of the performance of an agent is the skill
that he or she possesses. Sales are in a cycle of 3 stages. First, agents are recruited from the
market as employees, to be part d the agent body that sells policies to prospective customers
in the market. Firms always seek to hire more experienced agents from the market and to
tetain the better: performing agents (as they are economically more attractive for the firm).
Second, accumulated policy sales of policies in force form the basis of the future revenue
stream (as premiums). Policy sales and lapses are a function of the agents’ skill level. Third,
agents are compensated based on the sales made and lapses occurred in that particular year.
By joining a firm, agents increase the headcount of sales employees and add their sales skills
to the skill pool of the firm. From time to time, some agents quit the firm and some are
promoted. These decrease the headcount of sales employees and the aggregate skill pool of
the firm. If agents quitting the firm have lower than average skill level, those promoted will
have a higher than average skill level. It is thus a challenge to management to maintain and
improve the skill level of their agent base. A gent compensation is very important. If agents
perform above the performance level expected by the market (i.e. the average skill level
prevalent in the market, which is assumed to be 3 years in the simulations), the resulting
compensation is above market expectations. Lower than market standards performance
means inferior compensation. In turn, compensation affects the quit rate of agents (which
influences the lapse rate of new policies sold) as well as the attractiveness of the firm to new
agents who are considering whether to join the firm.
Managers supervise these sales agents. These managers are responsible for recruiting and
training the agents, besides monitoring their performance. Usually these managers are
themselves former agents who have been promoted into this role. Promotion from within is
an established industry practice. Managers at the branch level have to allocate their time
amongst their various responsibilities to see that none of the necessary aspects are being
ignored. Simultaneously, they have to implement the growth targets that are set by the senior
management in the headquarters - e.g. the headquarters may announce a rate of growth for
the agencies that has to be met at by the manager of the different branches.
Model Development
The model was developed in three stages: data collection, formulation and validation. The
field work during data collection was done with the help of a management consulting firm. In
this stage, more than 60 interviews were undertaken by four mid-level consultants, involving
around 40 managers as well as some ex-managers of the various branches and the corporate
office of the insurance firm in question; many of the senior managers were interviewed more
than once. These hour-long interviews were carried out in a semi-structured manner where
they were asked about the execution of their responsibilities. The aim of the interviewers was
to develop an understanding of the various policies, processes, process adjustments and
informal targets that were active. Wherever possible, written data and internal studies were
used to support and verify claims. Statistics about salesmen and their performance was a key
part of this; the numerical data about salesmen was analyzed to reveal longitudinal trends.
Interview content was probed to see which aspects of management were common to all and
which were distinct. The quotes were very useful in clustering the branches according to
differences in management practice and performance.
In the second stage, we developed a the ory to explain the evolution of different branches,
with the help of causal loop diagrams*, The variables and causal links from this analysis
include the feedback processes that generate the dynamics of interest. Separate diagrams
were created for each cluster of branches. These were merged to a unified framework so that
a single set of feedback processes could generate the main trajectories of interest.
Subsequently, each causal link was converted to equations or graphs to form the model where
the variable s vary with time. Literatures in various relevant disciplines were consulted to
justify the causal relationships. The model was further calibrated with the help of the
available numerical data. In the last stage, three independent industry experts validated the
behavior of the model by conducting their own experiments. Since these results conformed to
their expectations, we assume that the link between the structure of the model and its field
setting is valid. Further experiments were conducted by two industry experts who were on the
3 See Repenning & Sterman (2002) for precedents and Sterman (2000) for a description of the procedure.
10
research team. This model excludes the attempted tum-around efforts by managers when they
became aware of their worsening position, as it is not the focal objective of this paper. A
description of the core of the model follows whose overview is sketched in Figure 5.
The core of the firm consists of four sectors that address headcount, skill, productivity and
compensation. The headcount sector is the collection of processes directly concerned with
managing agent headcount, consisting of hiring, firing and promotions to grow the firm.
Managers recruit agents from the market to be part of the agent body that sells policies to
prospective customers. Firms seek to hire more experienced agents and retain the better
performing agents as they deliver more value. A fraction of the agents are always moving out
through resignations and firings. Promotions are decided in-house and the rate depends on
managerial vacancies, as the ratio of managers to agents is legally regulated. The number of
agents and the overall rates of agents moving into or leaving the firm have an important
impact on the dynamics of the skill pool of the agents.
The skill sector is about the management of agents’ sales skills. For simplification, multiple
dimensions of sales skill have been collapsed into one dimension, which is based on the years
of sales experience possessed by an agent. Even though the measure of this skill is fairly
intangible, it is one of the most important drivers of performance in the industry. The
movement of agents into and out of the firm has a corresponding impact on the firm’s skill
pool. The higher the skill level of those quitting the firm, the greater is the depletion of the
skill pool. It is thus a most important challenge to management to maintain and improve the
skill level of their agent pool. The productivity sector highlights the productivity and tumover
of the sales agents which adds to the product portfolio. The sales function provides the firm
with premiums for the life of the product's length, if they do not lapse. Policy sales and
lapses are a function of the agents’ skill level. The larger the product base which is defined as
the inventory of live policies sold by the firm, the larger the cash flow and revenue to the
insurance firm. We use skill level per agent in the firm as the key productivity and
profitability indicator.
The compensation sector models the mechanics of fixed and variable compensation, or
commission, for agents and managers. A gents are compensated based on the sales made and
the lapses that occurred in a particular year. It specifies how the level of skills, the lapse rate
and the quit rate of the agents affect compensation and, in tum, how the compensation level
11
affects the same three variables. If agents perform above the performance level expected by
the market, the resulting compensation is above market expectations. A performance lower
than market standards of performance, brings inferior compensation. Thus, the compensation
affects the quit rate of agents, which influences the lapse rate of new policies sold, as well as
the attractiveness of the firm to new agents who are considering whether to join the firm.
Sectors addressing the comparison of performance and managers’ reaction to the comparison
are outside the direct scope of the model; nevertheless these issues have been addressed after
the next section.
SIMULATION: EXPERIMENT DESIGN, RESULTS AND ANALYSES
There will be two experiments. Each of the experiments will show the trajectory of the time-
path of two firms (Alpha and Beta) having identical resource structure and policies regarding
strategic positioning but differing slightly in one of the policies that represent the operational
tole of middle management. Thus, any differential performance of the firms in an experiment
will be the result of a difference in such a policy. The simulations progressively add partial
models to the basic core structure detailed above. Its advantage is that the link between
structure and behavior is easier to grasp when the structure is developed in stages, with access
to intermediate results (Morecroft, 1984, 1985; Sterman, 2000). The addition of structure
through partial model creates new contexts, but otherwise the initial heterogeneity among the
two firms would be maintained. Comparing the results of one experiment with another will
show how this specific difference in policy and consequent resource interactions changes the
magnitude of resource heterogeneity and differential performance over time, resulting in
strategic consequences.
So far, the core model assumes that managers have infinite time. In reality, managers have to
allocate their time amongst different responsibilities such as recruiting, coaching and admin-
istration. In accordance with the actual events of the industry, allocation to administration
commences only some time after the simulation has started. In this case, managers must
allocate 20% of their time to that activity while the time needed for the other two activities
depend on the physical constraints implied by the chosen strategic position: i.e. to make up
for those who leave, to meet growth targets set by the top management team, and to improve
average firm productivity. However, the amount of time actually spent to train agents
depends on the managers’ subjective beliefs about its efficacy. It implies a situation where
they are free to adjust their allocation of time in different circumstances. This intangible,
12
idiosyncratic but subtle adjustment is an example of the balancing activity that middle
management is expected to perform. Since the precise allocation in this act of balance is
likely to vary, it introduces heterogeneity in the policies followed. The activities of recruiting
and coaching are detailed in the next section while the manner in which time is allocated
between the two is explained after that.
Recruiting and Coaching
Ina typical market for agents, one finds salespeople with different levels of experience
tanging from greenhorns to stars. Usually, senior management decide s the profile of
salespersons to be recruited for their firm, depending on the firm’ strategic intent. The effort
involved in recruiting a profile of experienced salespersons is, in a quantitative sense, similar
to that involved in recruiting a profile of salespersons with little experience. In effecting the
former profile, the recruiter would seek to determine if the experienced salesperson would be
compatible with the organization from the point of view of organizational culture and work
habits while in the latter, the recruiter would seek to determine the potential ability of the
salesperson and whether that potential can be developed further. These attributes are
intangibles and hence recruiting is not a responsibility that can be executed mechanically.
Even then, firms usually choose a profile dominated either by experienced salespersons or by
inexperienced salespersons , avoiding a mixture. This is because of contrasting beliefs of
management about the abilities of agents in general and the disadvantages in managing both
kinds of profiles. We will delve further into the relevant beliefs and economics when we take
up the implications of ‘coaching’. Managers, who are in charge of overseeing agents, are
typically in charge of recruiting. In keeping with the profile of recruits preferred, managers
interview applicants, determine which applicants are best suited to their firm and to what
extent they would cohere with their firm’s culture. Then, they make offers to the suitable
applicants, and if necessary, negotiate with them in order to persuade them to join the firm.
Quite a few managers in the industry are convinced that recnuiting is the only way to add
skills to their agent base. For them, sales skill is something innate; thus agents are incapable
of systematically improving their sales skills. These managers believe that agents with
inadequate skills will separate themselves from those with adequate skills in the course of
time. Therefore, there is really no need to track the development of an agent's skills. In
addition, the immediate return from hiring inexperienced agents is also greater. Others
13
believe that recruits with little experience can be coached to improve their sales performance
over time. The strategy of recruiting inexperienced agents makes economic sense if they
ultimately acquire skills that are superior to average market expectations. Such managers
spend about 40% their time coaching agents. The structure denoting how agents gain
experience and the impact of overall skills is shown in Figure 6. The Agent Skills Sector
accounts for agent learning representing it by a flow of skill into the stock Agent Sales’ Skill.
Labeled as Agent Coaching, it is an aggregate indicator of the total skill being added to the
skill pool of the entire sales force. Agent Coaching is the product of two quantities: the
number of agents (Agents) and the Agent Coaching Rate. Agent Coaching Rate is the rate at
which agents ‘leam’.
For those who believe that agents can learn, the rate of coaching varies from pool to pool,
implying differences in the rate at which different agents enhance their skills. Some factors
that influence this value are the attitude of management, the age and experience profile of the
agents, the quality of coaching, the coaching infrastructure and the expectation of the agents
themselves. Different experiences and personal learning abilities give rise to differences in
coaching rates. However, we simplify the model by assuming a uniform rate of coaching at
approximately 1 equivalent year of experience per agent per year; implying that an agent
gains this amount of experience for every year passed in the firm. The simple mechanism of
agent coaching as presented here provides no additional levers for managers. If agents learn,
then they gain a certain but fixed amount of experience with the passage of time - the rate of
skill improvementcan’t be altered in the structure shown in Figure 6 when the amount of
managerial attention is fixed
Time Allocation by Managers
This section specifies how managers specify their time allocation to meet responsibilities.
There are three steps involved: time needed for managerial responsibilities, time supplied by
managers (or available time) and the actual allocation procedure. Although managers’ time is
a tangible resource, the allocation of that time to various tasks and the process driving the
allocations can be quite intangible. The bottom third of Figure 7, below the shaded box,
depicts the structure of managerial responsibilities. Recruiting and coaching take up most
managerial time Total Time Needed is a sum of the components of Time Required for Each
Activity, which is a matrix type variable. The two components of the matrix are Time
Required for Each Activity (Recruiting) and Time Required for Each Activity (Coaching) and
14
are derived from the time required for each activity whichis calculated in Indicated
Recruiting Time and Indicated Coaching Time.
Indicated Recruiting Time is a product of the number of recmuits targeted (Target Number of
Recruits) and the amount of time needed to recruit one average agent (Days Needed per
Recruit). Apart from the time spent in interviewing candidates who finally accept the offer of
employment made by the firm to join, this parameter also takes into account the time spent in
interviewing to weed out unsuitable and undesirable candidates and the eventual time lost by
managers in making persuasive offers to attractive prospective agents who decline. Indicated
Coaching Time is a product of the number of agents to be coached (Agents) and the amount
of time ideally required for coaching the average agent (Target Days of Coaching per Agent).
It accounts for the time required for classroom coaching that is partic ularly vital for fresh
recruits and the amount of ‘face-time’ that every agent needs to spend with his manager to
cover monitoring of the quality of sales, short term performance and long-term progress
Some time is spent in motivating agents to aspire for higher sales.
The top third of Figure 7 above the shaded box denotes the structure of the time available for
managers in the firm. Total managerial time available (Managerial Time Available), just
above the shaded box, is a product of three factors: the number of Managers, the Relative
Efficiency of Managers and the number of days that a manager works in a year (Working
Days per Year per Manager). The number of Managers is supplied by the Manager
Manpower Sector while Working Days per Year per Manager is both an indication of the
capacity of managers’ work time and a converter of time from years to days (because it has
the units of days per year). Relative Efficiency of Managers is a function of the Effect of
Relative Manager Skill on Efficiency that indicates how the relative efficiency of a manager
changes with varying values of relative managerial skills.
The independent variable in Effect of Relative Manager Skill on Efficiency is constituted from
the ratio of two factors: Skill per Manager and Standard Manager Skill. The latter, from the
Manager Manpower Sector, indicates the skill level of managers in the firm. The former,
analogous to Standard Agent Skill, is the expected average skill of managers, in tune with
industry standards. It is measured in years of equivalent experience. The lower the value of
Skill per Manager below the industry standard of Standard Manager Skill, more dramatic is
the drop in Relative Efficiency of Managers. However, the increase from higher skills
15
happens to be negligible. This reflects the handicap of inexperienced managers while the
nature of the managers’ tasks is such that experienced managers cannot economize
significantly on their considerable experience.
The central portion of Figure 7 (the shaded box) illustrates the structure allocating available
time for the necessary managerial resporsibilities. It is achieved in three steps. In the first
step, the Shortfall (if any) between Total Time Needed and Managerial Time Available is
calculated. The minimum value, 0, indicates no shortfall; otherwise it indicates the absolute
amount of shortage (in days per year) that needs to be compromised through a reallocation
of the suggested time for the principal activities. In the second step, the absolute amount of
shortfall is separated into two parts (Shortfall of Each Activity ). It is a 2-component matrix
variable. Each part corresponds to the absolute amount of time that must be taken away from
the ideal amount of time desired for each activity (Time Required of Each Activity). This
calculation is a function of four variables and a parameter. We have already met three:
Shortfall of Each Activity, Shortfall and Managerial Time Available. The fourth variable,
Ratio, is the ratio between Indicated Recruiting Time and Indicated Coaching Time. The
parameter, Weight to Recruiting, is very important because it is the essence of how managers’
time is allocated. It is the only choice realistically available to the manager in determining
how he spends his time between the various responsibilities.
Ina proportional allocation policy, one may allocate x% of the time for recruiting and the
remaining (100-x) % for coaching. This kind of a-priori assignment is reasonable at the
planning stage, but it was seen from our interviews that managers don’t stick to this kind of
allocation. Itis more accurate to describe the time allocation policy as a policy of ‘priorities’.
We cite three extreme cases of different priorities to illustrate its meaning. The first - “equal
priority” - assigns a value of 0.5 for Weight to Recruiting. This does not mean that both tasks
will get equal share of the time but that each will get its fair share. If recruiting requires 20%
of managerial time, that share of time (and not 50%) will be put aside for recruiting. Equal
priority comes into play when compromises have to be made regarding the allocation of time.
If available time falls short by 10%, both recruiting time and coaching time are cut by 10% to
18% and 72% of the original time spent on these two activities.
The second example is full priority to recruiting: Weight to Recruiting has a value of 1. Here
the time allotted for coaching is a residual from the total time available. If the total time
16
available is just sufficient for the recruiting workload or even insufficient, then there is no
time left for coaching. The third example is symmetrical - full priority to coaching: Weight to
Recruiting has a value of 0. Here the allocation is just the reverse. These extreme priorities
capture different orientations about management which are tangible, credible postures that
can be adhered to and communicated about. Equal priority approach suggests equal
importance to both the tasks; management goes in for ‘across the board’ cuts when faced with
a shortage of relevant resources. By contrast, the other two approaches are clearly biased
towards one of the two kinds of activity. It shows the orientation of the balancing capability
when faced with a shortage of relevant resources. The different priorities are not different
policies but are different ways of implementing a balance in the broad time allocation policy.
However, these differences may be responsible for generating differential performance across
firms, under certain pattems of resource interactions.
In the third step, the absolute amount of time to be compromised from Time Required of Ea ch
Activity is used to calculate the fraction of the full amount of time that can actually be
dedicated to the activity, given the time required, the time available and the priority to the
different tasks. The relative amount of time that must be compromised is in the matrix Share
of Time Allowed for Each Activity. We refer to the elements of this matrix as Share of Time
Allowed for Recruiting and Share of Time Allowed for Coaching. This fraction is conveyed to
the Headcount Sector and the Skill Sector, whe re the amount of recruiting and coaching are
scaled down as per the availability and allocation of time, in a multiplicative fashion. When
there is no shortage of time, both the elements of this matrix take the value of one.
Simulation 1: The Dynamics of Preferential Time Allocation
The objective of this simulation is to bring out the consequences of the differences in time
allocation. Initially, when the administrative requirement is absent, there is enough time for
the responsibilities of recruiting and coaching. As mentioned before, the two firms - Alpha
and Beta - both have 100 agents each at the start of the simulation and 300 equivalent years
of experience in sales skills (Agent Sales Skills) implying an initial average value of Skill per
Agent of 3 equivalent years of experience per agent. Standard Agent Skill at Hire, Relative
Skill of Quits and Relative Skill of Promotions are all set to 3 equivalent years of experience,
which is the average level of the skills in the market. (Standard) Agent Quit Ra te is set at 0.20
per year while Agents Promoted is set at 0.025 per year. Standard Manager Skill is set at 4
years of equivalent managerial experience per manager, which is the same as the initial value
17
of the stock Manager Skills. Managers are assumed to be working 250 days a year; it is the
value assigned to Working Days per Year per Manager. The recruiting workload (Days
Needed per Recruit) is set at 4 days per recruit while the coaching workload(Target Days of
Coaching per Agen? is set at 21 working days per agent. These assumed values are within
the range prevalent in the studied firms.
These conditions render the issue of priority irrelevant. Performance of both firms adheres to
equilibrium expectations. In Figure 8, the flat line the chart for Skill per Agent indicate
adherence to steady state. This idyllic situation shows no impact of the heterogeneity that
exists due to a difference in priorities, towards generating differential performance. 5 years
after the simulation commences, we propose a reduction of 20% in the time available to
managers to account for the extra time required to meet regulatory standards. The
consequence of this change is that there is now inadequate time for managers to execute the
entire extent of their other responsibilities, as the time required to verify the regulatory
standards is legally binding. Managers now have to decide to what extent they will
compromise. In this context, the relevant initial heterogeneity gets revealed Whereas
managers in Alpha prioritize recruiting (Weight to Recruiting is 1), managers in Beta
prioritize coaching (Weight to Recruiting is 0). This is the only initial difference between the
two firms in this simulation.
We stress that this heterogeneity is a difference at the operational level, not at the strategic
level Such heterogeneity between the two firms exists due to a difference in priorities, which
in tum results from differences in motivation and its implication for balance. This is
congruent with self- determination theory (Deci & Ryan, 2000) which takes into account three
main aspects in the tasks of recruiting and training: ability development vs. demonstration
(Nicholls, 1984; Dweck, 1986), intangible vs. tangible rewards (Herzberg, 1982; Riedel et al,
1988) and extrinsic vs. intrinsic motivation (Thomas, 2000; Brief & Aldag, 1977). The nature
of the heterogeneity proposed for this experiment has not had any impact on the initial status
of the firms. In fact, senior management may expect that such differences are not capable of
sustaining differential performance, absent any initial competitive advantage. We surmise
that prioritizing coaching over recruiting indicates that one of the firms is going for quality of
manpower while the other is going for quantity, and thus differences may well crop up,
whether sustainable or not. The simulation should inform us about that sustainability.
18
In such circumstances, we need to account for both quantity and quality aspects of perform
ance. Quantity is indicated by the products contributing to the revenue stream (Net Product in
Force) and quality through Skill per Agent Net Product in Force is a stock whose inflow is
the product of three metrics: the number of agents engaged in selling policies (Agents), the
productivity of those agents (Sales P roductivity) and the lapse rate (Lapse Rate ). Figure 8 has
the trajectory of these two measures for a period of 20 years, which includes 15 years after
the reduction in time and 5 years prior to it. Note the very similar profiles for the trajectories
of the two firms in the left hand side. Though Beta initially takes the lead in sales, by the end
of the period under study, Alpha reverses this difference and is set to increase it further. T he
maximum difference of about 5.3% occurs just before 10 years are completed in the changed
environment. These trajectories indicate that perhaps the differences in this experiment can
not create a large impact but this is not so if we examine a few more relevant variables.
The right hand side shows the trajectory of Skill per Agent. Whereas Skill per Agent decreases
from 3 years to 2.35 for Alpha, it increases from 3 years to 5.39 for Beta. Basically, these
trajectories show a sustained divergence in the performance of these two fimms. It is in
complete contrast to the left hand side which shows hardly any degree of difference between
the trajectories. Figure 9 shows the corresponding trajectories of Achieved Growth Rate and
Agents in the same 20-year period. T here is a sharp drop in Achieved Growth Rate only for
Beta. Just past the sixth year, it collapses to less than 4.2% but thereafter it recovers slowly to
exceed 6.5% by the end of the twentieth year. The trajectories of Agents show a divergence,
due to the impact of different growth rates achieved for the firms, compounded over time.
Alpha accumulates to 1636 agents while Beta expands to 475 agents only. After the 5th year,
when the reduction in managerial time takes place, the number of agents added on for Alpha
is about 5.2 times that of Beta. The difference in the trajectories in the figures illustrates the
impact of the different priorities chosen by each firm, under the given conditions of scarcity.
Alpha, where recruiting was the priority, achieved the planned growth target, but at the
expense of their coaching respons ibilities. Due to this negligence, the level of their agents’
skill pool dropped. Beta’s priority to coaching saw the fulfillment of its coaching tasks at the
expense of recruiting. It was able to increase the level of their agents’ skill pool. Let us take a
closer look at the dynamic balance around the stocks called Agents and Agents’ Sales Skills to
investigate the precise reasons for the demonstrated divergence. In both cases, growth occurs
in a dynamic equilibrium for the first five years. Even though there are different priorities,
19
there is no impact in that period because of the lack of pressure on the allocation of the scarce
Tesource, managerial time. When the sharp reduction in time takes place, it has an immediate
impact as it brings into play the difference in the weights between the two firms, in the
parameter Weight to Recruiting which is the important constituent of the policy of time
allocation for management and has a key role in deciding the values of Share of Time
Allowed for Recruiting and Share of Time Allowed for Coaching.
Alpha responds to the shortfall in time by curbing coaching, which decreases the Share of
Time Allowed for Coaching to a value less than 1. The stock Agents’ Sales Skill has as one of
its sources of skill, the amount of leaming put in by agents (Agent Coaching), a product of
Agents and Agent Coaching Rate. However, Agent Coaching is also a product of Share of
Time Allowed for Coaching. When Share of Time Allowed for Coaching drops below 1, the
amount of skill ertering Agents’ Sales Skill drops and its immediate impact is to decrease
Skill per Agent below the value that perpetuates a stable situation. This decrease has three
consequences. The first increases Agent Quit Rate and Agents Promoted , which sum to the
number of agents leaving the sales force. So, more agents have to be hired to keep up with the
planned growth rate. This in tum implies an increase in the recruitment load, and given the
priority to recruiting, it aggravates the Share of Time Allowed for Coaching. The second
consequence is to decrease Agent Skill at Hire due to a drop in compensation, which reduces
Added Skill at Hire and further decreases the inflow of skills into the stock Agents’ Sales Skill
and thus Skill per Agent. The third effect, a a result of increased hiring, is to increase the
coaching load - but due to the priority to hiring, this just results in an even lower value of
Share of Time Allowed for Coaching. Thus, the three consequences reinforce each other to
keep on lowering the values of Share of Time Allowed for Coaching and Skill per Agent
Beta responds to the shortfall by curbing recruiting at the expense of coaching. This increases
the recruiting component in Shortfall of Each Activity and therefore decreases the share of
recruiting in Share of Time Allowed for Each Activity. The result is a decline in the number of
replacements hired and a relative increase in the amount of skills flowing into Agents’ Sales
Skill. The immediate impact is that it increases Skill per Agentabove the value that enables a
stable situation. This has three consequences. The first is to decrease the Agent Quit Rate and
Agents Promoted, which decrease the number of agents leaving the stock. Consequently the
number of agents required to be hired to maintain the planned growth rate decreases. This
implies a decrease in the recruitment load. Given Beta’s coaching priority, it counteracts the
20
decline in the Share of Time Allowed for Recruiting. The second effect is to increase Agent
Skill at Hire, which increases Added Skill at Hire and further increases the inflow of skills
into Agents’ Sales Skill and consequently Skill per Agent The third consequence, as a result
of decreased hiring is to decrease the coaching load - but due to the coaching priority and
higher Skill per Agent, this just results in a further increase in Skill per Agent. Despite the
counter consequence of the first effect, the overall impact is increased Skill per Agent
In both firms growth continues to take place - but not in equilibrium, since the dynamic
balance has been upset in different ways. The compounded impact results in different profiles
of Achieved Growth Rate and Agents. We have succeeded in achieving the objective outlined
at the commencement of this experiment - which was to show that a difference in balancing
time allocations can generate differential performance, even dramatically so.
Simulation 2: Managerial Attitude to Growth
So far, in the model above, we have assumed that managers strive to plan for growth at an
unambiguous, universal target rate. Interviews with industry experts confirm that such targets
are set by senior management in the corporate office. However, given the results above, this
is unlikely, since performance feedback influences goal setting (Latham & Locke, 1991).
After a few years of regularly failing to meet target, only a few managers were likely to
remain determined to reach target next year. These managers were unaffected by past
performance, due to reasons such as conscientiousness (Barricket al, 1993) and belief in self:
efficacy (Wofford et al, 1992; Locke & Latham, 1990). Perhaps inadvertently, others adapted
their aspiration levels (Wright et al, 1995), for reasons such as the attribution of their failure
to factors beyond their control. They would see the target as unrealistic, which does not take
into account the specific context that they have created at the branch level.
Once set, middle management (with direct responsibility for functions like recruiting and
coaching) is evaluated against its ability to achieve the target. Grappling with local issues and
imbalances, managers at this level bring into play an element of compromise / balance - they
implement growth in their own way by varying the actual value of the growth target they
pursue. This change is likely to be intangible, ambiguous or even opaque to senior
management because, when presented with a fait accompli, it would be difficult for them to
determine whether failure to achieve the requisite growth rate was due to external
circumstances beyond one’s control or purely due to timid planning. T hus, differences arising
21
from such balancing may persist, even while broad claims may be made that no such
differences exist.
Figure 10 represents the structure of the growth-rate implementation process. One key to it is
Achieved Growth Rate, a tangible measure of the instantaneous growth rate just achieved. It
is simply the net increase in agents divided by the existing agents. This instantaneous rate is
likely to fluctuate in the short-term. So, the perception of the growth rate is likely to be
anchored in the existing value of Perceived Growth Rate but would also be influenced by the
recently achieved growth (Achieved Growth Rate) in a continuous manner. Thus Perceived
Growth Rate is anaccumulation updated regularly through Change in Perceived Growth
Rate. The magnitude of Change in Perceived Growth Rate is proportional to the difference
between the current perception and the justachieved growth rate (Achieved Growth Rate).
The speed at which this difference adjusts the stock is determined by a time constant (Time
Constant 1), related to how much time it takes for management to effectively absorb the
change coming through. The initial value of Perceived Growth Rate is set to the Sales
Growth Target chosen by senior management.
For a variety of reasons, managers may not be able to reach the set target for growth in a
particular year. In fact, they may consistently under-achieve this target that is set for them.
For example, in our last set of experiments, management tried to achieve the set growth target
but those who prioritize coaching were not able to. In many cases, senior management will
accept the reasons given for non-achievement and not punish significantly for missing the set
target. However, it is the reaction of the managers unable to meet the target that interests us.
Managers who are not idealistic about achieving high standards would change their target
away, though slowly, from the originally value. The actual target rate in use is represented by
De-facto Growth Target. This is the other key to the intangible adjustment process as these
managers allow Perceived Growth Rate to influence D e-facto Growth Target.
Initially, they expect to achieve the set target, but after just a few years, their effective target
is dictated by their past feats rather than their supposed ideal target - e.g. if managers
achieve, say historically 10% growth instead of the planned 15%, they scale down their
expectations to grow at, say, 12% for the forthcoming year. The process of adjusting the
tension between Perceived Growth Rate and De-facto Growth Target is done through an
adjustment factor that depends onthe ratio of the two it is depicted by Adjustment Factor for
22
Change in Target The manner inwhich D e-facto Growth Targetis influenced by this
adjustment factor is similar to the process where Achieved Growth Rate smoothes Perceived
Growth Rate. Following an analogous structure, De-facto Growth Targetis updated regularly
through Change in Targ et. The magnitude of the flow is proportional to the adjustment factor
and a Time Constant 2 that is directly related to how much time it takes for management to
effectively adapt their expectations. Idealistic managers may also underachieve year after
year but, in contrast, still plan for the year ahead based on the desired growth rate , ignoring
their own historic performance. This context means that De-facto Growth Target is initialized
with the value of Sales Growth Target, set by senior management and remains unchanged,
irrespective of recent performance. There is effectively no link between Perceived Growth
Rate and De-facto Growth Target.
Similar to the last experiment, there are two firms whose priorities contrast in recruiting and
coaching. Here, however, the managers of both firms have a realistic attitude to the growth
target. The parameter Weight to Sales Growth Target serves as a switch between the
heterogeneous attitudes towards the implementation of the growth target. It is the difference
between the last simulation and this one - highlighting another aspect in the implementation
of balancing capabilities. Alpha-x and Beta-x have a value of Ofor this parameter which
represents the case with adaptive expectations (realistic attitude) while Alpha and Beta from
the previous simulation had a structure equivalent to the value of 1 - the case with adaptive
expectations (realistic attitude). Time Constant 1 is set to 0.2 year while Time Constant2 is
set to 4 years. Both Perceived Growth Rate and D e-facto Growth Targetare initialized to
Sales Growth Target, which is set at 15% per annum. As in the last test, the heterogeneity
between Alpha-x and Beta-x does not have any impact on the initial status of the firms.
Senior management may be unaware of the difference in context that is proposed in this
simulation. Even if they were aware, they may expect that such differences are not capable of
sustaining differential performance. As before, we surmise that prioritizing coaching over
recruiting indicates that one of the firms is going for quality of manpower while the other is
going for quantity, and thus differences may well crop up, whether sustainable or not.
Figure 11 presents the trajectory of Net Products in Force and Skill per Agentfor a period of
20 years, under conditions similar to the previous experiment. The maximum difference for
Net Product Sales in Force is quite small but Alpha -x is ahead towards the end of the time
frame. The proximity of the two trajectories indicates the subtle nature of the differences
23
between the firms in this experiment, in attitudes and in performance - particularly when
examined from an aggregate level typical of the top management team. The right hand figure
shows that, Skill per Agent falls from 3 years to 2.26 years for Alpha-x while Beta-x shows
an increase in Skill per Agent from 3 to 9.75 years. Like the previous experiment there isa
sustained divergence in this aspect of firm performance.
Figure 12 displays the trajectories of Achieved Growth Rate and Agents. It is similar to the
divergence we saw in Figure 9. Alpha-x grows from 100 to 1744 agents while Beta-x grows
to only 244 agents. Corresponding to the dramatically slower growth for Beta, Beta-x shows
a similar arrest after the fifth year when the reduction in managerial time is introduced. Its
overall growth rate over 20 years is only 4.56% compared to the general rate of 15.37% for
Alpha-x. In fact, in the last 12 years Beta-x grows at 0.017% - it is virtually stagnant. The
similarity in the trajectories of Alpha to Alpha-x (and Beta to Beta-x) indicates that the
analysis and explanation for the divergence observed between A Ipha-x and Beta-x is similar.
Managers in Alpha-x prioritize recruiting when there is a reduction in available managerial
time. The fact that they achieve their growth target (in terms of agents) leads them to aim for
somewhat larger targets. This accounts for the observed growth rate going beyond the official
15%; done at the expense of coaching, there is a drop in Skill per Agent. Conversely, Beta-x
prioritizes coaching to raise Skill per Agent, but at a huge concession to recruiting targets.
DISCUSSION & CONCLUSION
Two broad implications from the above experiments are obvious. First, the non-divergence in
the first five years of Figures 8, 9, 11 and 12 shows that a difference in priorities does not
automatically lead to a performance differential. Rather, it is just a contributor to the
divergence which comes about when the context is appropriate. There are other resources
whose interactions with the implementation of balancing allocations determine the degree of
divergence.
Second, the divergence in the above figures originates from the exogenous regulatory change
as the Alpha types spend a greater-than-appropriate share of their time recruiting. Though the
headcount target is met, neglect of training causes a cumulative weakening of the skill pool.
Alpha’s trajectories are very similar to Alpha-x because its ‘realistic’ managers do not have
anything to compromise about as Alphas meet their growth target anyway. Type Beta spend
more of their time training and accumulate superior productivity with obvious productivity
24
implications, but the headcount targets are not met; the actual growth rates fall well short of
what was intended. Trajectories of Beta-x show a smaller sales force size and actual rate of
growth but larger in productivity compared to Beta as the managers here have room to give in
on the growth target. This explains the wider divergence in the second simulation. These
outcomes provide evidence that the nature of interactions emerging from the balancing act is
a source of differential performance and competitive advantage.
There is negligible separation in Figures 8 and 11 which show an aggregate financial measure
combining productivity and absolute size. For the simulation results here, these two are
negatively correlated; so the critical information about the divergence gets suppressed.
Examining firm performance at such an aggregated level does not help us appreciate the
significant differences that arise due to the heterogeneity in the implementation of balancing.
This is a very interesting outcome because it shows the deviation, from the compromise made
to pursue a personal target, would be ambiguous, if not completely opaque to the top
management team. It would be difficult for them to determine whether failure to achieve the
tequisite growth rate was due to extemal circumstances beyond one’s control or purely due to
timid planning, when presented with a fait accompli. Thus, differences arising from the
heterogeneity in balancing may persist, even while broad claims may be made that no such
differences exist. Considering the performance of the real life firm in question, the masking
of the differences here shows how middle management effectively brings about a significant
deviation from the intended strategic positioning.
Since the mechanisms of the results are very similar in the two simulations, one might
conclude that while the strategy for executing the balance in recruiting vs. coaching makes a
clear difference to the firms’ performance, the strategy for the second kind of balancing
(adjusting the de-facto growth target) appears to have no significant impact. The hypothesis is
that if heterogeneity in this balancing strategy has a negligible impact, then the performance
gap within each firm type should be negligible. Comparing the left hand sides of Figures 9
and 12, the hypothesis seems verified by the trajectories of the achieved growth rate for
Alpha and Alpha -x. However, contrasting Beta and Beta -x reveals a clearly different story.
After the exogenous change, the growth rate of Beta-x continues to decline while that of Beta
starts recovering. The eventual impact of this difference shows up if one compares the Beta
firms in the first set of figures with the second set, e.g. right hand sides of Figures 9 and 12
25
The heterogeneity in balancing between the Beta firms does have an eventual differential
impact on the final size and quality of agents, while the Alphas are busy recruiting an ever
lower quality of agents. The key success factors show no impact until the slack in managerial
time is used up. Further, the heterogeneity in the second kind of balancing has a small impact
among the Alphas as they are committed to meeting recruiting targets, even in circumstances
averse to hiring. The impact on the Betas is larger as they do not prioritize hiring; this permits
greater deviations from meeting already set targets when they think the situation is right for
compromising on the tangible and quantitative aspects of growth. The fact that the same kind
and same degree of heterogeneity has differing impacts with the Alphas and the Betas implies
that the efficacy of a key success factor varies with the context, particularly from a static or
equilibrium point of view. The increasing performance differential among the pairs shows
that it also varies with time. These results suggest the insightthat the efficacy of key success
factors changes with respect to time and context.
A review of the seeming lack of reaction by the top and middle management teams is
justified, as the actual industry events parallel the simulation results. We have used extreme
values to emphasize differences, but the important similarity to the empirical events is that
managers, on eventually understanding the increasing lag in skills or seeing themselves fall
behind in size, did not change their practices. Interviews with senior managers and industry
experts revealed this was due to strong cognitive orientations. A part from taking advantage of
the delayed emergence and intangible nature of their poor performance, managers in the
Alpha firms rationalized away their falling behind through ambiguous interpretations about
the influence and state of the environment on their firms, as they believed coaching was
useless or were strongly motivated to attain the numbers that could be easily verified or
enjoyed recruiting much more than training. In contrast, the managers of the Beta firms,
whether or not they understood the eventual impact of training on the industry, believed in
the efficacy of training to create a better future for their agents and themselves, or enjoyed
training more than recruiting. A similar difference in attitude prevails when one contrasts
Alpha-x with Alpha, or Beta-x with Beta, by their ingrained tendency to compromise on
future targets when faced with repeated failures in their past endeavors.
The cognitive orientations describe above tum out to be significant barriers to flexibility
(Bukszar Jr., 1999), and they lead to strategic consequences (Miller, 2002). Even for the
minority of managers who were alert to the changing situation, it would be difficult to change
26
their behavior dramatically in a short period of time. We account for ate le ast five reasons to
explain such behavior.
First, it is well-known that it is hard to drop dispositional attributes like biases, likes and
dislikes which have been based on self-justifying assumptions and grow new ones overnight.
Second is the potential unfavorable reaction from their colleagues and partners when they
would perceive a change in identity or a deviation from accepted noms. Third, even if such
managers were to successfully change their habits ovemight, it would require significant time
to work through the depleted skill pool and build it up back again. This is because of the
inertial nature of the pool of agents and their skills; while it is easy to change particular
individuals, changing the properties of a group with accumulated skills is a different story.
Fourth, shifting the emphasis from recruiting to training would cause a clear drop in the
growth rates that were being achieved. It is doubtful whether such a drop would be accepted
in a transparent manner by the top management team, except in a crisis. Fifth, the trajectories
of measures for product sales or product portfolio size, which are aggregate balance sheet
measures of performance, reveal little and late to the top management team about the
increasing discrepancy of skills and firm sizes, which are eventually of strategic importance.
This prevented them from introducing different kinds of incentives; it is doubtful whether
such incentives would have made an impact for the better, considering the resistance they
would face about the legitimacy of new kinds of incentives. A part from their legitimacy, it
would take significant time to change behavior or weed out those with undesired attributes.
These factors make it all the more difficult to establish a true balance between actions that
create immediate benefits such as recruiting and those that are vital but need prolonged
investment before delivering results such as training. This significant management challenge
is even steeper when there is considerable uncertainty about the impact of change in the
extemal environment.
Conclusion
Models such as the one presented here are subject to hindsight bias because of retrospective
accounts. A simplified model hinders extensive generalization; yet integrating more theories
into a model produces results that are ever more difficult to interpret. Nevertheless, the model
is open to extension - e.g. one could bring in the impact of the financial performance of the
27
investing arm of the insurance firm through the price of its policy premiums and examine the
consequences on competitive interactions.
The main objective of the paper has been to recognize, introduce and develop the concept of
balancing by middle management as a meta-capability*, Through the simulation of the model
of the insurance firm, we have shown that, in apposite circumstances, the impact of this
balancing meta-capability, in interacting with resources, resource constraints and resource
linkages, is likely to create and perhaps go beyond, to increase differential performance even
amongst firms that limit heterogeneities initially to this meta-capability. We also explained
why these heterogeneities are likely to persist and therefore difficult to eradicate.
With respect to the RBV, we have discovered a first-order capability similar to dynamic
capabilities, although it does not change existing zero-order capabilities (Winter, 2003) or
create new ones. Let us call it dynamic balancing capabilities (DBC). Dynamic capabilities
are assumed to provide an indication of the potential of the firm to be dynamic by changing
capabilities from the evolutionary point of view, also known as a non-ergodic development
feedback effect; DBC complements them by adjusting the firm’s capabilities to avoid
ecological constraints encountered in the external environment and in the intemal resource
structure during the firm’s time evolution, principally by redirecting productive services. This
nor-ergodic dynamic feedback effect tries to ensure that the firm continues to grow smoothly,
thereby avoiding undesirable dynamic effects.
Of course, DBC could be refined to introduce the existence of another first-order capability -
the ability to motivate capacity expansions of the various types of tangible and intangible
resources present in the resource structure of the firm, in a timely manner, so that bottlenecks
and capacity constraints do not hinder the dynamic of the firm. From the theory point of
view, it would be interesting to study the interaction among these three first-order capabilities
so as to get a better understanding of the dynamic properties of resources and capabilities.
References:
Bamey, J.B. (1986a). ‘Organizational culture: Can it be a source of sustained competitive
advantage?’, Academy of Management Review, 11(3), pp. 656-665.
4a meta-capability is a capability that dictates the application of other capabilities. Here the meta-capability of
balancing dictates the allocation of ordinary capabilities like hiring and coaching to managerial time.
®Non-ergodic development and non-ergodic dynamic feedbacks are discussed in Khalil (1997, 1998-9).
28
Bamey, J.B. (1986b). ‘Types of competition and the theory of strategy: Toward an integrative
framework’, Academy of Management Review, 11(4), pp. 791-800.
Bamey, J.B. (1986c). ‘Strategic factor markets: Expectations, luck, and business strategy’,
Management Science, 32(10), pp. 1231-1241.
Bamey, J.B. (1989). ‘Asset stocks and the sustainability of competitive advantage: A comment’,
Management Science, 35(2), pp. 1511-1513.
Bamey, J.B. (1991). ‘Firm resources and sustained competitive advantage’, Journal of Management,
17(1), pp. 99-120,
Barrick, M. R., M.K. Mount, and J.P. Strauss (1993). ‘Conscientiousness and performance of sales
representatives: Test of the mediating effects of goal setting’, Journal of Applied Psychology, 78,
pp. 715-722.
Brief, A. P. and RJ. Aldag (1977). ‘The intrinsic -extrinsic dichotomy: Toward conceptual clarity’,
Academy of Management Review, 2(3), pp. 496-500.
Bukszar, J. (1999). ‘Strategic bias: The impact of cognitive biases on strategy’, Canadian Journal of
Administrative Sciences, 16(2), pp. 105-117.
Deci, E.L. and R.M.Ryan (2000). ‘The what and why of goal pursuits: Human needs and the self-
determination of behavior’, Psychological Inquiry, 11(4), pp. 227-268.
Dierickx I. and K. Cool (1989). ‘Asset stock accumulation and the sustainability of competitive
advantage’, Management Science 35(2), pp. 1504-1511.
Dweck, C. S. (1986). ‘Motivational processes affecting learning.’ American Psychologist, 41, pp.
1040-1048.
Herzhery, F. 1982. The managerial choice: To be efficient and to be hunan. Olympus Publishing, Salt
Lake City, UT.
Khalil, E.L. (1997). ‘Chaos theory versus Heisenberg’s uncertainty: Risk, uncertainty and economic
theory’, American Economist, 41(2), pp. 27-40.
Khalil, E.L. (1998). ‘The Janus hypothesis’, Journal of Post-Keynesian Economics, 21(2), pp. 315-
342.
Khalil, E.L. (1999). ‘Two kinds of order: Thoughts on the theory of the firm’, Journal of Socio-
Economics, 28(2), pp. 157-173.
Latham, G. P., and E.A Locke, (1991). ‘Self-regulation through goalsetting’, Organizational
Behavior and Human Decision Processes, 50, pp. 212-247.
Levinthal, D.A.. and J.G. March (1981). ‘A model of adaptive organizational search’, Journal of
Economic Behavior and Organization, 2, pp. 307-333.
Locke, E. A., and P.G.Latham (1990). ‘Work motivation and satisfaction: Light at the end of the
tunnel’, Psychological Science, 1, pp. 240-246.
Miller, K.D. (2002). ‘Knowledge inventories and managerial myopia’, Strategic Management
Journal, 23(8), pp. 689-706.
Morecroft, J.D.W. (1984). ‘Strategy support models’, Strategic Management Journal, 5(3), pp. 215-
229.
Morecroft, J.D.W. (1985). ‘Rationality in the analysis of behavioral simulation models’, Management
Science, 31(7), pp. 900-916.
Nicholls, J. G. (1984). ‘Achievement motivation: Conceptions of ability, subjective experience, task
choice, and performance.’ Psychological Review, 91, pp. 328-346.
Penrose, E.T. (1959). The Theory of the Growth of the Firm. Oxford University Press Inc., New Y ork,
NY.
Peteraf, M.A. (1993). ‘The comerstones of competitive advantage: A resource-based view’, Strategic
Management] ournal, 14(3), pp. 179- 191.
Porter, M.E. (1980). Competitive Strategy: Techniques for Analyzing Industries and Competitors.
Free Press, New Y ork, NY.
Porter, M.E. (1981). ‘The contribution of industrial organization to strategic management’, Academy
of Management Review, 6(4), pp. 609-620.
Porter, M.E. (1985). Competitive Advantage: Creating and Sustaining Superior Performance. Free
Press, New Y ork, NY.
Porter, M.E. (1996). ‘What is strategy’, Harvard Business Review, 76(6), pp. 61-78.
29
Prahalad, C.K. and G. Hamel (1990). ‘The core competence of the corporation’, Harvard Business
Review, 68(3), pp. 79-91.
Riedel, J.A., D.M. Nebeker and B.L. Cooper (1988). ‘The influence of monetary incentives on goal
choice, goal commitment and task performance’, Organizational Behavior and Human Decision
Processes, 42(2), pp. 155-180.
Sterman, J.D. (2000). Business Dynamics: Systems Thinking and Modelling for a Complex World
Irwin/McGraw Hill, New Y ork, NY.
Teece, DJ., G. Pisano and A. Shuen (1997). ‘Dynamic capabilities and strategic management’,
Strategic Management Journal, 18(7), pp. 509-533.
Thomas, K.W. (2000). ‘Intrinsic motivation and how it works’, Training, 37(10), pp. 130-134.
Wemerfelt, B. (1984). ‘A res ource- based view of the firm’, Strategic Management Journal, 5(2), pp.
171-180.
Winter, S.G. (2003). ‘Understanding dynamic capabilities’, Strategic Management Journal, 24(10),
pp. 991-995.
Wofford, J. C., V.L. Goodwin and S. Premack (1992). ‘Meta-analysis of the antecedents of personal
goal level and of the antecedents and consequences of goal commitment’, Journal of Management,
18, pp. 595-615.
Wright, P.M., J.R. Hollenbeck, S. Wolf and G. McMahan (1995). ‘The effects of varying goal
difficulty operationalizations on goal setting outcomes and processes’, Organizational Behavior
and Human Decision Processes, 61, pp. 28-43.
30
Back to the
FIGURE 1 - An Insurance
. Firm
proxy for performance is “~~ Reject or
agent productivity (sales /
op. exp.) Accept
ACTUARIAL CLAIMS INVESTMENT
Dept. Dept. Dept.
Pricing
of F Investmen
Policies Chala t of Cash
Basket of | [ Portfoli
‘Sale , Premium
Policy sf 0of™. fg > ae
Variations | Policie| NOR
5 | ™
Lapsed.
New | | Policies: _
kinds Operating
of Matured: Expenses
Policie | / Policies |
MARKETING | | AGENC
DEpE \ |Y¥ Dept. |~
ee Model
Boundary
IGURE 2 - Productivity of the 3 group
110000 Z
90000 #
70000 Val pon
30000 T T T
1989 1992 1995 1998 2001 2004
130000
Productivity is measured in UK £ per person per year
FIGURE 3 - Normalized Productivit
150 a
140 A
130 —
120
110 go
100 G 7 7 va
90 o
«0 a
70 T T T
1989 1992 1995 1998 2001 2004
The y-axis has been normalized, base set to 100
FIGURE 4 - Model
‘or perf Boundary
proxy for performance |. ejec or
agent productivity (sales /
op. exp.) Accept
ACTUARIAL CLAIMS INVESTMENT
Dept. Dept. Dept.
Pricing
of ; Investmen
Policies Claims t of Cash
Basket of | [ Portfoli
‘Sale . Premium
Policy sf 0of™. fg > ae
Variations | Policie| NOR
s | ™
Lapsed.
New | | Policies’,
kinds Operating
of Matured: Expenses
Policie | / Policies |
MARKETING | | AGENC
DEpE \ |Y¥ Dept. |~
ee Model
Boundary
FIGURE 5 - Model
Overview
manager
quit rate
growth
target,
MANAGER
MANPOWER
lost skill from
promotion
promotion Agents
agent quits
agents promoted
hires
HEADCOUNT
Agents
skill per agent,
achieved
de-facto growth \
growth rate
target PRODUCTIVITY
ATTITUDE TO
GROWTH TARGET i
product products
sales lapsing
. kill level
impact on a
lapse rate of hired
agents
agent
quit rate
y y
COMPENSATION SECTOR
MANAGERIAL
COMPENSATION
variable agent
compensation
FIGURE 6 - Skill Sector
modified by Agent
Learning
PRODUCTIVITY
HEADCOUNT |______ / SECTOR LS
SECTOR skill per/
agents quits S,
T agent~
\ HEADCOUNT
\
\pires SECTOR
relative skill tet skill
Be of quits from quits — aagants F
Me ze fi
“Ske OK _y __y| Agents’ caren al pepiners’
——- p2dded skill Sales Skill Si trom lost skill from
at hire promotions
ip
agent ug
| — level of learming relative skill of
at hire hired agents promoted agents
4
1
I
if Agent
COMPENSATION |[-_---~ ! ;
SECTOR meee of \Agents Learning
, Rate
compensation
HEADCOUNT
SECTOR
FIGURE 7 - Time Allocation
MANAGER MANPOWER SECTOR
standard fo 4 wince
manager /skill per \ working Gays per
skill of DAREgEr “Managers year per manager
al a
ds ye %
\
e i x P, SKILL
effect of relative relative + _ managerial
+ 9
manager skill ————> efficiency * time available SECTOR
on efficiency of managers
1
1
weight to !
recruiting |
li
; .
1
y share of time
shortfall ——_____» Shortfall of _____“,, allowed for
each activity each activity
+ i
1
!
1
ratio i
i
I
I
totaltime ~ s time required "
needed “4 oneach activity
+ HEADCOUNT
SECTOR
admin JN
target days of
days needed ke \). 2g tes coaching per
per recruit recruiting time coaching time agent in a year
4 A
1
target number |
of recruits | /
1
i)
1
\Agents
| H
[ HEADCOUNT SECTOR |
FIGURE 8 - Differences in
erformance |
Net products sol Skill per agent
108K 6
_
83K 5 =a
Za .
58K 4 7
33K 3
8000 2
() 4 8 12 16 20 0 4 8 12 16 20
Time (year) Time (year)
Units: policies per year (left) and equivalent years of experience (right)
Alpha: Full weight to recruiting
Beta: Full weight to training
FIGURE 9- Divergence in
Units: dimensionless (left) and agents (right)
Size
Graph for achieved growth rate Agents (size of sales force)
0.20 1600
0.15 1200
0.10 800
| 4
—_ “
0.05 cL -— 400 = a —
— —_
0.0 0
0 4 8 12 16 20 4 8 12 16 20
Time (year) Time (year)
FIGURE 10 -
Attitude to
Growth Target
time
constant 1 gt
_cachieved
/~” growth rate
_ +"
&
change in perceived
growth rate
HEADCOUNT
SECTOR
time
a 7 constant 2
-»| perceived weight to sales
: |growth rate growth target
“ | 7
|
Adjustment factor for
+
; change
thange in target in ange eg
6)
- rT oe
de-facto
growth
target
Diemer eeenssienes Sales Growth
Target
FIGURE 11 - Differences in
erformance |
Net products sol Skill per agent
105K 7] 10
e :
85K A 86 7
7 “
ag 7.2
65K 4 7
Se 5.8 ,
45K oe A 7
44 a ¥
25K -*
3.0 bissmnee hw Bond
5 22 eter enes
0 4 8 12 16 20 0 4 8 12 16 20
Time (year) Time (year)
Units: policies per year (left) and equivalent years of experience (right)
Alpha-x: Full weight to recruiting
Beta-x: Full weight to training -- = == = = -- -- -
S i Z e Units: dimensionless (left) and agents (right)
Graph for achieved growth rate Agents (size of sales force)
0.20 1800
i
| 1350 ra
. ?
0.10) : a
| 900 a
0.05 fe Y
e
\“ . a
. 450 0"
~, -
0.00) tel, Po
i goat “ . .
os"
ral
-0.05 0
0 4 8 12 16 20 ) 4 8 12 16 20
Time (year) Time (year)
CC BY-NC-SA 4.0