The Logic of Sizing a Service Organization
By
Merit Smith
Vice President
Director, Health Care Practice
Fall is in the
air, and it's budget time. It’s the season for “management dialogs” about
staffing the service operation. You know the ones we’re talking about:
 | The
executive says: “I’m concerned that given the staff you are projecting we will
not be able to achieve next year’s huge bonus paying financial plan.” |
 | The
hapless service director says: “We’re projecting a 6 percent growth in
transactions. You’ve cut my first round budget increase from 3 percent to 2
percent. What more do you want? As it is, we are going to be arguing about
unplanned overtime in March… Let’s get real.” |
They don’t really say these things, but they do
think them. And lacking logic for determining staff size, they make decisions
about staffing that are the managerial equivalent of a crapshoot. This article
will not solve all your budgeting and staffing problems. But it will give you a
way of thinking through the questions you should ask as you build next year’s
staff budget. Because some readers think in terms of numbers and some get
concepts best via graphics, we’ll give you the logic in words, numbers and
pictures.
How many people do we need to run the service center? The service center’s total staffing should equal its fixed staff plus its
variable (or transaction-related) staff.
The size of the fixed staff is determined by management
decisions about levels and layers, spans of control and how we design jobs for
managers and executives. The fixed staff is the staff needed if we wanted to
have the service center, but it didn’t have any transaction or volume flowing
through it. Increases in volume have little impact on how large the fixed staff
is.
What do we mean by “management decisions about levels and
layers, spans of control and design of jobs”? These management decisions are
really the result of organizational culture and philosophy.
What about the variable or transaction related staff? How
many should we have? Variable jobs are driven by two factors: transaction
volumes and transactional efficiency. The size of the variable staff should be
driven by the number of services they will deliver multiplied by the time needed
for each transaction.
The transaction volume is the number of services we must
deliver. A key measure for volume is “the submission rate of transactions per
source of transactions.” For example, if we get 50 written inquires per 1,000
contracts per month, we have a submission rate of .05/contract month. Smart
organizations know the submission rates for their major types of transactions
and use the submission rates in their budgeting. Submission rates are stable and
predictable over long periods of times but show a lot of seasonality.
Volume is driven by a number of factors: The size and
growth rate of the market we serve, our market share, the demographics and the
type of products we are servicing.
Many organizations overlook differences in the transaction
volumes created by differences in products. For example, in health care, 1,000
HMO members will generate many more phone calls than 1,000 PPO members will. New
customers unfamiliar with your products and processes may generate more
transactions and higher rework rates than customers who have been with you for a
number of years.
Transactional efficiency is measured by the number of
minutes needed to service each transaction. For example we know that it takes
the staff 20 minutes to deal with a written inquiry. Smart organizations know
two things about minutes "per thing”: how long it takes today and how long it
should take. Good record keeping and expense records tell us what is happening
today. Several things influence transactional efficiency. The key influencers
are:
 | Design of Work. There are two meanings to this
phrase. The first is how processes and work activity are designed and
structured. A work process can be designed in ways that fragment and divide
work, creating inefficient hand-offs and other non-value-adding activities.
The second meaning is about how we design our conditions of work. How many
hours of a typical staff member's year are available to provide service? A
typical worker may have only about 1,200 hours per year to deliver service.
Knowing the hours available per year is a key part of staff budgeting. |
 | Staff Efficiency. The hours that are available for
service have to be adjusted for staff efficiency. These adjustments are needed
because people work in different ways and have different levels of
productivity and skill. Fatigue and delay are also a factor. A rough rule of
thumb is that about 85 percent of available time is used in productive
service. |
 | Technology. Economics 101 told us that we can
substitute capital (i.e., technology) for labor (staff). Work processes that
are designed with intense use of technology can use less staff. A less
technology-supported process may require more staff. |
So let’s put all this together in a simple formula. Let’s
use one transaction type. The total staff is equal to the fixed staff and the
variable staff. The variable staff is equal to the transactions multiplied by
the time per transaction, and then divided by available hours per year. Using
about all the high school algebra I can muster is an equation: Y = a + b(x/c).
Y is total staff; a is fixed staff; b is the time per
transaction; x is the number of transactions; c is the number of hours our staff
is available for service.
Here's a diagram that gives you a picture of the logic
we’ve been discussing.

Let’s do a simple example to tie some of these
concepts together. Here are some assumptions and data. The question is "How many
people do we need to support the correspondence function in the service
center?" Here we go.
 | Our
corporate policy is that we will have one supervisor or team leader for every
eight service staff. |
 | Our HR
policy lets a typical staff member work xxxx hours per year. We think they are
highly effective. In fact, we think they are productive 90 percent of the time. |
 | We have
250,000 clients. The submission rate for written correspondence is .045 per
contract per month. |
 | Written
correspondence is a complex transaction. It takes us 25 minutes to service a
written inquiry. |
How many staff members do we need? How many team leaders?
Now that you have the logic, here are nuggets of experience
that might help you.
Fixing processes that are in place before you invest in new
technology. The savings you get from fixing the present processes may be able to
fund the new technology. If it doesn’t, at least it will make it easier to
implement the new technology.
If you have a good idea about the baseline staffing for
your organization, you’re not done. Baseline staffing tells you the
average-sized staff you need to meet your average demand for service. But
staffing to meet average demand will result in substandard and unnecessarily
expensive service.
Why? In the industries Nolan serves, the demand for service
is subject to significant seasonal variation. If you staff based solely on
average demand, in peak periods you will be unable to deliver acceptable
service. Unacceptable service will result in backlogs, increased call volumes,
increased error rates and rework. These symptoms drive increased temporary and
overtime staffing and higher cost of service.
But economic reality is that you can’t staff to your peak
demand. So what do you do? You adjust your baseline staffing to reflect some
part of peak demand, and you use other tactics to balance service supply-demand.
There are a variety of operational and technology approaches to this problem.
Helping clients understand and use these tactics are the “bread and butter” of
Nolan consulting. Using these tactics begins with a clear understanding of your
specific problems of seasonality.
If you aren’t yet using “seasonal adjustment factors” in
planning your staffing, it's likely that your staffing will either be too large
and expensive or will result in seasonally unacceptable levels of service.
Understanding seasonality is important to balancing service demand and service
delivery while achieving optimal cost.
|