What is capacity management?11.2
Operations managers are concerned with ensuring that the service process has sufficient resources
to deal with the anticipated levels of customer demand in such a way that quality of service meets
pre-set targets in the most cost-effective manner. This is a particularly difficult task when managers
are faced with very variable demand, not just in terms of volume but the variety of services
required, as you can see in the New Zealand water taxi firm described in Case Example 11.1.
Capacity management is a delicate balancing act because both underutilised and overstretched
resources can be disadvantageous. Underutilising resources has the potential to
damage the long-term success of the organisation in a variety of ways:
● Expensive resources not earning revenue lead to poor financial results. The airline that fails
to achieve a high load factor on its planes will struggle to survive.
● In many services, customers are suspicious of services that appear not to be busy. Banks
and similar financial institutions find that customers are not happy about using an empty
branch, and many diners prefer the ‘buzz’ of a busy restaurant.
● Service employees may become demotivated if underutilisation persists. Boredom and
concern for their long-term employment may lead to poor service attitudes, which again
lead to reduced customer satisfaction and lower profitability.
Conversely, resources that are overstretched also lead to problems for the success of the
organisation:
● Overloaded resources mean that many aspects of service delivery suffer. A sudden surge
of customers into a shop means that waiting times increase and staff cannot devote the
amount of attention to customers that is desirable.
● Staff who are continually overloaded make more mistakes and, in the longer term, may
decide to leave the organisation in search of less stressful employment.
● To deal with overload, staff may be drafted in to carry out tasks with which they are unfamiliar
or for which they are only partially trained. The potential for increased error rate is
high and, again, stress levels may be intolerable for some members of staff.
The task of capacity management is to try to achieve a balance between too much and too
little resource utilisation, within the financial and operational constraints. Capacity management
is concerned with putting a plan in place that makes the best use of resources to deal
with the forecasted or expected demand for services.
The Marlborough Sounds covers an area of around
4,000 square kilometres at the north end of the South
Island of New Zealand. The Sounds are a series of
fjord-like water inlets connected to the Cook Strait
which separates North and South Islands. The steep,
wooded hills and small quiet bays of the Sounds are
sparsely populated and access is difficult. Many of
the small settlements and isolated houses and holiday
homes are only accessible by boat. The main port
of Picton sits at the foot of Queen Charlotte Sound.
Serving the more isolated Pelorus and Kenepuru
Sounds is the small town of Havelock, famous for its
green-lipped mussels which are farmed in the Sounds.
Case Example 11.1 Pelorus Water Transport
Part 4 Deliver284
11.2.1 Defining service capacity
Before we discuss the methods available to try to balance capacity and demand we first need
to define what capacity is and how it can be measured.
Service capacity is defined as the maximum level of value-added activity over a period of
time that a service process can consistently achieve under normal operating conditions.1
We
can define and measure capacity relatively easily at the process level. For example:
● the number of calls a customer service agent can handle in the course of a shift
● the number of meals served by a restaurant during the lunchtime period
● the number of repair calls made by a computer service engineer during an eight-hour day.
It is important to note the words ‘under normal operating conditions’ and ‘consistently’. It
may be possible, in some cases, for an individual employee to exceed the throughput rate for a
short period. If call centre employees handle 120 calls over 8 hours (15 calls per hour), it may
be possible for them to achieve as many as 30 calls in one of these hours, but for this rate not
to be sustainable over any length of time.
As we will see later, in Section 11.6, overloading resources may appear to increase output.
Indeed, if analysed solely in terms of numbers of customer transactions completed in a given
period, this may seem to be the case. However, there may be an impact on the nature of the
service, the service concept and also the quality of the service provided. Service organisations
must take particular care to ensure that the service concept is not changed in the search for
John Beavon and Catherine Coates operate one of the few water taxi services from Havelock, Pelorus
Water Transport. In what they describe as ‘the best job in the world’, John and Catherine ferry passengers
and their small cargo around the Sounds. Passengers include residents with their shopping, holiday home
occupiers and their belongings, backpackers and their rucksacks, builders working on the houses, contract
workers, often loggers, and their smaller tools, and government environmental and building officers. They
are sometimes involved in medical emergencies evacuating seriously ill people to the road in Havelock. They
are also very happy to fill up a few places with tourists and take them along for the ride, doing their utmost
to extend the trip for them and show them local features such as gannet and seal colonies and mussel farms
or provide guided or unguided walks. They even serve afternoon tea on board!
The work is varied and as with any taxi not particularly predictable and so a degree of flexibility is required
not only by the operators but also by their customers. John explained:
We try to get about 10 to 12 people on every trip, and we go out usually two or three times a day. The
Sounds are very long and despite having a fast boat (a 24-foot custom-built 25-seater powered by twin
200-horsepower engines capable of cruising at 24 mph, 36 knots, which is pretty fast for water transport)
it can take over an hour to reach the furthest point on the Pelorus Sound. Usually the first person to ring
up and book the boat on a particular day has the say on what time we depart and where we go to. Then,
as others ring or email, we will fit them in around that. We may also have to start making adjustments to
the schedule to allow us time to pick the others up and load and unload their equipment. The vast majority
are happy when we ring them up and ask them to adjust the times; I guess they realise it might be them
we are trying to fit in next time. Obviously some timings are critical and can’t change such as when we are
bringing people back to Havelock to meet up with other transport.
The helpfulness of customers spills over to create a sense of camaraderie on board. The atmosphere feels
more like a convivial bar than a taxi, with John and Catherine on first-name terms with many of their clients.
For exclusive use of the boat for a trip John and Catherine charge around NZ$274 per engine hour. If
people are willing to share, as most are, they would aim to charge people 40, 60 or 75 dollars per head
depending on distance travelled.
There is one other water taxi based in Havelock, and several other companies offering tours, such as the
Mail Boat which covers three set routes three days a week. Access to the Sounds can also be by float plane,
based in Picton and travelling between Wellington (in the North Island) and the Marlborough Sounds.
Chapter 11 Managing service resources 285
greater productivity. For example, a restaurant may decide to encourage customers to leave
their table when they have completed their meal, in order to fit in a second sitting for dinner.
On the face of it, the restaurant has doubled its capacity, but customers may feel that the
nature of the service has changed and the level of service has deteriorated. Of course there are
strategies that the restaurant can adopt to manage this sensitively, but the operations manager
must be certain that in increasing productivity the desired service concept is maintained.
11.2.2 Measuring capacity
To manage capacity, we must be able to measure it. A simple measure is the amount of demand
in a specified time period. The parcel delivery service may have an overall measure of
capacity in terms of the total number of parcels that can be processed overnight; however, this
overall measure is not very helpful in managing the day-to-day operation. It is necessary to
develop a measure of capacity that is sufficiently detailed to give a ‘good enough’ estimate of
capacity. For the parcel delivery company, some of the key aspects to be considered include:
● The size, weight and value of the parcels to be moved – a package that is small but valuable
will provide more revenue per truck movement than something less valuable that takes up
a large amount of space.
● The geographical locations served by the company – rural districts have greater travel
times, and many inner city areas suffer from traffic congestion.
In determining capacity,a number of factors make the assessment of service capacity difficult:
● Service mix. If the service mix (the range of services provided) is made up of high volumes
of ‘runners’ (see Chapter 8), the capacity calculation is relatively straightforward. However,
once the service mix incorporates fluctuating volumes of ‘repeaters’ and ‘strangers’, the
calculation becomes more complex. The customer service agent may be able to handle
120 ‘normal’ calls in a shift; however, if some of the calls are complex enquiries or serious
complaints, for example, the number of calls handled will drop significantly.
● The impact of location. At first sight, the measure of capacity for a computer service engineer
or a telephone engineer would seem to be relatively straightforward. However, if we
consider the difference between an engineer operating in a major city and another engineer
dealing with rural communities, it can be seen that calculating capacity on the basis of
calls completed alone would be inaccurate, since it would take no account of travel times.
● The extent of intangibility in the service. Services with low degrees of intangibility are
relatively easy to deal with. The number of short transactions per hour in a fast-food
restaurant is relatively consistent. However, the customer-facing staff in a gourmet restaurant
may have greater discretion as to how to carry out their task. They may perhaps
spend time with customers in ‘building relationships’, with the result that the individual’s
capacity becomes more difficult to define. This calculation becomes more complex when
dealing with knowledge workers, who must combine short-term revenue-generating activities
with long-term research and development. Capacity in this case is linked more to
the individual service provider. It may be almost impossible for a manager to know when
someone is working to capacity when output is so variable.
● The ease of identification of resource constraints. The capacity of a process is determined by
resource constraints or bottlenecks. In the Karolinska Hospital example (see Case Example
11.3), it was clear that the resource constraint was the operating theatre. Finding ways
to increase the effective utilisation of this space was relatively straightforward. For more
complex systems, the identification of the key resource constraints may be rather harder.
An information systems provider may require a wide range of technical skills relating to
different applications and programming languages, but the precise requirements may not
be known until part-way into the contract.
Part 4 Deliver286
Major investments or disinvestments are required to deal with longer-term forecasted or
anticipated fluctuations in demand. Alton Towers Resort (see Case Example 11.2) made
significant investments every few years to install new or replacement rides to provide the
capacity to deal with the increasing numbers of guests it wanted to attract to its theme park.
It also invested in hotels and a water park to allow it to enter the year-round short break
market. One key issue with such large-scale investments is that this capacity usually comes
in large lumps, like a 180-room hotel, and so that capacity may not always be fully utilised
when it is first installed/opened. Later in its life that capacity might become stretched and
bookings might be turned away. As a result most operations managers spend a great deal of
their time dealing with more short- to medium-term capacity management, trying to balance
the day-to-day demand for their services with the available capacity. This section begins
with a discussion of the longer-term capacity issues then focuses on short- to medium-term
capacity management.
How can managers balance capacity and demand?11.3
Alton Towers Resort is Britain’s answer
to Disneyland. It is the UK’s leading short
break resort with a wide range of family
entertainment. The resort, a division of
Merlin Entertainments Group Limited, is
perhaps best known for its Theme Park’s
white knuckle rides such as Thirteen,
Air, Nemesis, Oblivion, Rita and Ripsaw.
Oblivion, for example, is the world’s first
vertical drop roller coaster. The ride lasts
160 seconds and reaches speeds of up
to 100 kph whilst pulling a G-force of 4.5
(astronauts only experience 3G at takeoff!).
Rita accelerates to 100 kph (62 mph)
in 2.2 seconds. The Thirteen ride, opened
in 2010, is billed as the scariest in the UK,
combining physical and psychological
fear. Some of the other less scary rides include Congo River Rapids, the Runaway Mine Train and Haunted
Hollow. Opened in 2009 was Sharkbait Reef by Sea Life. This includes ‘touch pools’, where guests can
interact with various underwater species, and a 10-metre ocean tunnel. In April 2010, a live webcam was
installed to allow internet users to watch one of the tanks via the Alton Towers website. Younger guests can
enjoy a range of special toddler and child play areas, and those who don’t care for the rides or play areas
can enjoy the gardens and floral displays, shops, restaurants and daily live shows. To help cope with the
British weather about a third of all the rides are either indoors or in covered areas. The range of eating places
includes McDonald’s, KFC and Pizza Hut as well as Alton Towers’ own branded family restaurants. Together
they produce the 260 tonnes of chips and four million cans of drinks that the guests consume each year.
While the Theme Park is open between March and November, the two hotels, Alton Towers Hotel and the
Splash Landings Hotel with its water park, are open all year round.
There is parking for over 6,000 cars and 250 coaches at the Theme Park, which entertains around three
million visitors a year. The cost of entry is currently around £30 for adults and £25 for children. Demand peaks
Case Example 11.2 Alton Towers Resort
Chapter 11 Managing service resources 287
11.3.1 Long-term capacity management
There are four main considerations for making long-term capacity decisions: location, capacity,
capability and resilience.
Location decisions
Location is the geographic positioning of a facility or facilities which is providing capacity.
Location decisions are often expensive and may have a significant impact not only as an investment
cost but also on operations costs, since location may be affected by local wage rates
and business rates, for example. Location may also have an impact on revenues, particularly
when the operation involves physical contact with customers. For operations that do not
require direct physical contact with customers, for example call centres and internet-based
service providers such as health or benefits advisory services, location decisions can be made
to minimise the physical costs of the buildings and the running costs of the operation. For
operations that need direct access to customers, expensive town-centre or out-of-town shopping
malls may be essential.
Location decisions are a balancing act between supply-side factors and demand-side fac-
tors.2
Supply-side factors are those that influence the costs and difficulties of a location decision.
The demand-side factors are those that influence revenues. Not all the factors below will
apply to every location decision but they are an indication of those factors that may need to
be taken into account.
Supply-side factors include:
● land costs – the costs of acquiring the land
● labour costs – wage costs, employment taxes, welfare provisions etc.
● energy costs – the cost of energy or the availability or even the consistency of the supply of
energy
● transportation costs – the costs of getting resources to the site and of transporting materials
to customers
● government factors – local taxes, capital restrictions, financial assistance and political
climate, and planning restrictions
● social factors – language and local amenities
● working environment – the history of labour relations and labour supply.
at about 50,000 visitors on Easter Bank Holiday Monday and usually runs at about 35,000 throughout the
summer. The busiest times are usually during the week. Fridays and Saturdays during the peak season tend
to be relatively quiet. The various activities in the Park reach peak demand at different times. The peak time
at the gate is 10.30–11.00; for the restaurants it is 12.30–1.00. The major rides are very busy all day with
queues reaching their longest in the early afternoon. Fast-track tickets can be obtained for the most popular
rides, offering guests an allocated time slot. These tickets can be purchased in advance. The Theme Park
employs around 350 full-time and 1,200 seasonal staff. The majority of staff live within a 20-mile radius and
to help cope with unexpected fluctuations in demand a pool of staff is available for work at short notice. For
flexibility most of the operators are trained to operate several rides.
Each year the company invests in new rides, attractions and infrastructure. For example, in 1994 £12 million
was invested in Nemesis. In 1996 the 180-room Alton Towers Hotel was built at a cost of £20 million to
allow the company to position itself as a major short break destination. More than 100,000 guests stayed in
the hotel in its first year. In 1998 £12 million was invested in Oblivion. Air was built in 2002 at a cost of
£12 million. Rita was built in 2005 for £8 million. Thirteen was added in 2010 at a cost of £15 million.
Part 4 Deliver288
Demand-side factors include:
● convenience to customers – the site’s accessibility for customers, including transport network,
parking, distance from markets
● labour skills – the availability of particular talents, skills, accents and cultures
● characteristics of the site – the intrinsic and maybe aesthetic appeal of the site
● image – the reputation of the surrounding area and the extent to which there are complementary
services in the vicinity.
Capacity decisions
Another key question is: how big should the facility be? For the package distribution operation,
some estimate of volumes to be sorted in a relatively short time window each night will
be required. Likewise, when deciding the size of a supermarket, call centre, airport, surgery or
cinema, the costs need to be weighed against forecast demand – not only short-term demand,
but also long-term demand, because the cost of changing facility size can be expensive and
sometimes difficult.
The two interrelated issues for operations managers are:
● Facilities can usually only be added in large – and expensive – chunks.
● Capacity needs to match demand.
Adding new facilities usually requires the organisation to commit significant amounts of
capital. This can be a risky business because long-term demand can rarely be predicted with
any great certainty. If necessary break-even volumes are not met, the facility will not pay for
itself. In some cases, such as a theme park where having sufficient customers creates the atmosphere,
the service may not be as good as it should be if volume targets are not achieved. If
volumes are exceeded, there may be significant localised problems for customers, resulting in
customer dissatisfaction and lost business. Given that the majority of forecasts will be wrong
(because they are only forecasts), operations managers will invariably suffer from the consequences
of over- or under-capacity.
Many airports have suffered from the latter problem; furthermore, owing to the length of
time it takes to design a new runway or terminal building, and to go through the planning
process and build the facility, volumes may again exceed capacity as soon as the new facility
is opened.
As with short- and medium-term capacity management (see Section 11.3.2), there are
three main strategies for long-term capacity planning:
● Plan to exceed demand forecasts. This strategy is appropriate where there is an expanding
market or the cost of building a new facility is inexpensive compared to the cost of, or
problems that would be created by, running out, such as electricity or water supply, or air
traffic control facilities.
● Build to forecast. This approach would balance the likelihood of not having enough and
having too much capacity and is appropriate where the costs and consequences of exceeding
demand are similar to those for not meeting demand.
● Plan not to meet forecast demand. This is an appropriate strategy where it is acceptable not
to meet demand or where the cost of capital is very high compared to the costs and consequences
of not meeting demand. Football clubs may be able to do this, using price premiums
and revenues from television companies to balance the books and even set money
aside for future expansion. The problem for some organisations that follow this strategy,
such as supermarkets, is that it might give the competition time and income to pursue an
aggressive expansionist strategy.
Chapter 11 Managing service resources 289
Developing a facility strategy involves steps that are easy to describe but difficult to
implement:
● establish a measure of capacity
● develop demand forecasts, ideally several forecasts including optimistic and pessimistic
ones, identifying the assumptions on which each is based
● identify alternative means of dealing with the forecasts
● undertake an assessment of the risk involved
● evaluate the alternatives.
Capability
It seems obvious that any new facility should be capable of doing what is required, but this
is not as easy as it sounds. There are some airports whose runways are too short to accommodate
some of the larger aircraft. A decision taken years ago on the length of a runway
when planes were smaller creates constraints on operations now. Doctors’ surgeries, too, have
changed significantly over the last few years, as doctors form larger practices to share growing
administrative costs and ease the burden of 24-hour cover. Surgeries also provide many
more facilities than previously, such as well-person clinics and routine surgical operations, for
example, putting stresses on facilities designed for a different way of working.
A key problem we face is in forecasting both demand and also the nature of that demand
and thus the nature of the services that have to be provided in the future. It is little wonder that
many operations management problems stem from the size and nature of the facilities available.
Resilience or flexibility
Although forecasting the size and nature of demand and future services is difficult, if not
impossible, the only thing an operations manager can do, apart from keeping their finger on
the industry’s pulse, is to try to ensure that their facilities have some degree of resilience or
flexibility.
Physical resilience can be created through either structural flexibility or developing the
potential of the infrastructure. Building flexibility or resilience into a facility can be done in
many ways:
● buying extra land to facilitate any possible future expansion
● having a flexible internal structure, with open-plan offices and movable walls
● using flexible equipment, such as cordless telephones or desk-sharing schemes
● adopting different methods of working, such as using more home-based workers
● developing contingencies – railway companies, for example, may plan to use different
routes if one route fails.
11.3.2 Short- to medium-term capacity strategies
There are three basic short/medium-term capacity strategies, although, as we will discuss,
many organisations employ a mixture of all three (see Case Examples 11.1 and 11.2). These
strategies are:
● Level capacity. In this case scarce or expensive resources are maintained at a constant level,
and the organisation must manage the consequential issues for customer satisfaction and
operational service quality.
● Chase capacity. The service organisation attempts to match supply to demand as much as
possible by building flexibility into the operation. The prime objective is to provide high
levels of service availability or fast response, in the most efficient manner.
Part 4 Deliver290
● Demand management. Rather than change the capacity of the service operation, the
organisation influences the demand profile to ‘smooth’ the load on the resources.
Level capacity strategy
The prime objective of this strategy is to maximise utilisation of expensive fixed resources.
An airline seeks to fly planes that are as full as possible with passengers paying the highest
fares. The key operational measure is the ‘load factor’, with the airline knowing that if it is
exceeding a certain figure (about 80 per cent for an international airline), it will be making
profit.
To achieve this level of utilisation, the service organisation may have to make a number of
trade-offs, most notably around customer perceived quality of service. Figure 11.1 illustrates
the situation in a hospital clinic. Here, the task is to make the most of the medical consultant.
The clinic has to solve the problem of always having enough patients for the consultant to
work on, with the added difficulty of there being a high percentage of ‘no-shows’. The clinic
has chosen to overbook appointments, and believes that it is better to upset a few patients
rather than lose valuable consultant time.
To deal with the no-shows problem, the clinic has made four appointments at the start
of each 15-minute period, estimating that one in four patients do not arrive and that the
consultant will require 5 minutes per patient. If all goes to plan, the first patient will be
seen immediately, the second within 5 minutes and the third within 10 minutes, but it
should be noted that they each have the same 2.00 p.m. appointment. In practice, some of
the 2.00 p.m. appointments will still be waiting when the next ‘batch’ arrives for a (supposed)
2.15 appointment.
Some general principles and issues can be drawn from this example about the level capacity
strategy:
● Resource utilisation goals are frequently achieved at the expense of customer satisfaction.
● Customers may receive inconsistent service levels (those with 2.00 p.m. appointments fare
better than those with later appointments).
● Customers (patients) accept (or suffer) this poor level of service because the service is valuable
to them and there may be no or few alternatives.
● There is a danger that the service provider may become complacent and not make attempts
to cut the emotional cost of waiting for the customer, making it potentially vulnerable to
competition (in this case private healthcare).
To overcome this problem of variable service levels, the service organisation may use yield
management (see Section 11.7) or queue management approaches.
0
1
2
3
4
5
6
Patients/
hour
Number
in queue
16
12
Average
capacity
Patient
appointments
Number
in queue
2.00 2.15 2.30 2.45 3.00 3.15 3.30
Scheduled
end of clinic
Figure 11.1 Level capacity strategy in an outpatients clinic
Chapter 11 Managing service resources 291
Examples of organisations that use the level capacity strategy as their dominant approach
include:
● Airlines, which need to maximise the revenue from their most expensive resource
(planes). The prime objective is to have planes flying as frequently as possible, preferably
full of passengers. This may mean that passengers do not always receive the service they
anticipated.
● Professional services, which may have a recognised expert in a specialised field. It is frequently
the case that the overall workload will not sustain another professional, leaving
clients with a choice as to whether they wait or find an alternative provider.
● Popular restaurants, which may intentionally not expand capacity in order to maintain
exclusivity. Having to book days, sometimes months, ahead in order to ensure a table may
enhance the service concept.
Examples of approaches adopted under the level capacity strategy include:
● Promoting off-peak demand. This is often combined with a pricing strategy to encourage
customers to switch. The organisation must be careful that this does not bring about a
change in service concept. A restaurant encouraging customers to move to less popular
times may institute a ‘happy hour’ with cheap drinks, which may damage the restaurant’s
reputation with existing customers.
● Queue management. This is dealt with in more depth in Section 11.5, but it is important
to point out here that making an assumption that customers will continue to queue can
be dangerous. It is, after all, sending the message that their time is relatively worthless, and
they are only prepared to wait because they anticipate that the service they receive will be
valuable enough to make the wait (lost time) worthwhile.
● Booking systems. Making forward bookings is a form of queue management. It allows
the organisation to schedule capacity ahead, and for customers to utilise queuing time
for themselves. Supermarkets have successfully utilised this system for their delicatessen
counters, issuing customers with numbered tickets to ensure that people are served in
order, and allowing customers to judge whether they have time to continue shopping before
their number is called. As with the physical queue, customers may not want to wait
and so may go elsewhere. Indeed, if the organisation has the reputation that customers
need to book ahead, it may lose potential sales if customers assume that there is no point
in trying.
Chase capacity strategy
This strategy is usually adopted by high-volume consumer services, since a major aspect of
their competitive strategy is the provision of ready and rapid access to service. For these services,
capital resource utilisation is rarely a prime goal, although cost reduction will be very
important. To explain this further, consider the following statements concerning a fast-food
restaurant:
● A key objective is to maintain short queue lengths. This is managed by staffing tills and
kitchen in line with expected demand.
● If the queues are too long, customers go to another fast-food outlet.
● The premises are not fully utilised: there are only about six hours out of the possible
24 hours when the facilities are 100 per cent utilised.
The challenge of these high-volume standard services is to develop volume flexibility (see
Section 11.7.2). In other words, the operation must be able to cope with wide ranges of customer
demand, providing consistent service standards at minimum cost. Figure 11.2 shows
the demand pattern for a fast-food restaurant, with a crew roster to show how the restaurant
manager schedules the staff to deal with the variation in demand.
Part 4 Deliver292
In this restaurant, the staff are organised into three categories: those on one of the core
shift teams, those working split shifts, and those working part-time. The split shifts allow the
manager to schedule staff for the forecast demand peaks. There will probably be a weekly
rotation between the core shifts and the split-shift personnel. In this example, the early morning
peak load is covered by employing part-time labour. In addition, the manager will have a
pool of labour to be contacted at short notice to cover absenteeism or an unexpected rise in
demand. A common strategy is to extend the length of the split shifts, with some organisations
operating a ‘compulsory overtime’ policy as part of their conditions of employment.
General principles and issues for the chase capacity strategy are:
● Most organisations operating the chase strategy must develop a high degree of volume
flexibility. In other words they must be able to respond to changing demand profiles. In
most cases this is achieved through employing staff on flexible contracts, allowing the
operations manager to decide working hours as required.
● Although a principal objective is to ensure that customer service targets such as availability
or response times are achieved, many of these service organisations fall into the
commodity category (high volume/low process variety). In contrast to the organisations
employing the level capacity strategy, they frequently have relatively little means of differentiation
and are therefore rather price sensitive. The challenge in adopting the chase
capacity strategy, therefore, is to ensure that costs are strictly controlled and that flexibility
is not achieved at any price.
Examples of organisations that employ the chase capacity strategy include:
● Retailers that need to deal with extremely high demand at weekends and after normal
office hours.
● Direct insurance companies operating extended hours through call centres.
● Theme parks, which may open up more attractions as demand grows.
Typical approaches to the chase capacity strategy include:
● Flexible staffing levels. Some organisations use flexible employment contracts, allowing
the operations manager to decide when staff will be working. In some cases staff will work
a standard core time, but in many retail organisations staff may not know when they will
be working beyond the next few shifts. Another approach is to employ part-time staff who
must work ‘compulsory’ overtime as and when needed. Although this gives flexibility, the
operations manager must be aware of the possibility of staff resentment at having to work
inconvenient hours, and the knock-on effect of poor customer service.
Figure 11.2 Chase capacity strategy for a fast-food restaurant
6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00
Part-
time
Split
shift
Split
shift
Daily demand profile
Number of
customers
Staff scheduling
Core shift team 1 Core shift team 2
Chapter 11 Managing service resources 293
● The use of subcontractors or temporary staff. Organisations may use temporary staff or subcontractors
to deal with short-term overloads. Although these workers may be readily available,
they may not be sufficiently trained or motivated to deliver service in the style of the
organisation. However, some organisations report that their temporary staff may be more
responsive and less complacent than long-service staff. Some call centres use organisations
that specialise in what is called‘peak lopping’. Excess calls are automatically routed to the organisation
and they are answered in such a manner that customers are unaware of the switch.
● Making use of customers. Many service operations may have the option of changing the
service process to utilise customers as temporary employees. In effect this is again changing
the service concept. Some regular customers may be very happy to be included in the
service process – clearing tables, or even serving other customers as well as themselves.
Demand management strategy
Most organisations operate a mixed approach to capacity management. Whether adopting
a principally chase or level capacity strategy, most service organisations also operate some
degree of demand management. Examples of this approach include:
● Pricing strategies. This typically takes the form of offering price incentives to encourage
customers to move to off-peak times. The ‘happy hour’ in the pub or wine bar is a good
example.
● Restricted service at peak times. The philosophy here is similar, though taking the form of
a disincentive. In this case the organisation may provide a limited service at peak times,
again encouraging customers to move to less busy times. Some restaurants operate this
policy, providing a limited menu at these times.
● Specialist service channels. Rather than provide a general service at all times, the provider
may choose to segment the demand and to allocate specific times for special needs.Doctors’
surgeries are a good example, with advertised times for services such as immunisations,
mother and baby clinics, and counselling provision. This allows the surgery to schedule
specialist resources to restricted times, often making better use of scarce resources.
● Advertising and promotion. Increasing public awareness of the service and informing customers
of special offers will stimulate demand. Bookshops not only advertise, but will also
stimulate demand by arranging sessions for authors to autograph their works. A particular
problem with advertising is that it tends to increase the inaccuracy of any forecasting
model used by the business. Although it is possible to track the effectiveness of advertising
in stimulating demand, it is often difficult to pre-judge the likely impact of a new campaign.
Putting the strategies together
Most complex service organisations use all three of the capacity strategies in different parts
of their operations, depending on the respective underlying cost models. Some examples are
shown in Table 11.1.
The prime objective of the airline is to maximise ‘load factor’ on flights, the utilisation of
its most expensive assets. It employs a number of strategies to ensure that it makes the maximum
revenue on each flight, using sophisticated yield management techniques (see Section
11.7.1) to help adopt the optimum pricing strategy to sell unsold seats as departure time
approaches. The airline may simply oversell seats in the belief that there will be a number of
‘no-shows’. Passengers who are then not able to obtain the seat they thought they had booked
need to be compensated in some way, although unless managed well that can significantly
affect customer satisfaction.
To maximise the opportunity for customers to book seats, the airline employs a chase
strategy in its sales department using relatively cheap resource (as compared with planes),
scheduling staff to meet forecast demand patterns. It is better to suffer slightly reduced productivity
here rather than lose potential seat revenue.
Part 4 Deliver294
The insurance company uses a level strategy for its actuarial staff (back office), in part because
they are relatively expensive, but more because they are often in short supply. The lead
time to recruit and train an actuary is measured in years rather than weeks and therefore it
makes no sense to attempt to chase demand.
Similarly the restaurant chain will operate a level strategy in its manufacturing function
because it has relatively fixed capacity: although it can increase capacity marginally by overtime,
significant increases can only be achieved by investing in another kitchen or restaurant.
Table 11.1 Capacity strategies
Level capacity strategy Chase capacity strategy Demand management strategy
International
airline
Ensure that planes are flying
with maximum payload as
frequently as possible
Schedule staff reservations
department to meet demand to
ensure bookings can be made
Promote off-peak demand
Try to maximise revenue from each
flight (yield management)
Insurance
company
Protect back-office experts
(actuaries and investment
specialists) from variations
in customer demand
Schedule direct sales
operation (call centre) to
provide maximum access
for customers
Influence selling cycle so as not to
coincide with policy renewal peaks
Restaurant
chain
Keep manufacturing of basic
food materials as close to
‘level’ as possible
Maintain high utilisation of
process plant
Draw up staff rosters to reflect
anticipated demand
Use part-time staff to manage
peaks
Call in staff for demand surges
Use promotional activity to
stimulate demand in quiet periods
Devise special offers to allow for
bulk-purchasing discounts
How is day-to-day planning and control carried out?11.4
In the previous section we discussed how operation managers decide a capacity strategy,or mix of
strategies,thus creating a capacity plan.This sets the broad parameters within which this capacity
may be allocated to specific customers and/or tasks in order to meet customer service and productivity
targets. This section looks at the mechanisms that may be deployed to micro-manage
this plan as effectively as possible. Day-to-day operations planning is concerned with creating a
‘schedule’ or timetable (often a daily or weekly schedule), based on the capacity plan, which
● allocates staff,customers,equipment and/or facilities to activities (often referred to as loading)
● decides what order things will be done in, e.g. which customers/orders to deal with first
(sequencing)
● shows what time each activity will start and finish (scheduling).
Examples of ‘schedules’ include:
● A table plan for a restaurant – as bookings are received, tables are allocated, giving an
instant picture of the loading for any given time period, and the likely start and finishing
times. This implies that bookings will be accepted on a first come, first served basis and that
customers without reservations may have to be turned away.
● A school timetable that shows where every student, and member of staff, should be at any
time of the day, what they should be doing and when things take place.
● An appointment book at a car servicing workshop, providing space for a given number of
standard services and more complex jobs, showing who will do which job and roughly how
long they are expected to take.
Operations control is concerned with making adjustments to cope with changes as they happen.
Chapter 11 Managing service resources 295
11.4.1 Creating a schedule
Most operations have rules or policies regarding the allocation of capacity. Sometimes these
rules are relatively informal, developed over time in such a way that most customers are satisfied.
Other operations, usually those with more volume and/or complexity, tend to have more
formal allocation systems. For example, most leisure centres have one large space or sports
hall, which can be used for a number of activities. It might be configured for two five-a-side
football pitches, four badminton courts or one tennis court. To create the schedule or timetable,
the leisure centre manager must carry out the following tasks:
● Decide the proportion of time that the sports hall will be configured for each of the activities.
● Ascertain the optimal schedule for these activities based on customer preference. For example,
badminton clubs may prefer Thursday evenings while five-a-side football may be
more popular on Fridays.
● Check the schedule for ease of transition between activities.It may be relatively easy to move
from tennis to circuit training because there are relatively few changes to make, but changing
from badminton to football might require more staff to move and set up equipment.
● Create a booking schedule for the various slots to allocate capacity (times) to specific
customers.
This schedule is required because the leisure centre does not have infinite resources.It is clearly
impossible to satisfy all possible customer demands and remain a viable service organisation.
The way that operations deal with this issue is to create sequencing rules in order to manage
the prioritisation of allocation. Here are six examples of sequencing rules:
● First in, first out (FIFO) or first come, first served. This is the approach used by many
consumer services. The leisure centre may allow customers to book up to two weeks
ahead for badminton for the Thursday evening time slots of 6.00–7.00 p.m., 7.00–8.00 p.m.
and 8.00–9.00 p.m. and will take requests in order until all the courts are full. This scheduling
rule is simple and has the advantage of being perceived to be fair to all potential
customers.
● Last in, first out (LIFO). There are rare occasions when this rule makes sense. The sequence
of loading a delivery truck would follow LIFO scheduling so that the first call would be to
deliver the products or materials closest to the tailgate of the lorry.
● Most valuable customer first. The leisure centre may allow the local football club to have
early access to certain parts of the forward schedule and to ‘block book’ capacity for training.
If the club needs to have the use of the hall on every Friday for training, this represents
steady income for the leisure centre, though it may be a source of annoyance for other,
smaller organisations that are excluded from using the hall for single occasions.
● Most critical first. Emergency services adopt variants of this rule, grading the nature of
each demand between critical and non-essential. Clearly, life-threatening situations call for
immediate action and these demands tend to override all other activities.
● Least work content first. In situations where demand far outstrips supply, this rule allows
more customers to be satisfied quickly. Airline check-in desks sometimes operate a version
of this rule, providing faster service for those passengers with hand baggage only. This
means that the total queue is reduced and fewer customers wait. The problem with this
rule is that some customers wait far longer than they would under FIFO and are potentially
more dissatisfied.
● Most work content first. With activities that have long process lead times it might seem
prudent to schedule the tasks with the longest time requirement first. This may not always
be helpful because, once started, these tasks may become lost in among other ‘work-inprogress’
and not progress as fast as they could. Using this rule alone does not usually
produce the best performance against customer requirement dates.
Part 4 Deliver296
Different rules may be applied in different circumstances.Where customers arrive together,they
generally expect to be treated equally and therefore FIFO will tend to be applied. However, this
is not always the case. At the Alton Towers theme park in the UK, it is possible to buy fast-track
tickets that enable customers to bypass the queues of customers who pay standard prices. This
obvious privilege can cause high degrees of ill feeling among‘standard’ customers.
Back-office processes frequently deal with longer lead times than front-office processes, and
may apply a number of scheduling rules. In the more complex situations it is possible to use a
simple algorithm similar to that used in manufacturing shop floor control, called critical factor
calculations. In this approach a comparison of estimated work content against requirement
date allows for a calculation of an urgency factor. Each stage in the process is provided with a
list of the most urgent jobs, enabling prioritisation. A spin-off from these calculations is a spot
check of the performance of the process. If all tasks fall into the ‘most urgent’ category, clearly
significant action is required to avoid major customer dissatisfaction.
11.4.2 Operational control
Systems for control, i.e. to enable the schedule to be adjusted, range from comprehensive,
complex and expensive to extremely simple.
At the complex and expensive end of the spectrum are the ERP (enterprise resource planning)
systems. These software-based systems, sold by companies such as SAP and Baan, offer
the capability to integrate a number of functions across the organisation. For example, sales
order processing systems can provide direct input into operations control, and then into supplier
management/procurement systems.
At the other extreme, many control systems are basic, but effective nonetheless. Examples
are the restaurant table plan, the school timetable and the appointment book, which act as
both schedules and control systems. All of these‘systems’ can be examined and assessed to see
if, and how, changes can be accommodated. If a teacher reports in ill, the ‘manager’ will have
to assess which lessons need to be covered and who might be free to cover them. A restaurant
guest who extends the size of their party may be accommodated on a different table; the extra
car service for an important client may be added as overtime at the end of the day. In all
cases the existence of a detailed schedule/control system is critical to managing the day-to-day
planning and control of the operation.
Having an effective control system can be a significant source of competitive advantage. The
ability of the organisation to provide an immediate response to the question‘When can you do
this?’ is a major factor in building customer confidence. In recent years, retailers have been able
to interrogate logistics systems so that customers can negotiate a delivery slot for their purchase
of furniture or electrical products, with some confidence that this promise will be fulfilled.
The Karolinska Hospital operating theatre (see Case Example 11.3) provides a good example
of the benefits of a good schedule/control system. Such a system should include:
● A clear customer flow. The schedule provides the opportunity for customers to be ‘flowed’
through the system to arrive at the right time and place. We might compare customers in
this way to the work-in-progress (WIP) inventory in a manufacturing process. If WIP is
minimised, customer delays are reduced, there is less disruption, and costs are kept to a
minimum. An organisation that encourages customers to arrive early because it is unable
to manage its schedules may have to invest in larger reception (customer-holding) areas,
or suffer a major decline in customer satisfaction.
● Ensuring supporting resources are available to meet the schedule. Once the schedule has
been established making best use of the availability of scarce resource, all other resources
must be scheduled to meet this plan. For the hospital operating theatre this will involve the
scheduling of people (operating theatre staff and surgeons), and of course physical inventories
such as blood, bandages and surgical instruments. Restaurants, likewise, will match
inventories of food to expected customer demand.
Chapter 11 Managing service resources 297
● Creating schedules for interlinking activities. The schedule allows the operations manager
to create a realistic plan for the service as a whole and its many interlinking activities. In
some organisations this is termed the master schedule. This then facilitates the production
of supporting schedules for all the resources that feed this master schedule, using a variety
of scheduling rules matched to each situation.
● Creating schedules for suppliers. Good information provides a sound basis for negotiation
with key suppliers. This gives rise to the potential for managing day-to-day activity more
accurately with less waste (see Chapter 12).
11.4.3 Managing short-term schedules and mediumand
long-term capacity plans
The detail of schedules will increase as‘time now’ approaches customer requirement dates. To
return to the leisure centre example, tomorrow’s schedule will tell us exactly which customers
have booked a badminton court, and at what time. Moving further forward in time to, say,
next month, the manager will be interested only in which evenings are allocated to specific
activities which can then be booked by specific individuals as that week approaches. In the
much longer term, say a year, the manager will need to know the size of the facility available
and any plans to extend the sports hall.
It is essential to know the time period ahead within which it is impossible to change the
schedule. A restaurant may be able to deal with changes in customer mix almost up to the last
minute, whereas the operating theatre in Case Example 11.3 will require more notice of change
in detail schedule. We will return to this topic later in this chapter when we consider resource
flexibility, which aims to reduce the period ahead where the schedule must be fixed.
Karolinska Hospital faced a crisis as the pressure on operating
budgets was rising. This prompted an investigation as to how
well expensive resources were being utilised. It was soon identified
that operating theatres were not being used effectively. In
fact, surgeons, operating theatre staff and of course the theatres
themselves were idle for more than 50 per cent of the time. It soon
became clear that the schedule of patients through the theatre
needed to be managed more carefully. The scarce resource was
time in the theatre itself, so the management looked at ways to
reduce the time that patients spent in the theatre. A significant
step forward created a separate patient preparation area allowing
this activity to be carried out in parallel rather than in sequence
with surgery.
Further investigation revealed that some delays were caused
because anaesthetists were called away to other parts of the
hospital. Adding anaesthetists formally to the operating room
staff team and creating an anaesthesia clinic to evaluate preoperative
patients also improved the efficiency of the system.
Once the throughput through the bottleneck had been increased,
more operations could be carried out in the same
Case Example 11.3 Karolinska Hospital, Stockholm
Source: Pearson Education Ltd/Corbis/BrandX
Part 4 Deliver298
Bottlenecks (the parts of a process that constrain or restrict capacity),and the resulting queues
of customers or their orders, are features of many service operations. Managing these well can
have a significant impact on both capacity utilisation and customer satisfaction.
11.5.1 Bottleneck management
All organisations need to understand their key resource constraints. A clear understanding
of these constraints or bottlenecks provides greater clarity as to what is a realistic estimate of
capacity. To return to our example in Section 11.4, the key resource constraint of the leisure
centre is likely to be the central sports hall. All other resources, such as other facilities and
staff, are likely to be linked to the most effective use of this space.
Bottleneck management, or the theory of constraints, is generally well understood in manufacturing
organisations. It is seen to be important to manage the bottleneck – the stage in
any process with the lowest throughput rate and which therefore determines the effective
capacity of the whole operation.3
In the same way, it is important for service operations managers to understand where the
bottlenecks exist in service processes. For example, a company providing loan finance needed
to increase the standard of service provided to its customers, while also increasing productivity
of its risk assessment process. The management was given the task of meeting increasing
demand without increasing resources. Initially, it was not clear how this could be achieved,
but when the process was mapped it became obvious that a problem lay with the actuaries.
Figure 11.3a shows a simplified version of the original process flow, with all proposals passing
through an actuary’s hands for sign-off.
As can be seen from Figure 11.3a, the original capacity was constrained by the throughput
of the actuaries at 15 proposals per hour. It was recognised that many proposals did not
require actuarial sign-off because the credit scores indicated a clear accept or reject decision,
which could be taken by more junior, less expensive staff. The initial processing took
slightly longer, but in the revised process (Figure 11.3b) only 50 per cent of the proposals
needed to be seen by an actuary. The capacity of the process therefore rose by nearly 50 per
cent to 22 proposals per hour.
This improvement was achieved simply by monitoring the activity levels within the process
and then deciding whether the current resource constraints were really bottlenecks or not.
timeframe, and waiting times were dramatically reduced. The unforeseen benefit was that it was now possible
to create a much more reliable schedule because the lead time between diagnosis and surgery was
dramatically reduced. Patients were happier because they were treated faster and there were fewer ‘noshows’
than there had been when lead times were longer. A consistent theatre schedule meant that any tests
required prior to surgery could be arranged with more certainty. Previously, these tests had been frequently
repeated as surgery dates had been delayed.
For Karolinska, then, this approach to scheduling has paid off in a number of ways. The theatre carries
out more operations per day, costs are reduced and patients are seen more quickly. Operating rooms were
reduced from 15 to 13, while the number of operations per day was increased by 30 per cent.
Source: This illustration is based on material from the video Time Based Competition from Harvard Business School (1993), and from healthcare
industry sources.
How do organisations manage bottlenecks and queues?11.5
Chapter 11 Managing service resources 299
Once the assumption that all proposals must be seen by actuaries had been questioned, it
became possible to improve response times and productivity simultaneously.
There are some general rules for managing bottlenecks:
● Ensure that only essential work passes through the bottleneck.
● Be ruthless in taking away non-essential activities from the bottleneck.
● Ensure that no substandard work passes through the bottleneck.
● Once you have established where the bottleneck is, devote proportionally more management
attention to it to ensure maximum throughput and therefore maximum effectiveness
for the process.
And finally, if you have a complex system, the best thing to do is not to try to move the
bottleneck. It may be difficult to manage but at least you know where it is!4
One way to identify a bottleneck is to observe where the queues of work or customers form
in the process. In simple processes, this is probably as good a test as any, but it is as well to be
wary. Queues may form much earlier in the process because people operating feeder activities
may work at a slower rate than is theoretically possible because they believe there is no point
in working flat out if the customer or the jobs will have to wait at the subsequent bottleneck.
Indeed, service operations might well decide to limit the number of customers accepted into
the process because of the potential dissatisfaction caused for customers who have to wait
much longer than anticipated.
It is also important to distinguish between long-term and short-term bottlenecks. Longterm
or fixed bottlenecks provide the best estimate of the capacity of the operation, and the
basic capacity management approach can be determined as a result. However, short-term
bottlenecks may frequently occur, giving rise to the need for immediate response. For example,
a key member of a call centre may be unexpectedly absent, which may mean that
capacity to deal with particular enquiries is dramatically reduced.
11.5.2 Queue management
Queues occur in most service activities. Indeed, for any operation using a level capacity
strategy queues are ‘designed-in’. Furthermore, no capacity strategy is perfect and queues are
almost inevitable. Queues may be lines of people visible to both the customer and employee,
or they may be invisible to one and/or the other, as with a queue of callers to a switchboard
or a list of customers awaiting a repair engineer.
Figure 11.3b Revised flow
Inform customer of
decision (30/hr)
Risk analysis –
actuary (15/hr)
Proposal input and
credit score (22/hr)
Definite
yes/no
(50%)
Needs analysis
(50%)
Figure 11.3a Original flow
Inform customer of
decision (30/hr)
Risk analysis –
actuary (15/hr)
Proposal input
(25/hr)
Part 4 Deliver300
While queuing theory can be used to calculate the number of servers required to meet forecast
demand, resource constraints and forecast inaccuracy invariably mean that operations
managers need to look for other ways to minimise the impact of queuing on their customers.
It has been shown that not only dissatisfaction with the wait increases with waiting time5
but
also dissatisfaction with the service as a whole.6
Given that perceived waiting time is usually greater than actual waiting time,7
the answer is
to try to reduce perceived waiting time, which can also be a great deal cheaper than employing
more servers! Ten principles of waiting have been suggested:8
1 Unoccupied time feels longer than occupied time. It is a good idea to try to provide customers
with something to do or forms of distraction so that the time passes more rapidly
for them. Some services show promotional videos to people waiting in a physical queue.
Waiting areas for lifts often have mirrors to enable customers to check their appearance.
Telephone call centres or helpdesks frequently play music while ‘on hold’, although this is
not universally welcomed.
2 Pre-process waits feel longer than in-process waits. Once customers feel that they have made
a start inside the service process and that something, however trivial, is happening, they
tend to feel happier.A simple acknowledgement by a server that they have been noticed can
have a significant impact. Also, using pre-process time in some way, such as completing a
form or making choices about the service, can reduce the perceived waiting time.
3 Anxiety makes the wait seem longer. Sometimes customers do not know whether they
have been forgotten or not, which can be allayed by giving them numbered tickets to
demonstrate that they are part of the system. Also, the nature of the service will have a
significant impact. If the customer is worried about flying or going to the dentist the wait
may seem interminable, possibly giving rise to some tense behaviour with service providers.
Customer-facing employees should be trained to observe the effects of anxiety and to
find ways of giving reassurance.
4 Uncertain waits feel longer than known, finite waits. Customers are generally more happy
to wait if the expected duration is known, and if there is a good reason for it. If the duration
is unknown, research suggests that customers become restless much more quickly.
Theme parks frequently position markers at known points in the queue informing customers
how long they should expect to wait. Of course, the real wait time is usually a little
shorter than this, with customers pleased that they did better than expected!
5 Unexplained waits seem longer than explained waits. Being provided with a plausible explanation
of a delay reduces uncertainty for the customer. It also gives the impression that
the organisation knows it should not take the customer for granted.
6 Unfair waits seem longer than equitable waits. Generally, customers expect that those who
arrive first should be seen first. Many organisations have replaced the multiple-queue/
multiple-server system with a single-queue/multiple-server approach because of the perceived
unfairness of being stuck in a slow-moving queue. This approach also eliminates
the anxiety as to which queue to join. In some cases, such as a hospital casualty department,
there may be a good reason why some customers are seen out of turn, but it still
seems to be necessary for there to be an explanation rather than for the provider to assume
that other customers will understand.
7 The more valuable the service, the longer customers will wait. The more complex the service,
and the more it is customised to the needs of the individual, the more likely it is that
customers may be prepared to wait. It should be noted, however, that this should not be
assumed.
8 Solo waiting feels longer than group waiting. The realisation that others are also feeling the
pain may reduce the customer’s anxiety of thinking that they have made the wrong choice.
If others think it is worth waiting, it confirms the customer’s decision to wait. Also, people
tend to talk to each other, providing a distraction from the length of the wait.
Chapter 11 Managing service resources 301
9 Uncomfortable waits feel longer than comfortable waits. By making queuing conditions
as comfortable and indeed as distracting as possible, the wait time will be perceived
to be much shorter. Uncomfortable conditions sensitise customers to the time
and poor service.
10 New or infrequent users experience their wait as longer than frequent users do. Frequent
users of a service may be attuned to a wait and they may be more relaxed because they
know what to expect. New or infrequent users are likely to be more anxious and uncertain,
so operations should consider trying to identify them and provide them with information
and reassurance.
A booking system is a queue, with the advantage to customers that they do not have to
physically queue for the service. The advantage for the service provider is that the operations
manager is better able to manage resources to meet demand. Alton Towers (Case Example
11.2) has in effect created a ‘virtual queue’ through its fast-track system. Visitors to the park
are given time slots to return to a pre-booked ride which allows them to use their time more
effectively. Supermarkets that operate a ticket system at their delicatessen counters are using
the same principle. In both cases, the service provider has found a way to ensure equity of
treatment for its customers and has enabled customers to make better use of time otherwise
spent queuing.
11.5.3 Queuing theory and simulation
Management scientists and mathematicians have studied the behaviour of queues, producing
statistical models to predict queue length and so on. Fortunately, few of us need to understand
the detail of these models, as there are a number of computer simulations available to help us
predict the implications of operational decisions.
Simply put, there are three key parameters to queuing theory: the arrival rate for customers,
the server rate, and the number of servers or serving positions available. The arrival
rate and server rate must be further understood in terms of their variability. Even if the
average server rate and arrival rate are the same, queues will still form if there is variability
in these rates. This situation is illustrated in Figure 11.4, which demonstrates the impact
of variability in arrival rates and process rates. In this case, the process time is reasonably
constant – perhaps because it is a standard process, possibly determined to a large extent by
automation or standard scripts – while customers arrive in a more random pattern. In this
Figure 11.4 The impact of variability on queues
0 605040302010
Elapsed time
(minutes)
1
2
3
4
Customers in queue
Process completion
Average: 12/hour
Low variability
Customer arrivals
Average: 12/hour
High variability
Part 4 Deliver302
simple system, there may be up to four customers in the queue, the first being served and
three waiting. The longest wait time for customers is therefore 20 minutes (4 ϫ 5 minutes),
which perhaps is surprising when we know that average process time matches the average
arrival rate. The average queue length is about 2.5 customers, implying an average wait time
of 12.5 minutes.
Computer simulations now provide invaluable information to the service operations manager.
In a more complex situation than that described above, it would be impossible to model
the likely outcomes, but a simulation can identify the impact of different queue designs, priority
rules and so on. Whether the situation is complex or simple, the key question remains:
‘How long is an acceptable waiting time?’
We do not provide a detailed analysis of queuing theory here. Fundamental texts cover this
area well.9
However, useful terminology includes:
● Calling population. This describes the customer base. In many cases, the calling population
may consist of a variety of groups, each requiring different things. For example, customers
contacting the computer company’s call centre may enquire about billing, delivery,
faults, purchase of service contracts, and so on. A key element of queuing theory is the
size of the calling population. Consumer services have so many customers that the calling
population is thought of as infinite and the probability of the arrival of a specific customer
is unaffected by recent events. If, however, the calling population is relatively small, as
exemplified by potential callers to an internal computer helpdesk, the probability of new
callers is reduced if significant numbers have already called.
● Arrival process. It is clearly essential to understand the arrival pattern for customers. Many
arrival patterns follow an exponential distribution. Intervals between customer arrivals in
a retail store in a busy period might follow this distribution, with the majority of intervals
being rather short, and long intervals being somewhat rare.
● Queue configuration. This describes the number of queues and their location. In many
retail operations there has been a shift from multiple queues linked to multiple servers,
because some queues seem to move faster than others, leading to customers moving between
queues and possible ill feeling as customers believe they have been treated unfairly.
A single queue leading to multiple servers has the advantage of demonstrating equity of
treatment.
● Queue discipline. Management will choose the rule to determine who gets served next
from the queue. The most common rules were discussed earlier (in Section 11.4.1), though
first come, first served is the most popular with physical queues.
● Balking. One key measure for service systems is the number of customers who don’t join
a queue that they perceive as too long. This is referred to as balking.
● Reneging. This measures the number who join a queue, wait for a while then leave the
service system due to the perceived intolerable delay.
● Jockeying. This is the term for customers who switch from one queue to another hoping to
receive service more quickly.
There is usually a point at which service managers find it difficult to cope with increasing
demand (the break point: see Figure 11.5). This is when managers and staff enter the ‘coping
zone’. At these levels of capacity utilisation things are just too busy – staff become stressed,
everything becomes a problem and, importantly, perceived quality, i.e. customer satisfaction,
declines along with revenues per customer.10
What happens when managers can’t cope with demand?11.6
Chapter 11 Managing service resources 303
Figure 11.5 The coping zone in a high-quality restaurant
High
Low
Customerperceivedquality
Coping
zone
Break point
For example, in a restaurant operating at high levels of capacity utilisation, in the coping
zone
● Customers have to wait a long time for service.
● There is increasing likelihood of ‘stock-outs’ (items removed from the menu).
● Customers feel rushed and under pressure not to ask too much from busy serving staff.
● Staff feel under pressure and are less likely to give courteous responses or the personalised
service expected.
The break point is usually reached before full or 100 per cent utilisation for two reasons.
First, it is often not possible to run any single resource at full capacity for any period of time;
staff in a restaurant for example simply cannot work ‘flat out’ all the time. Second, several
resources may be involved and while staff might be working ‘flat out’, only 80 per cent of the
tables might be in use.
Interestingly there may be problems at times of low utilisation, too, which affect both staff
and customers. In the case of the restaurant,
● The perception of the overall quality of service experience is low because the restaurant is
‘dead’. There is no buzz of conversation; there are often prolonged silences.
● Service may be slow, because although there are not many customers, the kitchen may not
be working at maximum effectiveness.
● In the same way, serving staff may be less attentive than might be expected, because again
they may not be busy enough to be fully tuned in to customer needs.
Figure 11.5 illustrates the profile of customer perceived quality against capacity utilisation
and illustrates how quality may suffer through both too many and too few customers in a
high-quality restaurant.
The shape of the profile, the break point and the size of the coping zone will vary between
organisations. Figure 11.6 illustrates the relationship between customer perceived quality
(satisfaction) and capacity utilisation in a nightclub. Here customers may not enjoy the atmosphere
until the place is crammed with bodies! Still, at some point the club’s resources will
start to struggle with the demand placed upon its resources – door staff, toilets and bar, for
example – as it enters the coping zone.
11.6.1 How to manage the coping zone
There are seven steps in building up this profile and managing the coping zone.
Part 4 Deliver304
Step 1 Identify the service concept
Underpinning the service concept of a high-quality restaurant is the belief that customers
may book a table for the whole evening. They will not be rushed to vacate their table since
the restaurant has no intention of selling the space twice in an evening. It is intended that
the service experience should be relaxed, with staff able to converse with customers and
make recommendations about food and wine, where appropriate.
This is a very different concept from a restaurant wishing to create a high-energy situation,
often with staff rushing around, and with the customers encouraged to eat up and leave. It is
important to be clear as to what the designed or desired service concept is, particularly as the
restaurant gets busy.
Step 2 Determine how capacity utilisation is to be measured
For the high-quality restaurant, the best measure of capacity utilisation is the number of tables,
and also chairs, that are occupied during the evening. To some extent, other resources such as
serving staff or kitchen capacity can be adjusted to the busyness of the restaurant area.
The unit of measure is often best taken at the lowest level that the business analyses or
controls performance. A call centre might look at the average loading of a customer service
agent on an hourly basis throughout the day, whereas a professional service firm might look
at an individual’s case load.
Step 3 Draw the outline profile
Figure 11.5 shows the relationship between customer perceived quality and capacity utilisation
as it exists for the majority of customers on the majority of occasions. This does not
represent all customers at all times. Some customers, for example, prefer the empty restaurant
and would rank it as high quality at low utilisation. For others, the occasion and their mood
will have a significant influence on where they would place themselves on the profile. This
data can be captured from aggregating customer satisfaction indices and comparing it to
utilisation at the time.
Step 4 Understand the nature and impact of the coping zone
While we accept that low utilisation is as much of a problem for quality of service as high
utilisation is – indeed in some ways it is worse because revenue is also low – here we are
focusing on operations working in the coping zone. It is important to recognise the signals
Figure 11.6 The coping zone in a nightclub
High
Low
Customerperceivedquality
100%
Coping
zone
Break point
Chapter 11 Managing service resources 305
that suggest this zone has been reached and be sensitive to them. Tempers flaring, customers
looking around or queues appearing may all signal that breaking point has been passed. It
is also worth undertaking some financial analysis to try to demonstrate the impact between
passing the break point and its effects on costs and revenues, for example.
Step 5 Determine the ‘ideal’ operating area
In Figure 11.5 we have identified the break point at 80 per cent capacity utilisation. At 100 per
cent utilisation, in the case of the restaurant, it would be impossible to seat any more customers.
There are two broad approaches that can be adopted:
● Operate at 80 per cent capacity utilisation. Operating at this point would suggest that the
restaurant could be losing potential revenue. It is true that it might receive lower shortterm
revenues, but it may also upset longstanding customers by appearing to be greedy,
squeezing as many people as possible into every available space. It is critical to understand
the difference between customer satisfaction at 80 per cent and at 100 per cent, and to what
extent this significantly reduces the customers’ likelihood of returning.
● Operate at 100 per cent capacity utilisation. Generally speaking, this is a short-term cashgenerating
strategy. It is more appropriate to theme restaurants that are the ‘place to be
seen’ for a period, before those customers that are concerned about fashion move on to the
next ‘in’ place. This strategy might also be appropriate for restaurants in holiday resorts,
which do not expect high levels of customer returns.
Our example restaurant depends on long-term customer retention and word-of-mouth
advertising. As a result, it targets its operations at the 80 per cent point. To manage this, the
owner has removed some of the tables to give a less crowded feel to the area, only replacing
them on particularly busy occasions.
As a result, the owner has made the 80 per cent a 100 per cent effective capacity. In other
words, at this point, the restaurant is making sufficient revenue to meet its short-term financial
goals, and is giving a high level of customer satisfaction to safeguard its future business.
It is worth noting that because there is a gap between 100 per cent effective capacity
and 100 per cent potential capacity, it is possible for the restaurant to be working at greater
than 100 per cent utilisation on some occasions. For many managers this is the norm in
their businesses!
Step 6 Understand why coping happens
Clearly it is impossible to maintain the capacity balance on 80 per cent capacity utilisation at
all times. Even if the restaurant has a booking policy, there is always the possibility that one
of the most valuable customers will book at the last minute and the owner will be reluctant
to turn this business away.
In other situations, the launch of a new service or periods of faster than anticipated growth
may put parts of the business into ‘coping’ mode. This has been seen recently in the customer
service departments of mobile phone network providers and banks following product
launches. In some cases, some of the coping might have been avoided if the company had
carried out some forecasting, or had simply communicated internally.
A key point here is to recognise that all but extremely resource-rich organisations will be
in the coping zone sometimes. If the coping zone is never entered, the inevitable conclusion is
that the organisation has too much resource.
Step 7 Develop coping strategies
Most organisations cope after a fashion. In the restaurant, all the diners are given food,
but perhaps not with the greatest customer experience. Likewise, on the crowded flight, all
Part 4 Deliver306
passengers get a meal and a drink, though those that are served last may have limited choice
and little time to eat before the aircraft starts its descent.
Left to their own devices, customer-facing staff will find their own ways of coping. Some of
these informal coping strategies will be entirely appropriate and innovative, using interpersonal
skills and intuition to judge how to handle each customer. Others might be less satisfactory,
typified by the following examples:
● Waiters who become overly focused on one task, making it impossible for customers to
attract their attention to make yet more demands.
● Doctors’ receptionists who, faced with a crowded waiting room, become extremely efficient
in their dealings with patients, to the point of rudeness.
● Retail assistants who ‘forget’ to offer a customer a range of services, knowing that if the
customer chooses one of these, their workload will increase.
Operations managers develop coping strategies based on one or more of the following:
● Giving more information to customers alerting them to possible difficulties. An example
is an electricity company that after a major storm places a recorded message on its help
line to say ‘If you’re calling about loss of power in this district, we should be able to
restore it within two hours.’ This reduces the load on overworked telephone lines and
operators.
● Intentionally reducing the service on offer, perhaps using a limited menu at peak times in
the restaurant.
● Being clear to staff about what really matters most for customers: concentrating on the
‘must dos’ rather than the ‘nice to dos’.
● Building resource flexibility by bringing staff from a lightly loaded area to assist with the
overload. Call centres manage this by switching calls to other centres, whereas Disney brings
managers from back-office functions to assist with customer-facing operations on busy
days. It is important to note that some of this resource may not be as efficient as the normal
workforce.
There is a very strong link between prolonged overload and employee stress (see Chapter
10). It is relatively easy for providers to deal with short-term, predictable overloads. If we
know we’re going to be busy for a week or two, we can prepare for it, and many people get a
‘buzz’ from working together to cope with a crisis. The real problem with coping comes from
protracted periods of overload, without hope of a let-up in the foreseeable future. Management
support and appreciation becomes extremely important at this stage.
If the operation is in the coping zone for prolonged periods, it may be necessary for managers
to give their staff ‘licence to underperform’. For example, nurses in a busy accident and
emergency department may not be able to carry out all their duties in the way in which they
were trained. If this persists for any length of time, this will lead to stress and possible burnout.
Part of the coping strategy, therefore, is to agree which bits of the service are ‘must dos’
and which bits can be safely left for the time being.
11.6.2 Coping: key questions
We have devoted a lot of space to coping because understanding how the organisation deals
with this area may give clues as to where capacity management must be strengthened. The key
questions to address are:
● What does the customer perceived quality/capacity utilisation profile look like for your
service or services?
● How does this vary by service process and by customer group?
Chapter 11 Managing service resources 307
● What measures or early warning signals tell you that you are about to enter the coping zone
(as opposed to measures like lost customers or increased complaints which tell you that
you were in the coping zone)?
● What suffers for customers when you enter the coping zone?
● What suffers for employees when you enter the coping zone?
● How could you manage the coping zone better to reduce the impact on customers and
employees?
● How could you avoid being in the coping zone as much?
Of course, coping will affect every part of the organisation, in areas where both chase and
level strategies are operating, although coping is perhaps more obvious in operations that
are employing a chase strategy. In effect, chase becomes level in the short term because the
organisation is not capable of adding another unit of capacity quickly enough to deal with
an unexpected surge in demand. Coping is perhaps more sensitive here because, as we have
noted, organisations employ this strategy when fast response or high levels of availability to
customers are particularly important. In such circumstances, customers are not usually prepared
to wait, either because the service is not particularly valuable to them, or because there
are alternatives available to them.
How can organisations improve their capacity utilisation?11.7
There are four important additional ways of trying to improve capacity utilisation:
● yield management
● building flexibility
● reducing capacity leakage
● getting organisational support for capacity utilisation.
11.7.1 Yield management
Yield management is employed extensively by hotels and airlines to deal with the fact that
their capacity is perishable (see Case Example 11.4). In other words, if the hotel room is not
sold tonight, the contribution from that potential sale is lost for all time.
Yield management is focused on determining the maximum revenue to be obtained
from the various segments served by the capacity at hand. Thus the airline estimates
how many full-fare-paying (business-class) passengers will book for any given flight,
and adjusts the remaining capacity for economy-class passengers and other discount,
pre-booked customers. As departure time approaches, the airline may release some
capacity to discount travel shops and, as a last resort at the very last minute, to stand-by
passengers.
Service managers must be aware, however, of the potential damage to the service concept
in using this approach. Full-fare-paying customers may be unhappy to discover that the
person in the seat next to them is flying for a fraction of the price. This may give the impression
that the airline is merely after every last dollar of revenue, with customer satisfaction of
minor importance. The Kowloon Hotel appears to have overcome this particular objection
by creating a completely new concept where the charging policy is clear and unambiguous.
Customers can therefore make their choice of eating time, knowing that they will be treated
equitably.
Part 4 Deliver308
11.7.2 Building flexibility
There are four basic forms of operational flexibility:
● New service flexibility. This is the requirement of the service operation to introduce new
services into an existing mix. It will be necessary to define how frequently this might occur
and the extent to which the operation will require new capabilities to achieve it. For example,
house loan (mortgage) companies are continually introducing new ‘products’ with
varying interest rates and repayment terms. In this case the frequency of new ‘product’
introduction is extremely high, but the requirement for new capability is low.
● Service mix flexibility. This is the ability of the operation to deliver more than one service.
A hotel may provide a number of services simultaneously dealing with business people,
holiday travellers, conferences and wedding celebrations.
● Delivery flexibility. This is the capability of the operation to change the timing of the activity.
Courier organisations are increasing this form of flexibility, offering different speeds of
delivery and a range of pick-up and delivery times.
Sheryl E. Kimes, Cornell University
Yield management, the notion of charging
higher prices when demand is high and
offering discounts at times of low demand,
has traditionally been applied in reservationsbased
industries such as airlines, hotels and
car rental agencies. Managers at the Kowloon
Hotel in Hong Kong felt that it might offer
them the solution to improving their restaurant
revenues.
The Kowloon Hotel on Nathan Road in
Hong Kong is well known for its sumptuous
all-day buffet. The buffet, which includes a
selection of sashimi, oysters, salads and desserts,
is open from midday to midnight. As is
typical with most restaurants, customers only
wanted to dine at particular times of day, and
the restaurant was often empty in the late afternoon and late evening. To deal with this problem, the Kowloon
Hotel’s managers decided to move away from a single price for its buffet and charge different prices depending
on when customers arrive.
When guests arrive (check-in) they now receive a ‘buffet zone pass’. The cost of the pass varies depending
on their arrival time. At noon, the price is HK$118. It increases to $128 at 1.00 p.m., but then drops back
to $118 at 2.00 p.m. The 3.00 p.m. price is even lower ($108), but then progressively increases from $128
at 4.00 p.m., to $168 at 5.00 p.m., $208 at 6.00 p.m. and $248 at 7.00 p.m. Following this peak, the price
gradually decreases back to $138 at 10.00 p.m. and to only $98 at 11.00 p.m.
Not only has this new pricing system resulted in a 33 per cent increase in revenue – which was attributed
to a fuller utilisation of the restaurant space, hence an increase in revenue per available seat hour (RevPASH) –
it has also proved to be a hit with customers, with extremely positive customer reaction. As a result, the
management has decided to continue the time-of-day pricing for an indefinite period.
Case Example 11.4 The Kowloon Hotel, Hong Kong
Source: James Davies
Chapter 11 Managing service resources 309
● Volume flexibility. This form of flexibility is required by many consumer services operating
a chase strategy. It refers to the ability of the organisation to change its level of output to
cope with fluctuating demand. Thus the call centre may deal with 6,000 telephone transactions
on a normal day, but may have to cope with twice that amount following an advertising
campaign or a new product or service launch.
It is critical to define carefully the type and extent of flexibility required in order to develop
effective capacity management plans.
Figure 11.7 demonstrates the notion of minimum effective lead times – the time it takes
to respond to a change in demand. If an unexpected surge of demand occurs, the call centre
manager can increase capacity by 50 per cent in four hours. This is accomplished by asking
staff to stay on after their normal shift, by calling in off-duty staff, and by bringing in other
staff from the organisation to man the phones. It should be noted that this is very much a
‘first’ or emergency response as some of the staff will not be fully trained and productivity
levels will probably suffer as a result. Within eight hours the call centre manager can
bring on line more capacity, perhaps from other call centres, if the increase in demand is
sustained.
This is a valuable tool in assisting operations managers to plan for foreseeable contingencies.
In addition to specifying the type of flexibility required, the service manager must also
consider the following:
● Range. How much flexibility is required? Does the call centre need to move from 6,000
calls to 12,000 calls or only to 8,000 calls per day? How many new services will be introduced
and how frequently?
● Response. How quickly must the change be made? Can the call centre change from 6,000
to 12,000 calls in four hours or will eight hours be good enough? Clearly the faster the
response, the more expensive it is likely to be.
● Effectiveness across the range. Most processes have an optimal range. It is unlikely that
they will be equally productive across the potential range; they are likely to strain at the
extremes.
● Cost of providing the flexibility. What is the premium both for the change in output level
itself, and for providing the capability in the first place? For example, providing training so
that more staff are multi-skilled represents a significant investment.
There are a number of approaches to building flexibility. These include:
● Flexible employment contracts. Employees may be asked to work a given number of hours
per week, month or year. In this way, staffing levels may be more easily matched to expected
Figure 11.7 Minimum effective lead times for a call centre
Sustainable
response
12,000
First
response
‘Normal’
operating
level
Callsperday
6,000
4
Response (hours)
8
Part 4 Deliver310
demand. Flexibility in this area may extend to the requirement for staff to move between
functions to cover fluctuating workloads. It is important to recognise that these staff may
not be as effective when carrying out tasks that are less familiar to them.
● Overtime. Asking staff to work longer hours is a common short-term measure for building
volume flexibility. However, it is generally accepted that using overtime in the longer
term does not give the required increase in output as employees simply expand work to
fit the extra time available. In this event, overtime is a poor approach, costing a great deal
in extra salary and creating employees who are overtired through being at work for long
periods.
● Short-term outsourcing. The development of relationships with other service providers
who can deal with short-term peaks in demand is a common practice, particularly in call
centre activities. This requires some pre-qualification of suppliers, but can provide flexibility
at relatively low cost.
● Menu-driven service (standardisation). Standardisation of service offers provides opportunities
for increased volumes of fewer activities. This creation of ‘runner’ activities (see
Chapter 8) smooths the workload and provides opportunities to provide a degree of customisation
or personalisation for customers. This is often referred to as the‘Dell’approach,
whereby customers can create their own configuration of computer from a wide range
of standard modules on offer. Education providers such as the Open University adopt
a similar approach in allowing students to tailor their courses, within pre-determined
constraints.
● Teamwork. The development of multifunctional teams allows a group of employees to
deal with fluctuating workloads rather than simply operate on an individual basis. There
are several benefits to this approach, including the support that group members give each
other, and the ability to ‘flex’ capacity very rapidly.
11.7.3 Reducing capacity leakage
Many operations managers carefully plan their capacity in ways we have discussed in this
chapter; however, sometimes managers are surprised to find they do not have as much capacity
as expected. Here are some possible explanations:
● Labour sickness and absenteeism. Prolonged periods of overload and compulsory overtime
usually prove counterproductive, with staff taking time off to recover. Alternatively, the
organisation may need to look at its management style, placing more emphasis on team
building rather than ‘command and control’ approaches.
● Labour underperformance. It is extremely common to find that call centres may have the
right number of ‘heads’ but they may be ineffective because there has been insufficient investment
in training. Alternatively, employee churn means that experienced staff leave just
at the point when they are becoming effective.
● Scheduling losses. There are times when staff are idle with too much capacity for the demand,
whereas at other times there is too much demand for the capacity to deal with. This
often arises because demand profiles are not understood or are too volatile, or where staff
preferences for work patterns do not fit with the business need.
● Costs of complexity. The more the organisation deals with a broad range of services, the
greater the possibility that staff deal with a greater percentage of tasks that may not be part
of daily routine. This potential ‘relearning’ may give rise to inefficiencies and rework.
● Quality failures. The need to deal with quality failures is clearly lost capacity. Of course,
part of the role of the call centre may be to deal with poor quality generated elsewhere in
the organisation. It is essential that the extent of this rework is understood and charged to
the appropriate location.
Chapter 11 Managing service resources 311
11.7.4 Organisational support for capacity utilisation
The challenge for operations managers is to understand capacity utilisation in the context
of a changing world. Many of the issues need to be resolved by the organisation as a whole,
rather than simply confined to the management of service delivery processes. Aspects of this
organisational support include:
● How is the service concept changing? To what extent do operations managers have‘visibility’
in the future strategic direction of the organisation? Without this inclusion in the strategic
development process it is unlikely that capacity planning and capacity utilisation and
development will be effectively carried out. Often resource managers are left to develop
capacity plans that have little relevance because the concept has been changed. For example,
many service companies providing repair and maintenance support for information
technology have changed the emphasis of their service concept away from servicing equipment
towards supporting customers. The nature and length of customer transactions have
changed beyond recognition, and a link is required between this change in strategic direction
and resource planning for it to be implemented.
● How well are the internal interfaces managed? A key role for the operations manager is to
manage the internal relationships as well as customer relationships. Co-ordination of marketing
promotions and new service introductions is vital, as is getting to the root of quality
failures and long-term quality costs. Successful organisations are often those that manage
the internal relationships well. This does not necessarily mean that everyone always agrees,
but that there is valuable internal debate. In fact, it could be asserted that an organisation
without some degree of conflict will not learn and move forward.
● How important is resource management in the culture of the company? In some sectors,
resource management is seen as a low-level task. The ‘stars’ of the company are often seen
as those who deliver the latest deal or solve the latest crisis. Resource management needs a
different type of ‘hero’ who is able to plan for the longer term and persuade the organisation
to think differently about resource management.
Summary11.8
What is capacity management?
● Capacity management is concerned with putting a plan in place that makes the best use of
resources to deal with the forecasted or expected demand for services.
● Service capacity is defined as the maximum level of value-added activity over a period of
time that the service process can consistently achieve under normal operating conditions.
● Capacity is influenced by a range of factors including service mix, location, intangibility
and resource constraints.
How can managers balance capacity and demand?
● Most organisations adopt a mixture of capacity strategies: level, chase and demand management.
The mixture should reflect the strategy of the operation.
How is day-to-day planning and control carried out?
● Day-to-day operations planning is concerned with creating a ‘schedule’ or timetable
based on the capacity plan, which allocates people, customers, equipment or facilities,
Part 4 Deliver312