Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Optimizing PC Refresh Cycles to Maximize Business Value
Wipro Consulting Boston, Massachusetts
Sponsored by: Intel Corporation
© 2006 Wipro Consulting. All Rights Reserved.
Copyright © Wipro Ltd. 2006. All rights reserved. No portion of this study can be used or reproduced without permission of the author. For additional reproduction rights and usage information, go to www.wipro.com. Information is based on best available resources. Opinions reflect judgment at the time and are subject to change.
Wipro / Intel PC Refresh Optimization October 2006 2
Table of Contents
Executive Summary ............................................... 2 Our Findings ......................................................... 3 Optimizing Refresh Cycles: A Quantitative Analysis.......................................... 4 Methodology .......................................................... 5 Process Focus........................................................ 5 Assumptions Used to Model a Representative Environment .................................. 7 Insights from Survey Responses and Data Analysis.................................................. 9 Comparing Studies............................................... 10 Comparing Optimum Refresh Cycles................... 11 Software Deployment Cost Trends ....................... 13 The Impact of Complexity on IT Responsiveness................................................ 15 Conclusions ......................................................... 18 Appendices........................................................... 19
Appendix A: A Glossary of Terms Used in This Survey ........................................ 19 Appendix B. Appendix B: Selected Demographics of Survey Respondents ............ 20 Appendix C. Overview of Wipro PSA.............. 21
Executive Summary IT managers who are responsible for their companies’ PC
strategies must be on the lookout for the formula that will
help them maximize the productive use of their PC installed
base, while minimizing the cost of purchasing, provisioning,
and supporting those PCs. In other words, they need to
optimize their hardware refresh cycles. But the mix of users
has become more diverse and IT support responsibilities
more complex due to the popularity of mobile computing and
wireless technologies. Users expect the same level of
responsiveness from IT whether they are using their desktop
PCs, mobile PCs, notebooks or handhelds in corporate
offices, conference rooms, remote offices, home offices,
airports, cafes, hotels or wherever in the world they happen
to be. Understanding the cost impact of PC refresh cycles
under these circumstances is critical to setting—and working
within—a realistic, cost-effective IT budget.
This paper presents the results of a new study designed to
help IT management realize optimal refresh cycles.
Conducted by Wipro Product Strategy and Architecture
(PSA) and sponsored by Intel, this study extends and
validates the work described in an earlier Wipro PSA
research paper titled “Measuring the Benefits of Mobile
PCs in the Enterprise.” The objective of this latest is to
provide up-to-date information that identifies lifecycle
management best practices for the entire PC mix
throughout an organization.
DEFINING PC PLATFORM AND USER TYPES
• Desktop Users – Referred to as “Desktop” throughout
this document, these employees use a desktop PC
during their working hours.
• Office Warriors – Or “Office” as they are referred to
in this document, are employees who use a mobile PC
(i.e. a laptop or notebook), and are physically present in
the office at least 80 percent of their working hours.
• Road Warriors – Referred to as “Road” in the body of
this document, these employees use a mobile PC (i.e.
laptop or notebook), and are physically present in the
office for, at most, 20 percent of their working hours.
• Mobile – Will be used throughout this survey to
indicate the total population of laptop and notebook
users, including both Office and Road Warriors.
Wipro / Intel PC Refresh Optimization October 2006 3
Our Findings Wipro PSA conducted extensive interviews with senior
technology managers from 30 organizations, representing
diverse industries and PC fleets ranging from 7,000 to
110,000 systems. Their responses regarding PC
management practices and associated costs formed the
basis for the statistic analyses in this study. Our key
findings are summarized as follows:
1. Configuration-related support costs for Desktop, Office
and Road PCs all increase directly with the length of
refresh intervals. While the primary cost drivers are
software deployment failures and help desk calls, the
rate of increase varies according to both platform and
user types. When compared to the findings from the
aforementioned 2004 study, the trend in these numbers
remains the same, with cost increasing as the lifecycle
extends. The real dollar values for each cycle have
lowered overall, but the trend is constant.
• Desktop PC management costs rise $93 between
service years two and three, and $127 between
service years three and four.
• Office PC management costs jump $159 between
service years two and three, and $172 between
years three and four.
• Road PC management costs rise $4 between
service years two and three, and increase by $150
between years three and four.
2. The optimum lifecycle for each PC platform type
represents a balance between fixed cost amortization
and the platform’s specific rate of support cost increase.
• For Desktop PCs, the optimum refresh interval is
three years – the same as in 2004.
• For Office PCs, the optimum refresh interval is
three years – the same as in 2004.
• For Road PCs, the optimum refresh interval is two
years – the same as in 2004.
Organizations can derive the maximum benefit from
managing PC lifecycles against these refresh rates. Those
companies already following these refresh cycles should
continue to do so, as this study and the previous one have
determined that these refresh cycle lengths should be
considered best practices.
3. The overall cost profile of office systems is closer to
that of road machines than desktops, which may seem
counterintuitive since the Office Warrior profile (80%
of time in the office) is more closely aligned with the
Desktop profile (100% of the time in the office). The
difference between the cost profile of Desktop and
Office is much larger than between Office and Road.
Types of usage (i.e., Desktop and Office are both
largely sedentary usage models) are less important
than similarities between systems (i.e., Office and
Road are both laptop systems).
4. Despite the fact that Office and Road acquisition costs
are nearly identical, the use and support costs for Road
are higher than for Office, which accounts for the gap
between the two profiles.
5. Minor software deployment or patching is more
widespread and automated than in the past. However,
as enterprises have consolidated around Windows XP,
the cost to roll out frequent patches to a large volume
of machines has driven up the overall total cost of
ownership for Office and Road systems.
6. By implementing a proactive client refresh—retiring
older desktop machines, reducing the number of
hardware configurations and reducing aggregate
support costs—companies can recoup the capital costs
of replacing desktop PCs with mobile Office machines
that will increase user productivity. Generally,
workers with Office machines benefit from the ability
to work with their mobile PCs while in meetings
inside their office environment, often utilizing
wireless connectivity to access critical data and
services. This ability to be always connected and able
to use computing resources substantially improves
productivity.
Wipro / Intel PC Refresh Optimization October 2006 4
FIGURE 1. The time required to complete software
security patch deployment increases for all platforms with
infrastructure complexity related to longer refresh cycles.
Increased exposure to risk coincides with longer
timeframes for completing patch deployments.
7. In all cases, greater complexity in the IT environment
correlates with higher maintenance and management
costs. In addition, a complex PC population can
impede the IT organization’s ability to deliver routine
but critical software testing and deployment services.
This can result in a significantly greater security risk
exposure. Finally, extending the refresh cycle beyond
the optimal rate dramatically increases the number of
hardware failures that occur during the patching
process which, in turn, requires high levels of
responsiveness (and, potentially, higher numbers of IT
personnel). The alternative is increased security risks
on neglected older machines. Figure 1 illustrates the
effort required to apply a security patch to a PC.
Optimizing Refresh Cycles: A Quantitative Analysis In PC infrastructure, diversity—a wide range of hardware
configurations across the installed base—is generally
considered to be a bad thing. As Wipro PSA has shown in
past research, there is a direct relationship between
diversity of hardware configurations and the associated
levels of hardware-related support costs. Many people have
assumed this relationship through the years. In this study,
Wipro confirms it by investigating the differing rates of the
support cost increase associated with each platform and use
type at planned system lifecycles ranging from two to five
years. Moreover, this study provides a quantitative analysis
of the economic benefits of controlling hardware-related
diversity and optimizing PC refresh cycles.
Where relevant, this latest study compares specific data
with the previous study from 2004, “Measuring the
Benefits of Mobile PCs in the Enterprise.” In general, the
patterns and best practices identified in 2004 remain the
same: Refresh cycles that stretch beyond an optimal point
create additional cost, complexity, security risks, and lower
productivity. IT should consider all of these factors when
planning the PC refresh cycle, since cost isn’t the only
significant issue.
As an example, in the current study, the data shows that the
cost difference between two- and three-year refresh cycles
for Road PCs is minor. But by pushing out the refresh to
three years, IT must be aware that the critical security risk
point is much closer, meaning that pushing the Road PC
refresh cycle beyond three years will not only dramatically
increase costs but will also increase the risk of a
detrimental security breach, which could be costly indeed.
Industries Surveyed for This Study
Telecommunications Media and Entertainment
Retail Manufacturing
Transportation and Distribution Energy and Utilities
Banking Government
Insurance Professional Services
FIGURE 2. Wipro surveyed the opinions of senior IT managers
from 30 companies and representing a variety of industries.
Wipro / Intel PC Refresh Optimization October 2006 5
Methodology In mid-2006, Wipro PSA analysts interviewed CIOs, IT
directors, and senior IT managers at 30 companies
headquartered in North America and Western Europe.
Companies selected for this study represent a diversity of
industries, management practices and user distributions,
with PC user populations ranging from 7,000 to 110,000.
Figure 2 provides the industry categories from which the
companies were drawn. All interview subjects are actively
involved in the planning and execution of the management
processes discussed in the survey. On average, they are
directly responsible for 47 percent of their respective
organizations’ total installed PC base.
IN-HOUSE OPERATIONS All the companies surveyed conduct the IT business
processes analyzed in this study in-house. Some use
contract staff for certain roles but retain final control and
accountability within the organization. This in-house
criterion ensures that all reporting reflects direct, hands-on
experience with actual management practice.
PC LIFECYCLES To facilitate cost comparison, the surveyed companies were
grouped according to their PC lifecycle length in two-,
three-, four-, and five-year categories. Interestingly, firms
refreshing on two- and three-year intervals tend to
proactively decommission desktop and mobile PCs.
Conversely, firms with longer lifecycles tend to “cascade”
older machines, passing them down through the ranks and
replacing them only after serious failure.
It is worth noting that since Wipro began measuring
desktop and mobile lifecycle support costs in 2003, we
have observed a steady decline in the practice of cascading
four- and five year-old PCs, accompanied by a clear trend
towards the proactive elimination of the oldest and most
costly configurations.
Process Focus Many variables are involved in the measurement of desktop
and mobile PC complexity. As expected in any sample of 30
firms, Wipro found some variation in terminology, standards,
tools, processes, networks, and systems. To keep the analysis
consistent across the 30 firms represented, we kept the focus
of the study on common IT business processes related to PC
hardware and software deployment, configuration-related
lifecycle support, and the costs associated with both
deployment and support.
HARDWARE AND SOFTWARE DEPLOYMENT PROCESSES INCLUDE: 1. Installing new client computers
• Hardware Qualification – The standard process of
qualifying new hardware configurations prior to
deployment
2. Installing new operating systems on existing PCs
• OS qualification – Testing and qualifying an OS
on existing desktop and mobile PC hardware
configurations
• OS deployment failure resolution – Resolving OS
deployment failures resulting from hardware
components or drivers
3. Installing major application software and updates on
existing PCs
• Application/update qualification – Testing and
qualifying an application or update on existing PC
hardware configurations
• Application/update deployment – Packaging an
application for automated deployment
• Application/update deployment failure resolution –
Resolving application deployment failures
resulting from hardware components or drivers
Wipro / Intel PC Refresh Optimization October 2006 6
FIGURE 3. This analysis examined only those tasks that pertain directly to managing desktop and mobile hardware
configurations (shaded tasks), and thus focused on a subset of desktop management processes. Tasks marked with a dagger (†)
are beyond the scope of this study but result in costs that may also be affected by hardware diversity.
LIFECYCLE IT SUPPORT TASKS RELATED TO PC HARDWARE CONFIGURATIONS INCLUDE: 1. Help desk support – The IT workload associated with
receiving, prioritizing, dispatching and resolving help
desk calls related to hardware components and drivers
2. Spares and replacements provisioning – Procuring and
distributing spare batteries, power supplies and
miscellaneous accessories
3. Decommissioning tasks – Activities related to
removing PCs from service, including data cleansing
and used equipment sales
4. Documentation – Tracking and documenting hardware
in the installed base
5. Training – Training IT staff and end-users on the use
and maintenance of hardware in the installed base
6. User relocation support – Activities related to the
movement of users and their PCs. For Desktops, this
includes physical movement of the systems,
reconfiguration and reconnection to the network. For
Office and Road Warriors this may include relocation of
base stations and network or system re-configuration.
THE SURVEY AND INTERVIEW Respondents were given detailed surveys that were
distributed several days before scheduled follow-up
telephone interviews. After respondents completed the
survey, the telephone interviews provided an opportunity to
clarify or expand upon their answers. Using specific
information from respondents, Wipro analysts generated
detailed labor cost data by determining:
• The frequency of particular events (hardware
qualifications, etc.)
• The typical time for accomplishing the tasks associated
with each event
• The labor rates for employees performing the tasks
Wipro / Intel PC Refresh Optimization October 2006 7
FIGURE 4. A graphical representation of the structure of the interview guide.
As illustrated above, the interview guide was structured to
follow the lifecycle of desktop and mobile PC systems.
Sections of the guide focused on the following:
• Gathering general background on the respondents’
companies, such as number of employees, number of
client systems, and IT staff hourly rates
• Capturing the costs of system management across the
lifecycle of desktop and mobile PC clients, including
sourcing and procurement
• Processes and costs related to the configuration, testing
and qualification of hardware and software
• Ongoing support and administration costs, and
• Client decommissioning
Assumptions Used to Model a Representative Environment Based on detailed data collected from interviews, Wipro
modeled a representative environment composed of 50,000
Desktop, Office and Road PCs in order to assess the impact
of environment complexity and lifecycle length on support
costs, while holding other variables constant. We believe that
many of the findings associated with the 50,000 PC model
are also relevant for smaller organizations with PC fleets of
10,000, or possibly fewer, because many of the larger
enterprises surveyed did not centrally manage their client
environments. Instead, they managed their client PCs either
by line of business or geography, thus emulating much
smaller organizations and limiting their ability to achieve
economies of scale in managing sizeable user populations.
Wipro / Intel PC Refresh Optimization October 2006 8
FIGURE 5. Assumptions used to model
a typical enterprise PC environment.
The assumptions listed in Figure 5 were used to model
typical desktop management practices, and to evaluate
the impact of complexity and lifecycle variation on desktop
PC support costs. This survey used U.S. and European
labor rates to associate desktop PC support practices with
labor costs.
SPECIAL NOTE ON THE MODEL ASSUMPTION: COST OF A NEW LAPTOP In the 2004 study, “Measuring the Benefits of Mobile PCs
in the Enterprise,” the model assumptions included a cost
of $1,750 for a new laptop PC. As hardware prices have
generally fallen since 2004, this study uses a model
assumption of $1,350 for a new laptop PC. The only area in
which this change created a difference is in the refresh
cycle for Road PCs. Using the original $1,750 laptop price,
the optimum refresh cycle for Road PCs in the new study is
three years. But at the current and more accurate
assumption price of $1,350 for a new laptop, the new study
shows that the optimum Road PC refresh is two years,
which is the same as the 2004 study.
The optimum refresh cycle for Office PCs, which are also
laptops priced at $1,350 in the model assumptions, is three
years. This is the same as the study in 2004.
Model Assumptions Value
Number of PCs 50,000
Number of Vendors 3
Number of Models Deployed per Year 2
Number of PC Configurations Deployed per Year 6
Number of Applications Qualified per Year 7
Number of Applications Deployed per Year 11
Number of Minor Updates Qualified per Year 38
Number of Minor Updates Deployed per Year 38
Frequency of Operating System Deployments (Years) 2.5
Level 1 Resource Burdened Labor Rate (U.S. Dollars) $42.48
Level 2 Resource Burdened Labor Rate (U.S. Dollars) $59.93
Level 3 Resource Burdened Labor Rate (U.S. Dollars) $82.41
Cost of New Desktop PC $800
Cost of New Office Laptop $1,350
Cost of New Road Laptop $1,350
Wipro / Intel PC Refresh Optimization October 2006 9
Insights from Survey Responses and Data Analysis Older Desktop PCs are more expensive to support
• Support costs jump from $344 to $471 per year as Desktop PCs age from three to four years—a 37 percent increase. This is due to increased software deployment costs, including OS and major and minor application deployments.
• Minor software deployment costs increase particularly fast, from an average of $125 at three years to $172 at four years and $218 at five—increases of 38 and 27 percent, respectively, over the three-year rate. Respondents told us that this correlated with incompatibilities introduced by older hardware (often unsupported), as well as the fact that older PCs tend to run older OS versions.
• Hardware-related help desk calls become more frequent and consume more time, leading to increased costs.
• There is a consistent pattern for Desktop PCs in this and the 2004 study. Specifically, the real dollar cost is higher, but the pattern of increase and optimal refresh cycle are the same.
Office Warrior PCs follow a similar trend
• Support costs jump from $592 to $763 per year as Office Warrior PCs age from three to four years—a 29 percent increase.
• As with Desktops, this is due to increased software deployment costs, including OS and major and minor application deployments.
• Minor software deployment costs for these PCs increase at a similar rate to desktops, from an average of $322 at three years to $399 at four years and $505 at five—increases of 24 and 27 percent, respectively, over the three-year rate.
• More frequent and lengthier hardware-related help desk calls become a major cost factor as Office systems age. Our analysis reveals that up to 20 percent of help desk costs are incurred when the user is out of the office, and that wear and tear on laptops is consistently higher than on Desktop systems
• When compared to the data for Office Warrior PCs from the previous study, the pattern remains constant, except now minor software deployment costs are significantly higher than in the previous study. This issue is covered in detail later in this paper.
Road Warrior PCs also become more expensive as they age
• Support costs start relatively high and increase markedly, going from $629 to $858 per year as Road Warrior systems age from three to four years—a 36 percent increase.
• As with Desktop and Office systems, this is due to increased software deployment costs, including OS and major and minor application deployments,
• Minor application deployment is the major cost contributor right from the start, growing to 61 percent of total costs by year five. This is consistent with the trend we saw in Office systems but more pronounced.
• The lack of decommissioning costs points to two phenomena for which we have only anecdotal evidence:
• Road systems are commonly recycled or cascaded as Office systems
• Many Road systems are not as thoroughly decommissioned as their Office and Desktop counterparts
• When compared to the previous study, Road Warrior systems follow the same trend and, as with Office systems, minor software deployments are the most significant aspect of overall cost.
Wipro / Intel PC Refresh Optimization October 2006 10
TO REDUCE COSTS, REDUCE THE NUMBER OF VENDORS
One result of our analysis has been to correlate the number
of PC vendors used by an organization with the per-PC
support costs incurred. Our data reveals that firms with four
vendors spend 20 percent more in support costs, on
average, than firms who use just two vendors. This is
consistent with our earlier finding that the number of
configurations directly influences support costs. Therefore,
CIOs can further lower their support costs by carefully
reducing the number of PC vendors they use, keeping in
mind that some PC vendors excel at helping organizations
manage costs.
Comparing Studies
As mentioned previously, key trends uncovered in the 2004
study hold constant in its 2006 counterpart, with few
exceptions. Specific data points that vary from the 2004
study have been addressed in the main body of the text.
That notwithstanding, the following table summarizes some
of the key issues, noting two interesting differences.
2004 Findings 2006 Findings
Optimum Desktop PC refresh is three years
Optimum Office system refresh is three years
Optimum Road system refresh is two years
Optimum Desktop PC refresh is three years
Optimum Office system refresh is three years
Optimum Road system refresh is two years
Hardware-related help desk support costs are the major
contributors to overall support costs for Road systems.
Minor software application deployment is the major
contributor to overall support costs for Road systems. The
“Software Deployment Cost Trends” section of this paper
goes into more detail on this issue.
Help desk call volume increases at three and a half years
into a PC’s lifecycle and then levels off.
Help desk call volume spikes dramatically at three and a half
years into a PC’s lifecycle and continues to climb. This
increase is much higher than previously seen, further
emphasizing the need to follow refresh cycle best practices.
Operating system heterogeneity is extremely high. Operating system heterogeneity is extremely low. A large
majority of organizations have standardized on the Windows
XP operating system for PCs since 2004. This introduces
new management issues that are discussed in the “Software
Deployment Cost Trends” section of this paper
Wipro / Intel PC Refresh Optimization October 2006 11
Comparing Optimum Refresh Cycles
FIGURE 6. Cost optimizing the Desktop lifecycle
THE OPTIMUM DESKTOP PC REFRESH INTERVAL: THREE YEARS
Our analysis indicates that longer lifecycles result in higher
support costs. Because of this, determining the most cost-
effective desktop lifecycle will optimize a company’s cost
to benefit ratio.
Figure 6 shows the sum of hardware and support costs
given an initial cost of $800 for a Desktop PC (excluding
keyboard, mouse, and monitor). A three-year refresh offers
the lowest total yearly cost at $610 per Desktop PC,
compared to $671 for a four-year lifecycle and $741 for a
five-year lifecycle. For a firm with 50,000 Desktop PCs,
this translates to an annual total cost of $30.5 million for a
three-year refresh,
$33.6 million for a four-year refresh, and $37.1million for a
five-year refresh.
Firms on a three-year lifecycle invest an additional $67 per
Desktop PC each year on hardware, deployment and
decommissioning costs compared to firms on a four-year
cycle, but they manage to save $127 per PC each year in
support costs. The result is a net savings of $60 per
Desktop PC per year—a $1.05 million annual savings for a
fleet of 17,519 Desktop PCs, which is the average number
for respondents in this study.
“We try to keep client PCs for three years. After that, they
are decommissioned and recycled.”
– Infrastructure Director, retail industry
Wipro / Intel PC Refresh Optimization October 2006 12
FIGURE 7. Cost optimizing the Office system lifecycle
THE OPTIMUM OFFICE PC REFRESH INTERVAL: THREE YEARS
Using the same logic and similar core assumptions, Wipro
analysts deduced that three years is also the optimal refresh
cycle for Office systems. Figure 7 shows the sum of hardware
and support costs for an Office system, given an initial cost of
$1,350 (excluding keyboard, mouse, monitor, and/or docking
station). A three-year refresh offers the lowest total yearly cost
at $1,042 per Office system, compared to $1,101 for a four-
year lifecycle and $1,217 for a five-year lifecycle—increases
of six and eighteen percent, respectively. The additional
security exposure and cost to mitigate this exposure increases
significantly from years three to four. Combining the cost
increase and security risk exposure makes for a compelling
case to refresh at three years.
For a firm with 50,000 Office systems, this translates to a
total yearly cost of $52.1 million for a three-year refresh,
$55.1 million for a four-year refresh, and $60.9 million for
a five-year refresh. On average, firms on a three-year
lifecycle thus invest an additional $112 per Office system
each year on hardware, deployment, and decommissioning
costs compared to firms on a four-year cycle. However,
they save $171 per machine in support costs each year. The
result is a net savings of $59 per Office system per year—a
$260,000 annual savings for a fleet of 4,410 Office
systems, which is the average number per respondent in
this study.
FIGURE 8. Cost optimizing the Road system lifecycle
Wipro / Intel PC Refresh Optimization October 2006 13
THE OPTIMUM ROAD SYSTEM REFRESH INTERVAL: TWO YEARS
The most cost-effective refresh cycle for Road systems
according to Wipro analysts is likely to come as a surprise.
Figure 8 shows the sum of hardware and support costs
given an initial cost of $1,350 for a Road system (excluding
keyboard, mouse, monitor, and/or docking station). A two-
year refresh offers the lowest total yearly cost at $1,304 per
Road system versus $1,308 for a three-year lifecycle,
$1,457 for a four-year lifecycle, and $1,626 for a five-year
lifecycle.
For a firm with 50,000 Road systems, this translates to a total
yearly cost of $65.2 million for a two–year refresh cycle,
$65.4 million for a three-year refresh, $72.9 million for a
four-year refresh, and $81.3 million for a five-year refresh.
Firms on a two-year lifecycle will invest an additional $225
per Road system each year on hardware, deployment, and
decommissioning costs compared to firms on a three-year
cycle. But they will save $229 per machine in annual
support costs. The result is a net savings of $4 per Road
system per year—a $7,684 savings on a fleet of 1,921 Road
laptops, which is the average number for respondents in
this study. While this difference is quite small, consider
that additional security risks increase as a Road system
moves into its third year of service, with that security risk
exposure dramatically increasing at 3.5 years. For these
reasons, two years is the most cost-effective and secure
refresh cycle for Road systems.
FIGURE 9. Side-by-side cost comparison of Road, Office and Desktop systems for various lifecycles.
Wipro / Intel PC Refresh Optimization October 2006 14
COMPARING OPTIMUM REFRESH CYCLES
Figure 9 presents the relative cost profiles for Desktop,
Office, and Road systems at one-year lifecycle intervals up
to five years. As expected, for any given refresh cycle,
Desktop PCs are the least expensive and Road Warrior
most expensive. Aggregate costs for Road systems are
markedly higher than Desktop and Office machines.
IT organizations that already use these refresh cycles are
utilizing best practices. They are not only saving their
companies money, they are also minimizing downstream
security issues and support headaches associated with older
PCs. IT organizations not following these best practices are
advised to look more closely at their own practices and
build and execute a strategy to align PC refresh with these
optimal cycles.
“My experience has been that older PCs cause more
problems than they are worth. We stick to our refresh
plan because I know that this saves my staff a lot of
time and effort.”
– VP of IT, insurance industry
Software Deployment Cost Trends One of the new issues that emerged in this study as
compared to the previous study is the rise in cost of minor
software deployments for Office and Road systems as a
percentage of the total cost of individual systems. This
points to the reality that IT is patching much more
frequently than in years past. Increased focus on responsive
patching is a very encouraging development, as is the
increased use of automated patching software. But while
these trends are positive, they skew support costs higher.
One of the points of data that sheds some light on this
matter is the consolidation around Windows XP.
FIGURE 10. In just two years, IT has largely consolidated its client operating systems around Windows XP.
Wipro / Intel PC Refresh Optimization October 2006 15
As Figure 10 shows, since 2004, IT has standardized
on the Windows XP client operating system. While
this has reduced the complexity of managing large
numbers of PCs (as compared to an equal number
running multiple OS types), it has also contributed to
raising the cost of minor software deployment due to
the higher volume of security patches needed to
support Windows XP. What’s more, this higher
volume of patches is being rolled out to a higher
number of laptops. In the previous study, Road
systems had a much higher cost for Help Desk calls
than for minor software deployment. The change
demonstrates IT’s strengthening commitment to
rolling out patches to Road Warriors and thus
minimizing downstream Help Desk costs.
Regarding client systems in general, when asked the
level of agreement with a strategy of deploying patches
at the time of release or on a scheduled cycle
independent of patch release, survey respondents
strongly favored deploying patches immediately. Figure
11 below shows the level of agreement for deploying
patches upon release on a scale of 1 to 7 with 1 being
Strongly Disagree and 7 being Strongly Agree.
Figure 11: Companies are very proactive with regard to
deploying patches as soon as they are released. This bar
graph illustrates the importance that IT survey respondents
place upon deploying security patches at the first
opportunity. Responses are measured on a scale of 1 to 7
with 1 being Strongly Disagree and 7 being Strongly Agree.
“We have worked hard to implement a proactive patching
strategy and automation tools so that we can greatly lower
the risk of a security breach.”
– VP of IT, banking industry
In summary, IT is responding more quickly to patching
requirements for Windows XP, but this increases the cost
associated with minor software deployment within the
overall PC cost.
The Impact of Complexity on IT Responsiveness It’s a truism that more complex environments are more
expensive to maintain, and that they impede the IT
organization’s ability to deliver critical services. This is true
for Desktop, Office and Road systems. Not surprisingly,
Wipro’s analysis validates that the following conclusions:
• Minimizing the number of images for the fleet of PCs
will also reduce complexity and lower support costs.
• Following the refresh cycle recommendations for
Desktop, Office and Road systems outlined above can
make a significant difference in IT responsiveness.
“We make every effort to reduce the number of images
because it lowers expenses when trying to deploy new
software, patches, and OS updates.”
– IT Manager, retail industry
REFRESH CYCLE IMPACT ON SOFTWARE DEPLOYMENT FAILURE RATES
Our analysis has established that IT departments are doing
more software deployments than ever before, and that the
cost of those deployments rises as refresh cycles lengthen.
This is due in large part to the fact that longer refresh cycles
mean markedly higher failure rates for operating systems as
well as for application deployments, both major and minor.
The unfortunate thing is that IT personnel assigned to resolve
Wipro / Intel PC Refresh Optimization October 2006 16
failed software deployments are effectively barred from
engaging in other activities of more strategic importance.
Figure 12: Regardless of the type of client system,
software deployment failures increase rapidly with
longer refresh intervals.
Two important trends are apparent in the data charted in
Figure 12. The first is that OS deployments suffer the
steepest rise in failure frequency, becoming prohibitively
inefficient and expensive for all client types after just 3.5
years. The second is that failure rates for Desktop, Office,
and Road systems are effectively identical. This is interesting
inasmuch as it dispels the widely held notion that client
systems on LANs or enterprise networks have much higher
rates of software deployment success.
THE IMPACT OF REFRESH CYCLE ON HELP DESK OPERATIONS
A consequence of longer refresh cycles that is often
overlooked is the impact of the additional complexity of the
installed base that results as time passes and aging systems
remain in service. This can have a substantial impact on the
IT Help Desk, and it isn’t positive. In fact, this added
complexity can have a double impact on firms, since:
1. IT employees are occupied longer resolving help desk
issues, and
2. Help Desk personnel are busier, a problem that is
exacerbated by the recent trend to downsize IT
departments.
3. There is also a hidden burden borne by Help Desk staff:
Older systems typically run older OS versions, and have
hardware subsystems that require specialized
knowledge—all of which adds to the training,
documentation and knowledge base required.
Wipro / Intel PC Refresh Optimization October 2006 17
FIGURE 13: Help Desk calls rise dramatically at 3.5 years
into a refresh cycle.
In analyzing the Help Desk call volume data collected in
our survey and displayed in Figure 13, two additional facts
become evident:
1. The number of calls from Desktop, Office, and Road
users is strikingly similar and grows at almost the
same rate as the installed base ages.
2. When client systems reach approximately 3.5 years of
age, the volume of Help Desk calls rises dramatically
for all three user populations.
Road system call volume is dramatically affected by the
refresh cycle length, emphasizing the need to proactively
refresh those systems as described earlier in this report. A
single four-year-old Road machine averages nearly eight
calls per user per month—the same as two of the same
machines at two and a half years old. Firms should
therefore see dramatic, measurable decreases in their Help
Desk call volumes as they modernize their Road systems.
ADDITIONAL REFRESH-RELATED SECURITY RISKS
Longer refresh cycles directly increase the time it takes for
firms to successfully test and deploy software patches.
Moreover, every un-patched system increases an
organization’s vulnerability and risk of successful attack, and
many analysts agree that firms have not effectively patched
their networks until more than 95 percent of candidate
systems are covered. Reducing this interval of heightened
risk exposure is a top priority for almost all the IT executives
we spoke with. But the challenges are formidable.
• A four-year Desktop refresh cycle adds an average of
48 minutes of effort per PC per patch when compared
with those on a three-year refresh. For 10,000 systems
this means an additional 8,000 hours per patching
event.
• A four-year Office refresh cycle adds an average of 77
minutes of effort per PC per patch when compared with
systems on a three-year refresh. For 10,000 systems,
this means 12,834 additional hours per patching event.
• A four-year Road refresh cycle adds an average of 156
minutes of effort per PC per patch when compared with
those on a three-year refresh. For 10,000 systems, this
means 26,000 additional hours per patching event.
FIGURE 14. The time required to complete software
security patch deployment increases as refresh cycle
lengths increase. This is due to increased complexity
of the installed base related to longer refresh cycles,
as well as the difficulty of working with older systems.
Additional deployment time translates to additional
security risk exposure.
Wipro / Intel PC Refresh Optimization October 2006 18
The values shown in Figure 14 reflect the average
additional time required to test and successfully deploy
software patches to an installed base of 50,000 PCs. The
efforts measured and included in this display include:
• Patch testing
• Resolution of hardware-related failures during
deployment
Efforts not measured include:
• Initial patch deployment (manual or automated). This
cost is dependent on several IT systems such as patch
management and systems management software that are
not within the scope of this analysis.
Of course the potential liabilities associated with security
breaches will vary by firm, industry, and the nature of the
threat. However, more proactive refresh cycles can have a
direct, dramatic, risk-mitigating effect regardless of whether
you are patching Desktop, Office, or Road systems. By
retiring older, more vulnerable systems, the enterprise
increases the productivity of its IT administrators, who are no
longer saddled with the ongoing, time-consuming need to
apply patches to aging PCs. A security patch on an older
machine can require an end user to surrender their machine
to IT and thus render them without access to corporate
computing resources. This is likely to create a loss of
productivity and can even lead to a business crisis should the
patching event occur at an important moment in the business
cycle such as the end of a quarter for a sales person. The
downstream effects of reduced IT responsiveness due to time
spent patching and repairing older PCs can be far reaching.
Conclusions Data gathered from 30 firms in a variety of industries shows
that, when it comes to PC installed base management,
limiting complexity and optimizing platform refresh cycles
lowers overall ownership expense substantially.
Because the ongoing support and administrative costs for
Desktop, Office and Road systems rise steeply over time,
the cost-optimized lifecycle for each platform is a balance
between initial cost amortization and continuing cost
minimization.
• For Desktop PCs, the optimum refresh interval
is three years
• For Office systems, the optimum refresh interval
is also three years
• For Road systems, the optimum refresh interval
is two years.
IT can save substantial sums of money by optimizing PC
refresh cycles. Beyond cost savings, IT can greatly
minimize security risks and support headaches by retiring
older systems. This study clearly shows that IT
responsiveness to security patching drops because of the
extra time needed to patch older PCs that are well beyond
their optimal refresh point. This trend can lead to a loss of
productivity not only for IT but for the end-user population
it supports.
By implementing a proactive refresh strategy for their PC
fleets, tailored to optimum lifecycle intervals, companies
can greatly reduce their total spending on PC infrastructure,
while improving enterprise operating efficiency and
security. Companies should closely evaluate these finding
and aggressively implement proactive refresh strategies
across their PC environments.
Wipro / Intel PC Refresh Optimization October 2006 19
Appendices APPENDIX A. A GLOSSARY OF TERMS USED IN THIS SURVEY
Office Warrior Employees who use a mobile PC (i.e., a laptop or notebook) and are physically
present in the office for at least 80% of their working hours.
Road Warrior Employees who use a mobile PC (i.e., a laptop or notebook) and are physically
present in the office for at a maximum of 20% of their working hours.
Deployment The process of placing an application in a distributed environment and making
the application available for use. Deployment can include such tasks as
installation, configuration, and administration of various parts of the application.
Desktop/Laptop Mix The proportion of PCs in an enterprise that are desktop models vs. laptop
models, e.g., a 70:30 mix.
Hardware Configuration A collection of PCs that use the same hardware driver stack.
Installed Base The PCs already installed and in use within an organization.
Operational Costs The ongoing costs of managing an installed base of PCs, such as deploying
software updates, answering user queries, fixing broken hardware or software,
etc. Typically, most if not all of these costs are labor-related. They may also be
called support costs or management costs.
Proactive Refresh Cycle A plan to replace PCs on a regular basis, not just when they break or
performance falters. With a three-year proactive refresh cycle, about 1/3 of the
PCs in an installed base would be replaced each year.
Refresh Cycle The planned length of time from when PCs are deployed to when they are
decommissioned and replaced by new PCs.
Software Image The “master copy” of an enterprise’s operating system, tools, drivers, and
applications that is cloned to end-user PCs. There may be one or multiple software
images. “Image engineering” is the creation and testing of a software image.
Total Cost of Ownership (TCO)
TCO accounts for all of the costs associated with procuring, deploying and
owning IT systems. TCO includes purchase, lease and maintenance costs for
hardware and software. It also includes labor costs associated with planning,
purchasing, testing, configuration, deployment, software updates, training and
technical support. Some TCO models, such as Gartner’s, also include end-user
costs (such as downtime and peer support); others do not. TCO models do NOT
take into account the end-user benefits that flow from a technology, such as
increased productivity.
Software Updates Changes to an existing deployed application or operating system, such as
hotfixes, patches, and point releases, as well as updated and new versions.
Wipro / Intel PC Refresh Optimization October 2006 20
APPENDIX B. SELECTED DEMOGRAPHICS OF SURVEY RESPONDENTS
Wipro / Intel PC Refresh Optimization October 2006 21
APPENDIX C. OVERVIEW OF WIPRO PSA OVERVIEW OF WIPRO PRODUCE STRATEGY & ARCHITECTURE PRACTICE The Wipro Product Strategy & Architecture (PSA) Practice is a division of Wipro Technologies, a global technology services
division of Wipro Ltd. (NYSE-WIT). Wipro’s PSA Practice has more than 10 years experience in researching, analyzing, and
documenting the business value of technology solutions. Wipro’s PSA practice helps enterprises and technology vendors develop
innovative and effective product and IT strategies that enable them to expand their market opportunities, extend their competitive
advantage and economize their business operations.
In addition to providing consulting to technology vendors, Wipro’s PSA Practice consultants and technologists work with global
enterprises and service providers in architecting and implementing large-scale systems. This practical, hands-on experience gives
Wipro’s PSA Practice consultants and technical architects first-hand knowledge that informs their business analysis work. For
further information, contact [email protected] or visit http://www.wipro.com/psa_practice.
Wipro / Intel PC Refresh Optimization October 2006 22
For further information, contact [email protected]
or visit http://www.wipro.com/psa
Copyright © Wipro Ltd. 2006. All rights reserved. No portion of this study can be used or reproduced without permission of the
author. For additional reproduction rights and usage information, go to www.wipro.com. Information is based on best available
resources. Opinions reflect judgment at the time and are subject to change.
Other names and brands may be claimed as the property of others.
Printed in USA/1006