400 words 3 hrs
Vincent666
CaseStudy-Butler.pdf
144
CASE STUDY I-3
The VoIP Adoption at Butler University
As CIO Kincaid looked back over the past 12 months, he took pride in the fact that Butler University had taken a major step forward with a network capability for the twenty-first century—the convergence of its data and voice networks. Yet the choices that had been made in the past year were not for the “faint of heart”: his IT organization had taken on a lot more in-house responsibility for voice communications than in the past at a time in which network security risks were heightened concerns. This also meant 24/7 visibility for the telephony staff people who for the first time were working alongside their data network counterparts.
Kincaid wondered: Did we make the right decisions? Have we taken on too much?
IT at Butler University
Butler University is a private liberal arts college in Indianapolis, Indiana. Founded in 1855 by attorney Ovid Butler, the university was comprised of five colleges and 20 buildings on 290 acres, and a student enrollment of 4,400 students in 2005, the year it celebrated its 150th anniversary. More than half of the enrolled students lived on campus and depended on the university for network and telephone services.
Butler’s Information Resources (IR) department (see Exhibit 1) consists of 40 staff members that service the technology needs of not only its students, but also approximately 900 faculty and staff members. The CIO, Scott Kincaid, reports to the president of the university; formerly a CIO at a major financial services firm, Kincaid was the first IT leader at Butler to be given the CIO designation. Reporting to Kincaid are four directors responsible for the university’s IT services: Network and Systems, Administrative Computing, Web Applications Development, and Instructional Technology. Part-time student workers are employed in all four of these areas, with an especially heavy dependence on students for help desk and instructional lab support services.
Campuswide, the IT organization was responsible for supporting over 125 servers, over 1,400 university-owned desktops, and approximately 7,000 network connections.
Since 2001, the university’s administrative systems have included PeopleSoft1 ERP modules for Human Resources, Finance, Campus Solutions (student administration), and Enterprise Portal.
Prior to 2005, Butler had utilized SBC’s2 Centrex serv- ice to provide 3,000 phone stations, with most faculty and staff having basic single-line analog phones. The Centrex system was an outsourced solution: essentially, all call switching was handled by a local telephone provider (SBC), rather than by a university-owned system, and Butler paid for the services on a monthly basis. Over the past decade, the SBC Centrex system had been very reliable, but it lacked more modern functions, such as intelligent call queuing. Voice mail was provided by a Centigram 640 system that provided a single voice mailbox to each dorm room, which meant that students had to share both their phone number and voice-mail box with their roommates.
The outsourcing arrangement enabled the university to avoid the costs of ownership and the day-to-day management of the communications equipment that would be required if it had implemented its own private branch exchange (PBX). While very reliable, the Centrex system was based on features over a decade old, and was not easily customizable. It had therefore become an impediment to implementing customer call center service features, such as advanced call routings. As departments grew and personnel changed, moving phones and phone lines was a labor-intensive process. SBC’s billing system was antiquated, prone to error, and required constant reconciliation efforts by Butler’s telecom coordinator.
Since Centrex was only partially meeting Butler’s needs, Information Resources had begun researching the telephony landscape in early 2004. Many vendors and many trade magazines were predicting that an in-house PBX would be more cost-effective than the outsourced Centrex system for telephone services. Butler could acquire and control its own system and then customize it accordingly. Even more intriguing was a newer alternative that would bring voice communications onto the same
Copyright © 2007 by Carol V. Brown, Associate Professor of IS at the Kelley School of Indiana University–Bloomington and Scott A. Kincaid, CIO, Butler University, Indianapolis, Indiana, with research assistance by Tony Easterlin and John Sacco, MSIS students at Indiana University.
1PeopleSoft was acquired by Oracle in January 2005. 2SBC was formed through the mergers of Ameritech, Southwestern Bell, and Pacific Telesis. SBC purchased AT&T Corp in 2005 and took the name AT&T inc. They then merged with Bellsouth in 2006 and consoli- dated ownership of Cingular, all under the AT&T name.
customizable.
such
prone
partially
Applications
organization
Sy st
em s
A dm
in is
tr at
or
T el
ec om
C oo
rd in
at or
Sr . S
ys te
m s
A dm
in is
tr at
or
N et
w or
k E
ng in
ee r
W eb
D ev
el op
er L
ev el
3
W eb
D ev
el op
er L
ev el
2
W eb
D ev
el op
er L
ev el
1
St ud
en ts
W eb
D ev
el op
er s
Sy st
em s
Su pp
or t
Pr oj
ec t C
oo rd
Sy st
em s
Su pp
or t
Pr oj
ec t C
oo rd
Sy st
em s
Su pp
or t
C on
su lta
nt
Sy st
em s
Su pp
or t
C on
su lta
nt
Sy st
em s
Su pp
or t
T ec
hn ic
ia n
Sy st
em s
Su pp
or t
A na
ly st
-M A
C
Sy st
em s
A na
ly st
Sy st
em s
Su pp
or t
T ec
hn ic
ia n
Su pp
or t S
er vi
ce H
el p
D es
k, L
ab A
ss is
ta nt
s an
d O
pe ra
to rs
H el
p D
es k
Sp ec
ia lis
t
H el
p D
es k
Sp ec
ia lis
t
O ff
ic e
Pr oj
ec t
A dm
in is
tr at
or
A dm
in is
tr at
iv e
Se cr
et ar
y
D ir
ec to
r W
eb A
pp lic
at io
ns
Jo e
In di
an o
D ir
ec to
r N
et w
or k
an d
Sy st
em s
D ir
ec to
r A
dm in
is tr
at iv
e C
om pu
tin g
Sc ot
t K in
ca id
C hi
ef In
fo rm
at io
n O
ff ic
er
D ir
ec to
r In
st ru
ct io
na l
T ec
hn ol
og y
Sr . S
ys te
m s
A na
ly st
Sr . S
ys te
m s
A na
ly st
Sr . S
ys te
m s
A na
ly st
Sy st
em s
A na
ly st
Sy st
em s
A na
ly st
Sr . S
ys te
m s
A na
ly st
St ud
en t A
ss is
ta nt
s T
ec h
T ea
m
M an
ag er
, C lie
nt Se
rv ic
es D
at ab
as e
A dm
in is
tr at
io n
Sr . S
ys te
m s
A na
ly st
D at
ab as
e A
dm in
is tr
at io
n
Sy st
em A
na ly
st /
Se cu
ri ty
A dm
in .
In st
ru ct
io na
l T
ec h
Sp ec
ia lis
t
In st
ru ct
io na
l M
ed ia
S er
vi ce
s Su
pe rv
is or
In st
ru ct
io na
l S ys
Su pp
or t A
na ly
st
In st
ru ct
io na
l T
ec h
T ra
in er
M ed
ia S
er vi
ce s
Sp ec
ia lis
t
St ud
en ts
M ed
ia S
er v.
M ed
ia S
er vi
ce s
Sp ec
ia lis
t
Sy st
em s
A na
ly st
H el
p D
es k
Pr og
ra m
C oo
rd
E X
H IB
IT 1
IT O
rg a n
iz a ti
o n
C h
a rt
a t
B u
tl e r
U n
iv e rs
it y w
it h
K e y P
la y e rs
i n
D e p
lo y m
e n
t o
f N
e w
S y st
e m
H ig
h li g
h te
d
145
network as data communications: a VoIP3 network solu- tion. Like the other alternatives, a VoIP solution appeared to be neither simple nor cheap.
A new in-house PBX would improve the treatment of callers with its Automatic Call Distribution capability, which put callers in a queue if all available staff were busy. Multiline telephone sets, an online caller directory capabil- ity, and better management reports would help to improve telecommunications services throughout the university and provide quicker access to personnel. For students, private voice mailboxes could be assigned with each student hav- ing their own phone number. Many organizations own and manage an in-house PBX system utilizing traditional voice circuits, but newer VoIP solutions was an interesting trend in the communications industry. The potential cost savings from placing voice traffic on the data network—VoIP— caught the eye of Butler’s technology group. However, VoIP was a relatively new technology, and embarking on a marketplace solution that was this new would mean that there wasn’t yet a “clear path to success.”
Only twelve to thirteen percent of the market had VoIP installed in 2004. Even though the telecommu- nications landscape was changing with an emerging trend of increasing IP lines, the current statistics were still daunting and left us wondering if this was a good path to take. Needless to say, we became intrigued.
—Scott Kincaid, CIO
The limited adoption rates weren’t the only risks Butler had to consider. By converging traditional data networks and voice networks into one, the reliability of the voice services on cam- pus would likely be less than what the campus had experi- enced for telephone services in the past: The VoIP phone sets would depend on network operations that weren’t used by tra- ditional analog phones. The Butler data network was known to be prone to occasional outages due to equipment failure, insufficient capacity at times, or odd behavior from rogue student-owned computers. Several constituents were con- cerned about the potential voice quality of IP-based telephony.
Coincidentally, in January 2004, the Information Resources group at Butler had obtained formal approval to begin a three-year program to upgrade all of the network switches, routers, and hubs, a major endeavor. The network upgrade funding would come from two sources: two-thirds would be supplied by university gifts and the remaining third would come from the annual IR operations budget spread over three years. So Butler was on a path to improve the network, but it was not in place now.
The thought of a data network being capable of accommodating high-quality, business-class phone calls was a real source of apprehension. This meant that perhaps the biggest risk that a VoIP PBX solution posed was intro- ducing a less proven technology to its users and gaining their buy-in to this emerging solution. Gaining buy-in would entail accepting the adoption of a new method of communication and new performance risks. Everyone was also worried about virus attacks that commonly affected data networks, but not traditional voice lines.
People were concerned we’d drag down our voice systems to the quality of data network. Our goal, however, was to bring the data network up to the quality of our traditional voice network.
—Joe Indiano, Director Network and Systems
Matching Butler’s Needs with the Network
Alternatives
In order to look at the alternatives, the Telephony Evaluation Team was formed in March 2004 (Exhibit 2). The team was led by Joe Indiano, IR’s director of network and systems. The team included the telecom coordinator, a data network engineer, the university’s Facilities Management telecom technician, and the CIO, Scott Kincaid.
Given that this evaluation was a new project to Butler, but had been done by many other organizations, Joe Indiano made the recommendation to engage an experi- enced consulting firm to coordinate a formal needs analy- sis, including end-user surveys. To find a qualified outside firm without spending months of analysis on that alone, Butler utilized the expertise of a professional trade group (ACUTA4) of which it was a member. Through the ACUTA listserv, they asked peers for recommendations regarding consultants who had helped other similar organ- izations with telephony system evaluations. This quickly identified two primary candidates, who Butler contacted and interviewed.
Butler decided to engage Dietrich Lockard Group, an independent telecommunications consulting firm based in St. Louis, Missouri, to help determine if the investment in a new phone system could improve communication and advance Butler’s mission.
To engage the university community in the telephony needs analysis, a formal User Advisory Group was created to work alongside the consulting firm. This committee included end users of the phone system: administrative staff
146 Part I • Information Technology
4 ACUTA: Association for Communications Technology Professionals in Higher Education.
3Voice over Internet Protocol allows for transmission of voice traffic over an Internet protocol network that is either public (the Internet, for example) or private.
obtained
Case Study I-3 • The VoIP Adoption at Butler University 147
Director, Netwk &
Sys
Telecom Coordinator
CIO
Butler participants Outside participants
Facilities Mgmt Telecom
Tech
Network Engineer
Telecom Evaluation Team
Dietrich Lockard Group
Berbee
User Advisory
Group CFO
Server Admin Systems
Support CISCO
SBC
User Dept. Coordinators
Student Techs
Help Desk
Time Warner Telecom
PeopleSoft developer PR Comm.
Specialist
EXHIBIT 2 Project Participants at Cutover, July 2005
from Admission, Finance, Student Life, Facilities Management, the libraries, and faculty. In June 2004 the IR Telephony Evaluation Team, a Dietrich Lockard Group consultant, and most of the User Advisory Group went to Chicago to attend the SUPERCOMM telephony confer- ence. SUPERCOMM hosted leading-edge technology forums and exhibits and allowed the Butler staff to get a jump-start on industry trends. The team was able to “win- dow shop” to get a feel for what products were available and to help devise an appropriate solution for enhancing telecommunications on the Butler campus. The trip also allowed the User Advisory Group and technical staff to become more acquainted and begin to work as a team. The Butler participants began to feel more comfortable with the possibility of acquiring a PBX and implementing newer VoIP technology as they learned about ways to address some of the risks.
After returning to campus, Dietrich Lockard Group consultants conducted surveys, interviews, and focus
groups with staff, faculty, and students. The team found that many administrative users were displeased with the out- dated Centrex and Centigram capabilities, although they lauded the communication system’s high reliability. For example, all the offices with heavy call volumes had only basic calling trees to help route calls and lacked manage- ment reporting; the analog phones could only handle one call at a time. Moves, additions, and changes to current phones required a technician to physically visit each phone. Like many other universities, Butler also lacked any means to communicate to the entire campus if time-sensitive emergency information needed to be disseminated.
The student surveys revealed that nearly 92 percent of students residing on campus had a cellular phone, but virtually the same number of students used their regular room phone as well. While almost two-thirds of the students preferred cell phones, one-third still preferred a regular telephone. Moreover, 73 percent of the students indicated that their campus-provided voice mail was
148 Part I • Information Technology
important and they clearly wanted a private voice mail- box. These results were somewhat of a surprise to the Telephony Evaluation Team and the User Advisory Group, given the visible penetration of cell phones on campus. It meant that cell phones, with their limited minutes and sometimes spotty coverage, did not meet all of the needs of the typical residential student.
Conventional wisdom in the cellular age would be that traditional land line phones are not important. However, facts don’t always line up with intuition. We discovered that our peer institutions were continuing to provide regular phone service in dorm rooms, and therefore it was a no-brainer to continue to provide it as well. I would have bet money that the phones could have been pulled out, so it was a great surprise to me.
—Scott Kincaid, CIO
From all of the needs analysis discussions, a long list of needs and potential opportunities had been identified. In June 2004, based on the analysis of the data collected, a list of key criteria for a new system was developed, with three priority groupings: must haves, very important, and nice- to-haves (see Exhibit 3). Joe Indiano and Dorothy Lockard (president of the Dietrich Lockard Group) then used these needs analysis results to help pinpoint the best opportuni- ties within the context of Butler’s strategic plan, and five strategic goals were established:
• Improve communications with the students and within the student community, including the allot- ment of private voice mailboxes
• Improve the handling of callers to high-volume call areas on campus
• Leverage newer services such as multiline and self- labeling telephone devices and online directories as well as improve training
EXHIBIT 3 Key Criteria for New System (based on focus groups, surveys, and consultant interviews)
Critical Issues (from master list)
Improved handling of calls to high volume offices M
Performance reporting to manage calling to high volume offices M
Improved handling of callers to operator position M
Emergency alert campus-wide M
Position for future services/technology & changing expectations M
Improved training for end-users M
Communication with students using their preferred medium (e.g., IM) M
VM privacy for students in administratively efficient method I
Indication of message waiting to student I
Ensure donors reach Advancement anywhere, anytime I
Ensure campus police can locate key administrators in an emergency I
Standard features and functionality that improve efficiency and are common at many institutions (at minimal recurring cost)
I
Telephone instruments that reduce need for training and relabeling I
Long delay in 911 cell call response, not routed to campus police I
Administratively efficient means of assigning a single phone number for a student while at Butler I
Enable faculty/staff to make and receive calls using Butler identity (i.e., Butler athletics recruiters, advancement, faculty at-home.)
I
Easy to use video conferencing I
Conference calling for up to 6 parties at no extra cost N
Efficient means to integrate VM with E-mail N
Larger conference calls, ad hoc and meet-me N
Cell phone coverage inadequate N
Time saving system administration tools N
M = Must Have, I = Very Important, N = Nice to Have
strategic
Case Study I-3 • The VoIP Adoption at Butler University 149
5Lightweight Directory Access Protocol provides an online, fully- indexed telephone directory service developed and freely distributed by the Regents of the University of Michigan. [Eudora]
• Provide more immediate access to specific Butler personnel
• Remain competitive with peer institutions in the level of services offered, particularly those that impact students and their families
As part of the business case, cost estimates for vari- ous alternative telephony systems were determined. These cost estimates incorporated the investments in hardware, software, phone equipment, and on going expenditures, based on projections of what each option would cost given industry research on pricing and the effort it would take to support each option ongoing.
Based on the cost-benefit estimates, the decision was made to consider four alternative scenarios:
1. Continuing with the current solution: Centrex serv- ice outsourced to SBC
2. Continuing with the outsourced Centrex service— but with a significant investment in upgrades and new “bolt-on” equipment
3. Acquiring an in-house PBX system, using either tra- ditional equipment or VoIP
4. Continue with the current solution (Option 1)—but adding an independent 50-seat VoIP system for a few selected offices
With Butler’s needs analysis completed, the results were presented to the university vice presidents in July 2004. Butler management then gave the team the go-ahead to solicit bids from vendors, but was nervous about the potential investment.
Request for Proposal to the Selected Vendors
Based on the strategic goals and the key criteria developed by the User Advisory Group, the Telephony Evaluation Team and the consultants crafted a detailed request for pro- posal (RFP). Dietrich Lockard Group identified vendors that appeared to be well-aligned with the needs of Butler. Dietrich Lockard Group contacted the six vendors and identified the best local implementation partners (manu- facturer resellers) for each of the manufacturers. Four of the vendors, through the local resellers, submitted bids. SBC presented Butler with three bids: continue the out- sourced Centrex service option, add equipment to the serv- ice, or move to an in-house PBX using Nortel equipment.
Then the User Advisory Group and the Telephony Evaluation Team began the journey of listening to vendor presentations, interviewing vendor representatives, evalu- ating all of the features, making visits to selected vendor customers, and striving to do an apples-to-apples compari- son of the various proposals. The number of features and factors considered was extensive with 400 different data
points provided in the RFP. To summarize the criteria, a three-level priority scheme (shown in Exhibit 3) would be used as a weighting scale to help ensure that the selected alternative would best address Butler’s critical issues.
Butler’s main criteria were total overall cost and meeting the needs defined in the RFP (as summarized in Exhibit 3), and most venders could meet Butler’s require- ments. However, during the comparative analysis, issues bubbled up as key differentiators. Local support from an experienced reseller was proving to be imperative if Butler was going to move all of its telephony in-house. Experience with new VoIP systems and integrating other enterprise systems (i.e., PeopleSoft, LDAP5) would be cru- cial as well. With Macintosh users representing 25 percent of the faculty population at Butler, the university also wanted a system that offered equivalent features, such as video conferencing capabilities, for both Mac and PC users. An effective and flexible integrated emergency noti- fication system, which allowed for a simultaneous broad- cast of text and audio to all campus phones, as well as other data network security concerns were other key differ- entiators reflecting their increased priorities in a post-9/11, Internet era.
In addition, what the manufacturers were planning to invest in R&D were also considered. Most telephony ven- dors currently sold both IP-based PBX systems and also tra- ditional analog systems, with several promoting the ability to interconnect the two types of systems. But during last- minute visits to several of the manufacturers’ headquarters, it became obvious to the Telephony Evaluation Team that the manufacturers were putting all of their R&D dollars into IP-based systems. While VoIP was certainly the more risky approach, the Telephony Evaluation Team came to the uncomfortable realization that if they invested in a new ana- log PBX, they might end up with a system that had a limited life. Suddenly, staying with the SBC Centrex service for a couple more years looked like a serious option.
Vendor Selection
Butler narrowed the field down to five main candidates from a list of nine potential options presented to them.
The final alternatives included the following:
• Continue with SBC’s current Centrex service (out- sourcing option)
• Continue with SBC’s Centrex service, but add a Nortel 50-seat IP system
strategic criteria
criteria
150 Part I • Information Technology
• Implement a new system to support VoIP PBX in- house - Cisco (represented by Berbee Information
Networks) - Mitel (represented by MVD Communications) - Nortel (represented by SBC)
Side-by-side comparisons were made of what each vendor solution offered in regard to each critical issue, and what additional capital and/or ongoing support costs would be incurred to satisfy these requirements.
Following the review of this range of options, and the scorings of the options, the Telephony Evaluation Team, working in conjunction with the User Advisory Group, chose option 3: to acquire its own in-house system and to acquire a Cisco IPT system, with products and implemen- tation services provided by a Midwest Cisco partner, Berbee Information Networks. Joe Indiano and Scott Kincaid presented this to the Butler senior management, and after careful consideration, obtained the go-ahead to proceed. Butler informed Berbee and Cisco they had won the bid on the day before the Christmas holidays began.
Contrary to what was seen in the popular press, the team’s analysis did not show that an in-house IP-based sys- tem would be cheaper than the old Centrex solution. However, most of the needs identified by staff, faculty, and students would be addressed with only a minimal increase in costs. For example, each high-volume call area on cam- pus would now have Automatic Call Distribution capabili- ties. Butler would be gaining a system to issue campuswide emergency alerts via the new IP-based phones. Additionally, each student who previously shared a line in a residence hall room would now have their own unique phone number and private voice mailbox, which also moved with them as they changed rooms from year to year. (Changes to room assign- ments would be entered into PeopleSoft and an automated interface would carry the change into the telephony system.) When students received voice mail, they would also receive an e-mail notifying them of the message.
The strategic objective was to support IR’s mission of providing good services and communication between students and faculty. Cisco’s system fulfilled the identified needs for voice communication, but Butler was concerned about more than just equipment specs and cost.
During the sales cycle, vendors usually promise all types of grand improvements and outcomes. But the written contracts they provide promise only to deliver specific pieces of hardware and software, with very limited warranties, and customers often spend most of their time negotiating on price. In this project, we were able to successfully negotiate for the original
RFP and the vendor’s proposal to Butler to become part of the final contract. This kept everyone focused on delivering a fully integrated system, instead of just focusing on the hardware, and held the vendor accountable to the claims in their proposal.”
—Dorothy Lockard, President, Dietrich Lockard Group.
Furthermore, what Butler was ultimately implementing was an integrated set of software applications that operate on the IP network, not just a “black box” phone system. The understanding of the IP protocol and how it would fit into Butler was, in the end, second to how the multiple components of the system were to work together in a secure and quality fashion. Additionally, identifying who within Butler owned each piece, how the data would flow from point-to-point, and where data was being replicated were all important questions that needed to be answered for this application system. Creating seamless integration with other systems such as PeopleSoft, the housing sys- tem, and Butler’s LDAP directory was one of the values of the newer software-based systems. Essentially, there were multiple interfaces to existing systems, more options, and a vast array of decisions to make about how all of these fea- tures were going to work in unison.
Thus, when choosing the primary vendor, the imple- mentation support provided by a Cisco partner was extremely important.
Since this was really an application system, we were more concerned with who was going to help us deploy the system. Of course, the manufacturer is important as we are using their equipment, but even more important is someone with large-scale applica- tion implementation experience.
—Scott Kincaid, CIO
The new Cisco system configuration included the following investments:
• Cisco Call Manager with multiple forms of redun- dancy
• 1200 IPT stations for faculty and staff • 1800 Analog ports via VG248 gateways for student
phones • Cisco Unity Voice mail with mailboxes for each student • Cisco IPCC Express for six call centers • Cisco Emergency Responder for campuswide E911
capability • Berbee InformaCast alert system • ISI Infortel call accounting
Case Study I-3 • The VoIP Adoption at Butler University 151
The IR team considered leasing the new telephony system, but after analysis chose to purchase and finance it in-house by redeploying the monthly fees they had been paying for the Centrex solution.
Butler’s business model for its original data network replacement investment was a three-year funding plan. However, given the change from an outsourced telephony solution to an in-house solution that depended on the data network, there would have been a lot of risks if Butler con- tinued on the three-year implementation plan for the data network. Instead, Butler decided to compress its three-year network replacement project into one year, using the bulk of their allotted funds at the beginning of the project to reduce the total implementation time.
After the new telephony system was chosen, Butler had to select a vendor to supply local voice circuits. Another RFP had to be initiated for the circuits, and the decision was made to change to a new voice carrier, Time Warner Telecom. The university negotiated the installation of two independent communication links to the campus, and Berbee helped split the new telephony servers between two campus locations—a capability not possible with traditional PBX systems. A total of 17 Windows servers ranging from Call Managers, to an E911 Emergency Responder, to a Call Accounting Database server, were implemented between the two onsite locations to provide backup and disaster recovery capabilities. The Berbee InformaCast software would make it possible for university-wide alerts to be made in audio and/or text, simultaneously or by zone, within seconds, to all IP phones.
The new focus became deploying a single network that would continue to provide reliable service for voice as well as provide more reliable services for traditional com- puter applications. Butler already had the necessary fiber backbone in place but needed to replace selected old wiring with new Ethernet cable. Improvements in the new con- verged network, such as installing an uninterruptible power supply (UPS6) in all network closets and replacing old wiring, would now benefit the new telephony system as well as aid in the support of the traditional data network services.
Contract work began during the holiday break, and the contracts were signed in mid-January 2005. Additionally, the department accelerated the pace on the network replacement project and began to implement new routers and switches across campus. To mitigate the relia- bility risks of the new system, the new network design included more segmentation, protections such as intrusion detection, and quality of service (QoS7) levels. With functions such as QoS levels—new for Butler—the network
would be able to logically separate traffic on the converged network to protect voice transmissions from other traffic, including malicious activity. New monitoring tools would be incorporated in the VoIP system as well to observe where calls were originating and going, and to measure the quality of calls placed through the university. This would help the IR group pinpoint where problems were occurring and help them manage the network in order to provide a level of quality for voice communications similar to what was traditionally provided via old-style analog lines.
Implementing the VoIP System
The number of people formally involved in the project quickly began to increase. Berbee assigned a full-time project manager to the Butler project in January 2005. The director of network and systems, Joe Indiano, would serve as Butler project coordinator and play a key role by ensur- ing integration with other existing systems. The CIO would serve as project sponsor and had the role to elimi- nate roadblocks within the organization and to facilitate communication between key stakeholders in the university, such as the deans and other vice presidents. The CIO would also need to communicate the strategic vision of the project and address any concerns that staff, faculty, and student users had with regard to the new system.
The newly formed IP Implementation Team included other key IR personnel as well—telephony staff, network engineers, PeopleSoft developers, and help desk staff, as well as Butler’s Facilities Management staff to help with the logistics of the installation. A majority of the IP Implementation Team members were from the implementa- tion partner, Berbee Information Networks with Cisco staff helping as needed. Additionally, Berbee Information Networks’s security practice was engaged to ensure that the new network design would allow data and voice traffic to coexist but with telephony servers that were protected from student desktops and outside security threats. While not anticipated at the beginning of the project, it was also decided to keep the Dietrich Lockard Group consultant on the team to ensure that Berbee and Cisco implemented a system with all the items defined by the RFP and vendor contracts. Lastly, some 18 students were used to help with the testing and rollout.
Through this implementation, not only were data and voice network technologies converging, but also the former data and telecommunications personnel that had previously worked in separate units were merging:
You have to keep in mind that convergence within the project not only refers to the convergence of technologies but also the integration of processes and people as well. When you have voice and data
6 Uninterruptible Power Supply provides a backup power supply for a cable voice adapter and the telephone attached to it. [Cisco] 7 Quality of service is the set of techniques to manage network resources. [Cisco]
152 Part I • Information Technology
traffic merging, the respective personnel also merge. Data people approach projects differently than telephony people and use different terminology. So managing this merging of the staff and disciplines was essential. Along with the size of the implemen- tation team, there were a lot of people to coordinate to pull this off.
—Scott Kincaid, CIO
Butler’s “Infinitely Personal” Communication Plan
A communications coordinator was added to the team to help establish the message that the university would send to all users to help ensure acceptance of the new system. Managing the fear factor of a major system overhaul was viewed as crit- ical, because Butler’s users had experienced data network failures before. One of the techniques used for this was to
redefine “IP” to stand for “infinitely personal.” This set the tone for the teams to focus on the ways the network changes were empowering people to communicate more freely, not on the fact the calls were being transported via IP data packets.
We chose an IP system not because it was VoIP, but because, overall, it could provide the best set of serv- ices to meet our needs.
—Joe Indiano, Director Network and Systems
Reassuring users that a personal approach was being taken helped communicate that this was a decision to help everyone, not just to improve the technology of the univer- sity. The IP—infinitely personal—communications plan sought to educate users by first raising awareness and then holding training forums (see Exhibit 4A). Communicating
What Who/Target Purpose When Type/Method(s)
New Phone and Voice Mail System Teaser
Bluemail and my.butler users
Raise awareness May 31–June 27
On Bluemail, IR labs and my.butler homepages
New Phone and Voice Mail System Teaser
All Campus Raise awareness that a new phone system is coming
May 26 Email
Sign up for training and when and where
Fac/Staff on campus now
Get people to go to training May 27 Email
Training Dates and times on web
All Campus Make training information available at all times in a central location
May 27 butler.edu/ir/ documentation/ telecom.html
New Phone and Voice Mail System & Training
Fac/Staff on campus now
Raise awareness and get people to training
June 1 All campus Voice mail
Ask for support Scott to Cabinet Ask VPs/Deans to encourage their staffs to attend training.
June 1 Email
Coordinator verification of phones in their areas
Coordinators Ask coordinators to verify the phones in their areas.
June 3 Email
Early notification of cut dates.
Pilot team and coordinators
Give early notice of anticipated dates for phone system cut to groups.
June 6 Email
Emergency Broadcast System info
Greek house corps and presidents
Make Greek organizations aware of the opportunity to receive emergency information from Butler
June 7 Email
Messages to communicate in training
Students in training
Communicate important system information in person
June 13 in training class
Verbal and Print
Last reminder about training
All campus Final reminder about going to training
June 15 All Campus Voice Mail
Phone delivery Fac/Staff Announce phone delivery and training reminder
June 15 Fac/Staff Alert
EXHIBIT 4A Communication Plan: Sample Page
New Phone System Project Communication Plan
Case Study I-3 • The VoIP Adoption at Butler University 153
important system information to users in person via departmental staff and management reminders for attend- ing training sessions were emphasized. Although much of the communication was done via e-mail, other media included all-campus voice mails and printed materials distributed throughout the campus.
Messages sent out to the university community included why a new system was being implemented, where and when training would occur, how to transition from the old voice mail to the newer version, who to call with problems, the date and time of cutover, and how to return old phones after users began using their new equipment (see Exhibit 4B).
Pilot Program
Before a full-scale dissemination of the new VoIP phones, a pilot program was implemented to help test features and gain user feedback. In April 2005, approximately six weeks before the system was to “go live,” 40 phones were issued to high-volume phone users, such as department coordinators and high-level secretarial staff. Over roughly three weeks, the users helped to test the various features of the new system and to fine-tune any other aspects that pilot users had noted. The overall goal of the pilot program was to gain buy-in from the user community. Ideally, pilot users would not only share their enthusiasm for the new
system with their coworkers, but would also become a co- ordinator and trainer within their department.
IR needed input from the people who would be using the phones the most. They wanted someone from Admission involved because of our large call vol- ume. Since I manage the team that handles incoming calls, it made sense for me to volunteer to be a phone coordinator.
—Susie Bremen, Senior Associate Director of Admission
The pilot program began with an information ses- sion, followed by a series of training sessions that took place in three labs and were run by a firm that Berbee had contracted with for the training. Members from the implementation team and consultants from Dietrich Lockard and Berbee were on hand to facilitate informa- tion sharing and initial training. Pilot users were pro- vided with the opportunity to try various features of the new phone and interact with the new capabilities through role-playing.
They made the training fun. They had gift bags and giveaways, and encouraged questions and discussion from the very beginning. They started out with a
EXHIBIT 4B Example of Communication Message to Staff
overall
154 Part I • Information Technology
PowerPoint presentation to explain terminology, benefits, and how the new technology was going to improve our departments. The training sessions became very interactive. . . . The IR team sent out information regularly—leaflets with timelines, cards with contacts, and worksheets to complete after test- ing various features available with the new phones. We tested such features as “parking” calls, forward- ing calls, and setting up conference calls, and we completed Phone Feedback Forms after experiment- ing with newly-learned features.
—Susie Bremen, Senior Associate Director of Admission
The pilot program helped to create a type of “phone envy” among some users; some of those without IP phones had a chance to see how they were used and became intrigued with the new features of the technology. By bringing the technology closer to the users, it also brought home the message of being “infinitely personal.”
While many users were excited about the new phones and their functionality, others did not fully under- stand how the IP phone would actually work and were apprehensive about the change:
It was a smooth transition for me [but] it was hard for a lot of people to comprehend; they didn’t under- stand the concept of running the phone through a computer connection instead of a telephone jack.
—Kristine Butz, Associate Director of Financial Aid
I already had exposure to call routing and advanced phone systems. There is a learning curve. Some peo- ple were apprehensive because the phone was com- plicated, and they were also afraid that the system would crash and service would not be available.
—Kathy Harter, Senior Assistant Director of Operations
Managers of high call-volume offices were aware that the capabilities of the new system would greatly improve their department’s call center processes, as the functionality allowed for detailed reporting and could help department managers analyze and benchmark their call center employees. But these capabilities also contributed to additional concerns by users:
Some of the functionality was actually scaring some users; they feared they would be monitored because
we would have the capability to evaluate call volume and length. It gave some users the notion that “Big Brother” would be watching.
—Susie Bremen, Senior Associate Director of Admission
The Cutover to the New System
With the pilot behind them, the implementation team felt confident in their ability to move forward with the full roll- out planned for June 2005. Butler’s move from traditional telephony to an IP-based system meant they didn’t have to do a traditional cutover where the old phones are pulled and new ones installed at the exact same time. Instead, the new IP-based phones were distributed to all users during the weeks before the cutover, training sessions were arranged for all faculty and staff, and yet the old analog phones were still available. Even users at the new call cen- ters were able to make and receive internal calls on their new IP phones while the old Centrex phones handled all external calls prior to the cutover.
However, a variety of technical problems threatened to derail the roll-out plan—and later became known as a “Week from Hell”:
• The supposedly simple “plug and play” process of connecting the new IP-based phones failed due to a software bug; all multiline phones had to, unexpect- edly, be manually registered one-by-one using their unique machine address code (MAC).
• The new telephony servers were rebooting haphaz- ardly, and the vendor decided to replace and rebuild every single unit.
• After a two-month effort, 2,000 extra phone numbers, which Butler needed to provide private numbers and voice-mail boxes, were found to be missing—and SBC claimed it would take 45 days to resolve the issue.
• Both the Food Service and Bookstore offices, which are independent non-Butler companies but located on campus, were found to have no Ethernet wiring in their areas.
To further complicate matters, the SBC service rep- resentative whom Butler had come to rely upon for the past four years suddenly resigned, and the IR telecom coordina- tor’s husband was robbed at gunpoint, bound and gagged with duct tape!
Our week from hell was a huge setback . . . but I didn’t want this project to be delayed, and neither did any of the team.
—Joe Indiano, Director Network and Systems
helped
technical
concerns
under-
Case Study I-3 • The VoIP Adoption at Butler University 155
1/04 2/04 3/04
Began systematic plan to replace data network over three years
IR Telephony Evaluation Team formed
User Advisory Group formed
User Advisory Group & IR Telephony Evaluation Team attended SUPERCOMM telephony conference in Chicago Issued formal RFP and had
meetings with various vendors
Vendors scored on features, vendors, and implementation partners; Evaluation of finances and options with CFO
Implementation planning begins with Berbee; System build begins; Hardware ordered and installation begins
User pilot conducted with 40 phones over 3 weeks
Delivery of phone sets; Decided to delay implementation cutover date
Miscellaneous clean-up as well as 2nd training session for faculty and staff
Hired Dietrich Lockard Group to assist in needs analysis including student surveys
Needs Analysis complete; results presented to management and RFP started
Vendor presentations along with customer site visits for Mitel, Nortel, Cisco
Decide to go with Cisco and Berbee and start developing contract based on RFP
Contracts signed; Revised network infrastructure replacement plan; Berbee assigns Project Manager; IP Implementation Team formed
Completed network replacement and wiring upgrades; Training sessions offered to faculty and staff; Week from hell occurs June 13th–20th
July 11th actual cutover to new system
4/04 5/04 6/04 7/04 8/04 9/04 10/04 11/04 12/04 1/05 2/05 3/05 4/05 5/05 6/05 7/05 8/05 9/05
EXHIBIT 5 Project Milestones: January 2004–August 2005
One of the options allowed for in the Berbee contract was for Butler to alter the implementation plan to allow for a split roll-out: phase one of the roll-out could be done in the summer to make sure things went according to plan, and then the phase two users would be implemented during the Thanksgiving holiday break. However, the team decided to push through and not let the unexpected setbacks create more damage than necessary, and the decision was made to delay the cutover date by just two weeks.
We decided not to split the roll-out since everyone was motivated to do it all at once. We didn’t want to kill that spirit. Also, we felt that creating more work during Thanksgiving time would place even more of a burden on our team members and the vendors.
—Scott Kincaid, CIO
On July 11 the system went live! SBC migrated all Butler’s phone numbers to Time Warner Telecom, who then routed all calls to the new system, and Centrex was turned off. Calls got through to the right places and the new call center software managed the calls. Only a few
lines and features were missed, so the planning, testing and roll-out was clearly valuable. However, none of the departmental burglar alarms worked! The alarm panels depended on analog phone lines to communicate with the campus police department and were now being con- nected via a Cisco analog-to-VoIP gateway device. However, the alarm panels, based on 1970’s technology, would not work even with the analog-to-VoIP gateway; apparently Butler had not tested the devices before implementation. So Butler found some unused Centrex lines, and the alarms were reconnected using traditional analog circuits.
Otherwise, the newly converged, in-house voice- and-data system was fully functional. No one reported any problems with the voice quality.
Having been a part of various large system changes in the past, I have seen my fair share of “blips” and overall catastrophes and I was pleasantly surprised that I have seen no such issues from this transition, at least not yet.”
—Kathy Parsons, University Budget Director
push
156 Part I • Information Technology
Post-Implementation Reality
Altogether, the new system took approximately one year from the initial planning process to the final implementa- tion (see Exhibit 5) and included a $1.5 million investment by Butler. The new system gave Butler a robust “soft- PBX” capability based on a series of integrated servers linked to PeopleSoft, LDAP, and the network. This also resulted in six full-featured call centers allowing students and outside callers to get better help regarding such items as student accounts, admissions, financial aid, and box office tickets. When departments moved to different offices, they could simply take their IP phone with them and the number was automatically redirected, without IR intervention. Additionally, faculty and staff were able to utilize a directory of campus phone numbers (available on
the telephone’s display area) that was updated daily by the PeopleSoft system.
However, the new IP system was not cheaper and, in fact, was actually somewhat more expensive as IR had to add a full-time staff member to manage the new system. The new system involved many more components than existed with the outsourced Centrex system. In addition, the IR group at Butler now has the responsibility of maintaining networks not just for data, but also voice.
Soon after the euphoria of a successful initial imple- mentation ended, CIO Kincaid was wondering, “Did we make the right choices?” In their quest to improve commu- nications, should Butler have taken on the responsibility of managing an integrated voice and data network, given the current security threats to computer networks?
cheaper
