Return to the Integrated Systems Project Home Page

 

Integrated

Systems

Task Force

 

 

Study Group #3

Final Report

November 27, 1996

 

 

Index

I. Executive Summary

A. Mission Statement
B. Cost Matrix
C. High Level Assumptions

II. Background

A. Methodology
B. Membership
C. Significant Contributors

III. Appendices

A. Cost Comparison
B. Explanation of Costs
C. Best of Breed Costs
D. Single Vendor Costs
E. Assumptions
F. Conversations with Other Schools
G. Questionnaire Responses

IV. Diagrams

A. Interfaces Needed for Best of Breed Replacement
B. Interfaces Needed for Integrated Replacement

V. Glossary

I. Executive Summary

A. Mission
The mission of this group is to estimate the implementation and on-going costs of a best-of-breed approach and a single-vendor-partnership approach to replacing our current applications with an integrated set of systems.

B. Cost Matrix (from Appendix A)

 

       

    Best of Breed

      Integrated  

Acquisition and Implementation    

$16,396,000

$15,267,000

Costs for 5 Years

    $37,262,000

   $26,604,000

5-Year Cost

    $53,658,000

   $41,871,000

 

C. High Level Assumptions

 

  • Full replacement of 3 core systems categories with Client/Server technology.

Student

Financial

HR

Financial Aid

Purchasing

Payroll

Student Accounts

GL/AP/AR

Benefits

Grade Records

  • 5 Year Horizon
  • Interfaces for best of breed and integration for single vendor
  • Similar "look and feel" across applications for input and output
  • Shadow systems will be similarly affected by either best of breed versus integrated
  • Client Workstations will be similar for either approach
  • Customization will be limited for either approach

    (For additional information on Assumptions, See Appendix E.)

 

II. Background

A. Methodology
To attain our goals, the study group garnered information through various means:
  • Sent an informal request for information to 22 firms known to market integrated and/or best-of-breed administrative applications. Seven firms responded to our request with information related to software and potential customization costs, hardware requirements, recommended training, and projected on-going costs. Not all respondents provided all requested costing information.

 

  • Interviewed individuals at 14 peer universities throughout the U.S. and Canada by phone/fax concerning their institution's experience related to recent implementation of administrative applications. An exhaustive list of questions/discussion points was compiled by the study group as a discussion guide while conducting the interviews. A synopsis of the interview findings is enclosed in Appendix F. A synopsis of questionnaire responses is enclosed in Appendix G.

 

  • Tapped the experiential knowledge acquired at the UVA Medical Center during their recent (and ongoing) implementation of PeopleSoft financial applications.

 

  • Attended multiple vendor presentations that provided direct exposure to the technical architectures and types of functionality available from leading vendors. The question-and-answer sessions provided opportunities to clarify equivocal issues that might have otherwise remain unresolved. Those firms conducting presentations included: Buzzeo, PeopleSoft, SAP, and SCT.

 

  • Conducted site visits to James Madison and Virginia Tech, universities at varying stages of implementation of Financial, Human Resource, and Student systems developed by PeopleSoft and SCT respectively.


Vendor responses to the University's informal request for information varied substantially in quality and quantity of information. Some vendors related a pricing structure that included a flat rate for each functional component, other vendors included a graduated scale with positioning determined by the relative amount of user interaction with the software, and still others used a combination of these pricing structures. To overcome these problems, the study group normalized the software costs provided by each vendor with regard to the overall functionality delivered and the pricing structure employed. Vendors were contacted for clarification as required.

The derivation of estimated costs, other than initial software acquisition costs, was accomplished using a combination of information from trade articles, cost information from other universities, knowledge about current UVA environment, and the group's experience. A detailed description of the cost derivation is provided in Appendix B.

We then performed reasonableness checks of our figures against information obtained from other universities and vendors.

 

B. Membership

Barbara Deily, Leader
Shelley Payne, Facilitator
Peg Howard, Scribe
Virginia Bergland
Bernie Hill
Brad Holland
Pat Lozier
Shirley Payne
Robbbyrda Preston

C. Significant Contributors

Ellie Withers
Belinda Stevens

III. Appendices

A. Cost Comparison

Cost Comparison

Best-of-Breed

Single-Vendor

Acquisition and Implementation Costs:

Out-of-pocket Costs:

1. Purchased Application Software

1,321,000

1,882,000

2. Vendor Consulting/Customization

2,937,000

2,937,000

3. Hardware

1,688,000

1,688,000

4. Vendor-Provided Training

100,000

50,000

Note 1

Committed Costs:

5. Technical Staff Training

306,000

102,000

6. User Training

1,080,000

540,000

7. Technical Staff Time

4,464,000

4,018,000

8. User Staff Time

4,500,000

4,050,000

Total One-Time Costs:

16,396,000

15,267,000

On-going Costs:

Year 1

Out-of-pocket Costs:

9. Hardware

513,000

513,000

10. Software

1,137,000

259,000

Committed Costs:

11. Technical Infrastructure Costs

4,612,000

3,374,000

Note 2

12. User Personnel Costs - Support

1,048,000

1,048,000

Note 2

13. User Personnel Costs - Training

54,000

27,000

Year 1 Total

7,364,000

5,221,000

Year 2

Out-of-pocket Costs:

9. Hardware

513,000

513,000

10. Software

1,177,000

331,000

Committed Costs:

11. Technical Infrastructure Costs

4,612,000

3,374,000

Note 2

12. User Personnel Costs - Support

1,048,000

1,048,000

Note 2

13. User Personnel Costs - Training

54,000

27,000

Year 2 Total

7,404,000

5,293,000

Year 3

Out-of-pocket Costs:

9. Hardware

513,000

513,000

10. Software

1,221,000

364,000

Committed Costs:

11. Technical Infrastructure Costs

4,612,000

3,374,000

Note 2

12. User Personnel Costs - Support

1,048,000

1,048,000

Note 2

13. User Personnel Costs - Training

54,000

27,000

Year 3 Total

7,448,000

5,326,000

Year 4

Out-of-pocket Costs:

9. Hardware

513,000

513,000

10. Software

1,269,000

400,000

Committed Costs:

11. Technical Infrastructure Costs

4,612,000

3,374,000

Note 2

12. User Personnel Costs - Support

1,048,000

1,048,000

Note 2

13. User Personnel Costs - Training

54,000

27,000

Year 4 Total

7,496,000

5,362,000

Year 5

Out-of-pocket Costs:

9. Hardware

513,000

513,000

10. Software

1,323,000

440,000

Committed Costs:

11. Technical Infrastructure Costs

4,612,000

3,374,000

Note 2

12. User Personnel Costs - Support

1,048,000

1,048,000

Note 2

13. User Personnel Costs - Training

54,000

27,000

Year 5 Total

7,550,000

5,402,000

Total 5-year On-going Costs:

37,262,000

26,604,000

Total 5-year Costs:

53,658,000

41,871,000

For explanations on how cost figures were derived, see Appendix B.

Note 1: Some vendors include training in their software cost; others break it out as a separate cost.

Note 2: This cost includes staff time and department overhead.

B. Explanation of Costs

General
For cost items pertaining specifically to the UVA environment, we developed our own formulas based on past experience, number of system users, labor costs, etc. We then checked them for reasonableness against comparable costs reported by other institutions.

The vendor-related costs in the Best-of-Breed column are the sum of our estimated costs for:

  • Geac (formerly DBS) Financial System
  • Geac Human Resources System
  • ISIS, the University's in-house Student System

(See Appendix C)

The vendor-related costs in the Single-Vendor column are the average of our estimated costs for:

  • Buzzeo
  • Datatel
  • PeopleSoft (using costs provided by the vendor)
  • JMU's costs to implement PeopleSoft's suite

(See Appendix D)

All numbers are rounded to the nearest thousand. Numbers in this document that also appear on the Cost Comparison sheet are highlighted.

We are showing on-going costs as if all systems were implemented prior to Year 1. A schedule for replacement has not been developed so we could not allocate our costs accordingly. Therefore the timing of on-going costs may differ from that shown in the chart.

Out-of-Pocket Costs for Software Acquisition and Implementation

1. Purchased Application Software
Best-of-Breed: (See Appendix C)
The software costs for the two Geac systems were provided by the vendor. However, Geac's pricing structure consists of a base amount and a per-user amount. If we were to purchase both the Financial and the Human Resources systems at the same time, we would pay the base price for both systems, but only one per-user cost.

The costs for ISIS were based on the 1993 ITC document "Cost of Recent Application Development Projects".

The cost of the Peoplesoft implementation at the Health Sciences Center was used for comparison purposes only and was not included in our calculation.

Single-Vendor: (See Appendix D)

The software costs were provided by the three vendors and JMU.

2. Vendor Consulting and Customization

The vendors we surveyed could not give us costs for this item because it varies so much from one client to the next. Therefore we developed our own assumptions and formula.

We assumed that over the course of an 18-month implementation, we would use the services of three FTE for nine months (this might actually be more than three part-time consultants, working varying hours throughout the eighteen months but totaling 4,500 hours over the implementation period).

We estimated an average hourly rate of $200 per person. The total charged by the vendor, then, would be $900,000.

Travel and lodging expenses are an additional cost to us. We estimated that the number of person-days for the three FTE for nine months would be 585. At an estimated daily charge of $75, the expenses to us would be $43,875. For travel expenses, we assumed that each consultant would stay here during the week and return home over the weekends. An airline ticket would be needed each five person-days, or 117 tickets during the implementation period. At an estimated cost of $300 per ticket, the total travel cost would be $35,100.

The estimated total of the vendor charge, expenses, and travel for a single application, rounded to the nearest thousand, is $979,000.

The cost for all three applications, purchased from separate vendors or as part of a single-vendor integrated suite, is $2,937,000.

3. Hardware

We believe that the total hardware cost to replace our administrative systems with client/server systems will not depend on whether we follow a best-of-breed or a single-vendor strategy. In either case, we estimate we will need three RS/6000 servers for each application at a total cost of $250,000 for a grand total of $750,000.

We expect that there will be a one-time cost for new workstations. We believe that approximately three fourths of the 500 power users will need new workstations prior to implementation to use the new systems effectively. This would be a cost of (1,250,000 x 3/4) = $937,500. Therefore our total hardware cost will be $1,688,000.

In addition, we assume the workstations and servers for the University administrative system users will be replaced on a four-year schedule; these costs are included as yearly on-going costs.

4. Vendor-Provided Training

Because of the disparity in vendor pricing structures, these figures are unreliable. Some vendors include training in the software purchase price; in our sample only Geac and Datatel showed it as a separate amount.

Best-of-Breed:

For the Geac systems, our estimate of $50,000 per system is based on the vendor's recommended training schedule. There was no external training for ISIS, a system developed here at the University; all training was done in-house. The $100,000 total for our comparison represents the sum of the costs for the two Geac systems, at $50,000 each, and ISIS, at zero. ($50,000 + $50,000 = $100,000)

Single-Vendor:

The only integrated system vendor to provide separate training costs was Datatel. We used their figure of $200,000. The $50,000 total shown on our comparison document is the average for this vendor, at $200,000, and the three others, at zero.

Committed Costs for Software Acquisition and Implementation

5. Technical Staff Training
This cost represents the personnel costs to the University; it does not include vendor charges, which are shown in item 4.

Best-of-Breed:

We used JMU's figure of 24 days' training per technical staff member for each of our estimates. We expect there will be ten technical staff members on the implementation team for a single application. Twenty-four days' training for one person, at a fully-loaded labor cost of $53/hour, will cost $10,176. For ten people, the cost will be $102,000 (rounded).

For three applications from different vendors, the full cost would be incurred for each application, for a total of $306,000.

Single-Vendor:

We estimate that there will be a significant savings in training cost under a single-vendor approach because of the common programming language, database, tool kit, etc. We believe that the total cost for all three applications will be $102,000. Twenty-four days total for all applications.

6. User Training

Best-of-Breed:
We believe that we will have approximately 500 "power users" and 500 "casual

users". This is based on knowledge of approximately 300 departments who would most likely have 1 to 2 power users. This 500 also includes the central administrative users. We are also assuming a similar number of casual users in each department.

Based on JMU's experience, we estimate that each of our power users will require 24 hours' training on each of the new systems. We used an estimated fully-loaded labor cost of $30/hour for a grade 13 employee, for a per-person training cost of $720. The total cost to train our 500 power users on a single application, then, is $360,000.

For all three administrative systems, the cost will be $1,080,000.

Single-Vendor:

Again, we expect substantial savings in our training costs in a single-vendor, integrated system environment. Because of the common look-and-feel, we estimate that the cost to train the power users on an integrated suite of applications will be half that to train them on best-of-breed systems, or $540,000.

7. Technical Staff Time

Best-of-Breed:
We expect that ten technical FTE will be needed over an 18-month implementation period for a single application. This represents 30,000 person-hours (using 2,000 hours/year). Using our standard fully-loaded labor cost of $53/hour, we determined that the technical staff cost for the implementation would be $159,000 per person, or $1,590,000 for the team.

However, these 30,000 hours include the twenty-four days already allocated for training, above. Therefore, our final technical staff cost for a single application is $1,590,000 minus the training cost of $102,000, or $1,488,000.

The total technical staff cost for all three applications is three times $1,488,000, or $4,464,000.

Single-Vendor:

The cost of software implementation includes that of temporary interfaces among applications. We estimate that these interfaces account for 20% of the technical staff cost, or $893,000.

We determined that half as many temporary interfaces will be required under a single-vendor approach as will be needed for best-of-breed systems. (See Diagrams A and B.) Therefore, we subtracted half the cost of temporary interfaces, or $446,000, from our total best-of-breed implementation costs for a cost of $4,018,000 for a integrated, single-vendor approach.

8. User Staff Time

 

Best-of-Breed:

We expect that the user representation on the implementation team, like the technical members, will consist of ten FTE (however, these ten FTE might consist of many more people devoting a small portion of their time to the project). This represents 30,000 person-hours (using 2,000 hours/year). Using an estimated fully-loaded labor cost of $50/hour for a professional staff member, we determined that the user implementation cost would be $150,000 per person, or $1,500,000 for the team.

The total user staff portion of the implementation team cost for all three applications is three times $1,500,000, or $4,500,000.

 Single-Vendor:

 

User implementation costs, like those for technical staff, include the cost of temporary interfaces among applications. We estimate that these interfaces account for 20% of the implementation cost, or $900,000.

We determined that half as many temporary interfaces will be required under a single-vendor approach as will be needed for best-of-breed systems. (See Diagrams A and B.) Therefore we subtracted half the cost of temporary interfaces, or $450,000, from our total best-of-breed implementation costs for a cost of $4,050,000 for the integrated, single-vendor approach.

 

On-going Out-of-Pocket Costs

9. Hardware
Our estimated yearly hardware cost has two components: workstation replacement cost and server depreciation. The estimate is the same for either a best-of-breed or a single-vendor strategy.

We used estimates of 500 power users for the administrative systems, and $2,500 for the cost of a single workstation, for a total cost of $1,250,000. An annual replacement cost of one-fourth of the total, or $313,000 per year (rounded) is shown. This cost is the same for a single application under the best-of-breed approach, and for all three applications under either approach.

We estimated that there will be another 500 casual users of these systems. We did not include workstation costs for these people because we expect that whatever client/server system(s) we use will have a Web-based front-end for this user community and will not require a high-end workstation.

Our one-time server cost estimate (discussed previously) is $750,000 for RS/6000 servers. We included, in our on-going hardware costs, a yearly depreciation and maintenance allowance of $200,000 for the servers, assuming four year replacement.

The total yearly hardware cost is the sum of the $313,000 workstation replacement cost and the $200,000 server depreciation allowance, or $513,000.

10. Software

For all our estimates we used vendor-provided figures for our first-year software license fees. We then assumed that the fee would increase by ten percent each successive year.

(See Appendices C and D for first-year software license fees.)

Again, the final best-of-breed total is the sum of our three sample systems, and our final single-vendor total is the average of our four sample systems.

On-going Committed Costs


11. Technical Infrastructure Costs (including staff time and hardware and software environment)

We used the document "ITC Cost of Services for Administrative Applications" as our source for the University's current support costs. We used these numbers as a starting point and applied assumptions about the effect of the client/server environment on the cost of support to derive our estimates of future costs.

Our total current cost for the support of our mainframe-based administrative systems is $5,476,000 per year. Of this, $3,340,360, or 61%, is the cost of personal services. This is our current yearly on-going ITC support cost, which is a committed cost.

We identified three component costs within this overall support cost: troubleshooting, software upgrade installation, and other costs. We expect that the troubleshooting and software upgrade components will be significantly affected by both the change to a client/server environment and the decision about whether to purchase an integrated application suite from a single vendor or continue with a best-of-breed procurement strategy.

  • Troubleshooting: The baseline support portion of our committed cost is $647,613, and we estimate, based on ITC experience, that half of that, or $323,806, is attributable to troubleshooting.

 

  • Software upgrade installation: In our current environment, approximately three percent of our committed cost, or $100,211, is devoted to installing vendor software upgrades.

 

  • Other costs: The remaining portion of our committed cost after subtracting the costs of troubleshooting and applying software upgrades is $2,916,343. Much of this cost is the cost of customization, and, depending on decisions made regarding changing University policies and procedures, our actual costs could be significantly different-either higher or lower. We cannot make a determination as to what decisions will be made in the future about software customization, so we are using our current figure for this cost item.

 

  • Overall support costs: In a client/server environment, overall costs do not tend to go down, although they may be allocated differently. For example, while less time may be needed for systems development, more time may be needed for distributed systems management.

 

Best-of-Breed:

Troubleshooting: Because of the far greater complexity of the multiple-vendor client/server environment, we estimate that troubleshooting will consume four times the resources it does now, for an annual cost of $1,295,225.

Software upgrades: We know from the experience of other institutions that purchased client/server systems have much more frequent upgrades than mainframe systems. We estimate that this more frequent schedule would double our current costs, to $200,422 per year. And, because of the greater complexity of the software environment, we estimate that the cost of upgrades will double again with the new applications, to $400,844 per year.

Other costs: We are using our current figure for this item (see earlier discussion), or $2,916,343.

Total for best-of-breed: The total cost for all three components is $4,612,000 (rounded).

Single-Vendor:

Troubleshooting: We estimate that the simplification of the software environment and clear vendor accountability in a single-vendor approach would be one fourth the cost of best-of-breed troubleshooting for a yearly cost of $323,806 ($1,295,225 divided by 4).

Software upgrades: We expect that the streamlining effect of working with a single vendor would reduce our cost to apply upgrades to one third of the best-of-breed cost estimate, for a cost of $133,615 (400,844 divided by 3).

Other costs: We are using our current figure for this item (see earlier discussion), or $2,916,343.

Total for single-vendor: The total cost for all three components is $3,374,000 (rounded).

12. User Personnel Costs - Support

Best-of-Breed:
We assume that system support will require approximately half the number of FTE from the user areas as from the technical areas for troubleshooting. We divided our technical (in terms of user time) troubleshooting cost of $1,295,225 in half and then added the software upgrade cost of $400,844 for a total cost of $1,048,456.

Single-Vendor:

We assumed that the user cost for single-vendor would be the same as for best-of-breed - $1,048,000.

13. User Personnel Costs - Training

We assume a turnover rate among the application systems power users of five percent per year, or twenty-five new users each year. The new users will require the same training given to all power users when the systems were first implemented, or 24 hours' training for each application. At an estimated fully-loaded labor cost of $30/hour for a grade 13 employee, the training for a single user will cost $720. The cost for all twenty-five new users would then be $18,000.

Best-of-Breed:

Under a best-of-breed approach, the new users could need the entire training course for each of the three applications. The total cost would be three times the $18,000 for each application, or $54,000.

Single-Vendor:

Because of the common look and feel of a single-vendor, integrated suite of systems, the portion of the user training devoted to general screen navigation would be needed only once for each user, rather than three times. We estimate that this would reduce the cost of training by half, for a final cost of $27,000.

 

C. Best of Breed Costs

Best-of-Breed Costs

Vendor:

DBS
Financial
DBS
HR
PeopleSoft
(Med Ctr - GL)
ISIS

Combined

Acquisition and Implementation Costs:

Out-of-pocket Costs:

Software

767,000

489,000

355,000

65,000

1,321,000

Vendor Consulting/Customization

979,000

979,000

-

50,000

2,937,000

Hardware

563,000

563,000

280,000

563,000

1,688,000

Vendor Training

50,000

50,000

-

-

100,000

Committed Costs:

Technical Staff Training

102,000

102,000

-

25,000

306,000

User Training

360,000

360,000

-

-

1,080,000

Technical Staff Time

1,488,000

1,488,000

150,000

4,066,000

4,464,000

User Staff Time

1,500,000

1,500,000

300,000

400,000

4,500,000

Total One Time Costs:

5,809,000

5,531,000

1,085,000

5,169,000

16,396,000

On-going Costs:

Year 1

Out-of-pocket Costs:

Hardware

388,000

388,000

313,000

-

513,000

Software

200,000

200,000

200,000

737,000

1,137,000

Committed Costs:

Technical Infrastructure Costs (Including Staff)

1,384,000

1,614,000

-

1,614,000

4,612,000

User Personnel Costs - Support

314,000

367,000

367, 000

1,048,000

User Personnel Costs - Training

18,000

18,000

18,000

18,000

54,000

Year 1 Total

2,304,000

2,587,000

531,000

2,736,000

7,364,000

Year 2

Out-of-pocket Costs:

Hardware

388,000

388,000

-

-

513,000

Software

220,000

220,000

220,000

737,000

1,177,000

Committed Costs:

Technical Infrastructure Costs (Including Staff)

1,384,000

1,614,000

-

1,614,000

4,612,000

User Personnel Costs - Support

314,000

367,000

367,000

1,048,000

User Personnel Costs - Training

18,000

18,000

18,000

18,000

54,000

Year 2 Total

2,324,000

2,607,000

238,000

2,736,000

7,448,000

Year 3

Out-of-pocket Costs:

Hardware

388,000

388,000

-

-

513,000

Software

242,000

242,000

242,000

737,000

1,221,000

Committed Costs:

Technical Infrastructure Costs (Including Staff)

1,384,000

1,614,000

-

1,614,000

4,612,000

User Personnel Costs - Support

314,000

367,000

367,000

1,048,000

User Personnel Costs - Training

18,000

18,000

18,000

18,000

54,000

Year 3 Total

2,346,000

2,629,000

260,000

2,736,000

7,448,000

Year 4

Out-of-pocket Costs:

Hardware

388,000

388,000

-

-

513,000

Software

266,000

266,000

266,000

737,000

1,269,000

Committed Costs:

Technical Infrastructure Costs (Including Staff)

1,384,000

1,614,000

-

1,614,000

4,612,000

User Personnel Costs - Support

314,000

367,000

-

367,000

1,048,000

User Personnel Costs - Training

18,000

18,000

18,000

18,000

54,000

Year 4 Total

2,370,000

2,653,000

284,000

2,736,000

7,496,000

Year 5

Out-of-pocket Costs:

Hardware

388,000

388,000

-

-

513,000

Software

293,000

293,000

293,000

737,000

1,323,000

Committed Costs:

Technical Infrastructure Costs (Including Staff)

1,384,000

1,614,000

-

1,614,000

4,612,000

User Personnel Costs - Support

314,000

367,000

-

367,000

1,048,000

User Personnel Costs - Training

18,000

18,000

18,000

18,000

54,000

Year 5 Total

2,397,000

2,680,000

311,000

2,736,000

7,550,000

Total 5-year On-going Costs:

11,741,000

13,156,000

1,624,000

13,680,000

37,262,000

Total 5-year Costs:

17,550,000

18,687,000

2,709,000

18,849,000

53,658,000

 

D. Single Vendor Costs
Single-Vendor Costs

Vendor:

Buzzeo
Datatel
PeopleSoft
JMU
Averaged

Acquisition and Implementation Costs:

Out-of-pocket Costs:

Software

1,800,000

1,140,000

3,400,000

1,187,000

1,882,000

Vendor Consulting/Customization

2,937,000

2,937,000

2,937,000

1,125,000

2,937,000

Hardware

1,313,000

1,313,000

1,313,000

1,313,000

1,688,000

Vendor Training

-

200,000

-

-

50,000

Committed Costs:

Technical Staff Training

102,000

102,000

102,000

440,000

102,000

User Training

540,000

540,000

540,000

-

540,000

Technical Staff Time

4,018,000

4,018,000

4,018,000

604,000

4,018,000

User Staff Time

4,050,000

4,050,000

4,050,000

268,000

4,050,000

Total One Time Costs:

14,760,000

14,300,000

16,360,000

4,937,000

15,267,000

On-going Costs:

Year 1

Out-of-pocket Costs:

Hardware

513,000

513,000

513,000

54,000

513,000

Software

61,000

200,000

578,000

197,000

259,000

Committed Costs:

Technical Infrastructure Costs (Including Staff)

3,374,000

3,374,000

3,374,000

67,000

3,374,000

User Personnel Costs - Support

1,048,000

1,048,000

1,048,000

-

1,048,000

User Personnel Costs - Training

27,000

27,000

27,000

-

27,000

Year 1 Total

5,023,000

5,162,000

5,540,000

318,000

5,221,000

Year 2

Out-of-pocket Costs:

Hardware

513,000

513,000

513,000

54,000

513,000

Software

249,000

220,000

636,000

217,000

331,000

Committed Costs:

ITC Personnel Costs

3,374,000

3,374,000

3,374,000

67,000

3,374,000

User Personnel Costs - Support

1,048,000

1,048,000

1,048,000

-

1,048,000

User Personnel Costs - Training

27,000

27,000

27,000

-

27,000

Year 2 Total

5,211,000

5,182,000

5,598,000

338,000

5,293,000

Year 3

Out-of-pocket Costs:

Hardware

513,000

513,000

513,000

54,000

513,000

Software

274,000

242,000

700,000

239,000

364,000

Committed Costs:

ITC Personnel Costs

3,374,000

3,374,000

3,374,000

67,000

3,374,000

User Personnel Costs - Support

1,048,000

1,048,000

1,048,000

-

1,048,000

User Personnel Costs - Training

27,000

27,000

27,000

-

27,000

Year 3 Total

5,236,000

5,204,000

5,662,000

360,000

5,362,000

Year 4

Out-of-pocket Costs:

Hardware

513,000

513,000

513,000

54,000

513,000

Software

301,000

266,200

770,000

263,000

400,000

Committed Costs:

ITC Personnel Costs

3,374,000

3,374,000

3,374,000

67,000

3,374,000

User Personnel Costs - Support

1,048,000

1,048,000

1,048,000

-

1,048,000

User Personnel Costs - Training

27,000

27,000

27,000

-

27,000

Year 4 Total

5,263,000

5,228,000

5,732,000

384,000

5,362,000

Year 5

Out-of-pocket Costs:

Hardware

513,000

513,000

513,000

54,000

513,000

Software

331,000

292,820

847,000

289,000

440,000

Committed Costs:

ITC Personnel Costs

3,374,000

3,374,000

3,374,000

67,000

3,374,000

User Personnel Costs - Support

1,048,000

1,048,000

1,048,000

-

1,048,000

User Personnel Costs - Training

27,000

27,000

27,000

-

27,000

Year 5 Total

5,293,000

5,254,000

5,809,000

410,000

5,402,000

Total 5-year On-going Costs:

26,026,000

26,030,000

28,341,000

1,810,000

26,604,000

Total 5-year Costs:

40,786,000

40,330,000

44,701,000

6,184,000

41,871,000

 

E. Assumptions

Scope:

 

  1. The "costs" we estimated are quantifiable costs; other groups will be considering intangible costs and benefits. Since our group is not charged with doing a cost/benefit analysis, we will not be looking at benefits and cost reductions that would result from the implementation of an integrated system. Our estimate is solely for the tangible costs to implement each alternative.
  2. We did not include costs incurred prior to acquisition such as the procurement process, site visits, etc. Acquisition costs in terms of institutional personnel resources will be lower for an integrated system because of economies of scale, and higher for a best-of-breed strategy because of multiple selection and purchasing processes.
  3. The costs included in our estimates will be University costs only; any costs to the Medical Center to interface with our new systems are not considered a part of this project.
  4. The cost estimates include the full replacement of three of our core administrative systems (Student, Financial, and Human Resources), over the next five years.
  5. Providing an estimate for continuing with the status quo was not part of our mission; however, our estimates include some already committed costs --for example, even if we don't replace any systems we will still incur the personnel costs associated with supporting our present systems, and we will still proceed with our planned workstation replacement program.

General Assumptions:

  1. A best-of-breed strategy would result in three different databases and technical architectures, one for each of the three core systems.
  2. Under a best-of-breed approach, our systems would still require some level of technical integration. The estimate reflects interfacing but not integrating all best-of-breed systems.
  3. The cost to upgrade client workstations should be the same regardless of whether we decide on an integrated or best-of-breed approach.
  4. The implementation will be faster and less expensive for an integrated system because of the common database and technical architecture. For example, we will need to write more temporary interfaces between applications under a best-of-breed approach. (See Diagram A.)
  5. Training costs for technical users, power users, and end-user departments will be greater under a best-of-breed approach.
  6. Ongoing maintenance costs will be greater under a best-of-breed approach, because of multiple data bases and data structures, and more complex upgrades.
  7. Acquisition costs in terms of institutional personnel resources will be lower for an integrated system because of economies of scale, and will be higher for a best-of-breed strategy because of multiple selection and purchasing processes.
  8. Regardless of which alternative we decide on (integrated vs. best-of-breed), we will be moving to a client/server environment.
  9. Estimates of the difference in troubleshooting and problem resolution are based on the amount of time we spend on production problems now. We expect an increase in the time spent because of the greater complexity of the client/server environment, and, initially, because of the learning curve associated with the new technology. We also feel that the time needed to resolve problems would be significantly less with a single-vendor, integrated system than in a best-of-breed environment.
  10. Support for client/server systems should require approximately the same number of people currently required to support our mainframe systems, but they will be allocated differently. Fewer will be needed at the application level, but more will be needed to support the infrastructure.
  11. The current University network is robust enough to handle system traffic.

Cost Assumptions:

The standard workstation needed will be a Pentium 133, with 32 MB Ram, a 1-GB hard drive, and a 17-inch monitor, running Windows 95. The average cost per workstation will be $2,500. The total workstation cost will be $1.25 million. Under our four-year replacement plan, we can expect an average yearly cost of $313,000 (one-quarter of $1.25 million).

Our new systems will have 500 power users and 500 casual users. 150 users could be using the system concurrently at any time. Casual users probably will not need Pentium computers for the new systems. Any computer that can support a web browser will be sufficient. Note that the casual users of the new administrative systems may need Pentiums for their other work, but the cost cannot be attributed to this project.

  • Average personnel cost is $53/hour for ITC, $50/hour (average professional employee) and $30/hour (average grade 13) for users.
  • We will assume that software license and vendor support costs will rise 10% per year.

Unquantifiable Cost:

  1. The cost of redundant data (same data in different systems, e.g. addresses) is unquantifiable.
  2. Some costs of the mainframe to client/server migration are unquantifiable, although there is a possible saving for an integrated approach as compared with a best-of-breed approach. This category includes distributed system management tools, the infrastructure software needed on any large system that we will need to acquire when we move from mainframe to client/server applications. Some application systems may include some of this software, and would satisfy our need for that function in a single-vendor environment. In a best-of-breed environment, however, even if one application included this software, it might not work for the other applications, and we would have to purchase it separately.
  3. The impact of system replacement on supplemental and shadow systems should be the same regardless of whether we decide on an integrated or best-of-breed approach. It was noted, however, that this cost will be determined in large measure by the institution's willingness to change its business practices.
  4. The cost of add-on features is unknown, since it will vary considerably from one vendor to another. This category includes functions such as resume scanning, data warehouse, security, laser check printing, and EDI.
  5. The small costs of system replacement--e.g., the costs to departments to replace their SAS reports-- will be the same under either approach.
  6. Under a best-of-breed approach, the cost must include an integrated front end, probably Web-based, and Warehouse reporting. We did not quantify this cost because the technology in this area is still under development. We believe this cost to be in the range of several hundreds of thousands of dollars.
  7. Vendor supplied costs were pre-negotiation costs. It is very possible that additional costs could be uncovered during a negotiation process. It is also possible that savings or additional costs could be realized.

 

F. Conversations With Other Schools

UNIVERSITY OF WISCONSIN

I spoke with Scott Manley (smanley@doit.wisc.edu), whose title is Strategic Consultant.

They have chosen to use PeopleSoft for their Student Information systems, while maintaining their existing systems in other areas for the time being. Their primary reason for keeping the existing systems in those areas is cost. The acquisition and implementation of PeopleSoft for Student Information areas will consume their entire discretionary spending allowance for the next three years.

Their HR system is relatively new, developed in-house, and thought to be mostly YR2000 compliant. For financial accounting systems, which are older, they feel they can address YR2000 problems in the course of routine maintenance, more or less. (It's something of a political issue on campus that the decision was made to spend the entire discretionary budget for three years on SIS needs).

Their IT division was not exactly trying to take the approach that they did, but sort of happened upon it. They were looking for a package to replace just their Financial Aid system, (prompted primarily by YR2000 considerations), and invited TRG and PeopleSoft to do presentations. Users from several other areas were invited to attend, and were so excited about PeopleSoft that they really spearheaded the effort to acquire it for all Student Information functions. Wisconsin is joining Universities of Michigan, Minnesota, Iowa, Ohio State and perhaps others in a joint effort with PeopleSoft to develop their Student Information System. An important factor in taking the step at this time was the opportunity to join with these other universities as 'charter members', which would enable them to acquire the software at half price and to acquire Beta software. Another factor was an outrageously high estimate for making YR2000 changes and converting to a relational database.

The contract was signed three months ago. They expect the effort to take about two years to initial implementation. Teams have been formed to develop requirements for all Student Information functions, and leaders of these teams form a cross-functional team.

They have made a definite decision to change business processes to conform to the purchased software rather than the reverse. An outside consulting firm, Proforma (he thinks Chicago-based), has been brought in to expedite that process, at $3500 a day. Some user departments don't know what their current systems actually do. Proforma works with them to determine that, compare it to PeopleSoft, do 'fit analysis' and decide what practices need to change. They have very committed and enthusiastic users. Initial high-level scope process by Proforma is complete (after about one month), and they expect results reported at any time.

PeopleSoft will be implemented on RS6000 under AIX using Oracle, which they feel will keep their IT people quite busy.

UW-Madison's purchase price for the PeopleSoft student information system software was $1,025,000. This is a discounted price based on their status as a charter member of the software development group. Six functional modules are included for this price. They are:

  • Admissions
  • Student Records
  • Financial Aid
  • Student Financials
  • Academic Advising
  • Campus Community

The original project budget for the PeopleSoft implementation was $10,490,160.   The implementation timetable calls for production status for the Campus Community, Admissions, and Financial Aid modules in time for the Fall 1998 semester. The remaining modules will be production in Fall 1998, if possible, but otherwise in production no later than Fall 1999.

For more information on the PeopleSoft Integrated Student Information System project at UW- Madison, please feel free to contact the project management team:

  • Don Wermers - Project Manager (Registrar)
    • email: DON.WERMERS@MAIL.ADMIN.WISC.EDU
      address: 750 UNIV AVE MADISON, WI 53706
      building: PETERSON OFFICE BUILDING, A W 130C
      phone: 608-262-3964
      title: REGISTRAR (L)
      division: ACADEMIC SERVICES
  • Ron Niendorf - Associate Project Manager
    • email: RON.NIENDORF@MAIL.ADMIN.WISC.EDU
      address: 500 LINCOLN DRIVE MADISON, WI 53706
      building: BASCOM HALL 356
      phone: 608-262-2025
      phone2: 608-262-0130
      title: DIR, COMPUTER SRV (S)
      division: ACADEMIC SERVICES
  • Tom Scott - Associate Project Manager
    • email: TSCOTT@DOIT.WISC.EDU
      address: 1210 W DAYTON ST MADISON, WI 53706
      building: COMPUTER SCIENCES AND STATISTICS 2112
      phone: 608-263-1750
      division: INFORMATION TECHNOLOGY
      title: Director - Strategic Consulting

UNIVERSITY OF OREGON

I spoke with Jim Bohle (jbohle@oregon.uoregon.edu). His title is Assistant Director of Administrative Services.

They made the decision about 8 years ago to use SCT in what has turned out to be an integrated systems approach. They had a somewhat archaic 12-15 year-old home grown system running on mixed platforms in mixed languages, etc. This system was only for Student Information functions; HR, payroll, and financial functions were performed centrally by the state.

In 1988 they started looking at packages because they felt that they would become bogged down in endless haggling over details if they attempted in-house development. They wanted something that would run on DEC VAX and use relational database. They looked at SCT BANNER and IA. Implementations which were PC-based were not considered. As vendors were invited in for a series of demonstrations, other schools in the state were invited and became interested. They liked the demos and saw SCT as a well-managed corporation with good technological understanding and use.

When financial, HR and payroll functions were transferred to them from the state and they were ready to acquire a new HR/Payroll system, they were apparently not committed to maintaining an integrated approach, and also considered PeopleSoft. They decided to continue with SCT BANNER because it would allow them to stay integrated, they already had it, and they liked it. This was another decision that was made by all the state higher ed schools in concert; the larger schools tended to like SCT BANNER and the smaller ones tended to like PeopleSoft.

The original decision to go with SCT BANNER was made in 1989 and they have implemented the Student Information and Financial Information functions (FIS implementation included student health, physical plant, and a couple of other functions). They will be implementing HR/Payroll soon. They brought functions in one at a time, and extensively customized them to meet user preferences rather than asking users to alter business practices. (Since that time, SCT has incorporated some of the functions they customized into their base product, so they hope to operate less in the customized mode in the future). Examples of functional areas where they felt that SCT was deficient for their purposes are telephone registration and degree audit. SCT tries to cover a wide range of customers and has not always been able to provide the desirable depth for particularly specialized client applications, but are improving in this area.

The SCT system itself was not the major expense. Oracle has been most costly in terms of ongoing costs, and has not been easy to deal with in terms of licensees, etc. For instance, they had a simple machine CPU upgrade which was referenced in their licensing agreement, and Oracle wanted $400,000 to change the licensee, finally settling for $100,000. Other than that anecdote, he had no specific cost information to share.

Their current SCT BANNER systems are not client-server, though the new HR/Payroll functions will be and they will probably move in that direction with existing functions in the next 12 to 18 months. This will require a lot of workstation upgrading. They wrote their own client-server Data Warehouse system using Oracle and Graphical Query Language to extract information and report. They have also built an in-house query system using Oracle forms for the Student Information system.

They have high levels of user satisfaction primarily because they catered to user needs and preferences rather than requiring them to change. This high level of customization has made it much more difficult to maintain when there are new releases and upgrades. The user departments were also very involved in the entire process. The greatest resistance was encountered in the financial areas, but people came around, and they achieved some substantial improvements.

No consultants were used in the Student Information systems implementation. They have used consultants (he didn't know who), in evaluating processes and doing fit analysis in the financial and HR/Payroll areas. There is a lot of feeling that they wasted a lot of time and money and didn't contribute much.

They have been very happy with SCT. Feel that they are technologically strong and that they respond quickly and expertly to any problems they've had. They see them as continually improving their expertise in higher ed applications areas, getting smarter and smarter, and see them looking ahead and staying on top of changing technologies.

They also feel that they benefit greatly when making these kinds of decisions from close cooperation with the other higher ed institutions in the state.

UNIVERSITY OF TORONTO

I spoke with Chris Handley, Director of Administrative Computing & Director of Re-Engineering. He is something of an expert on change management and re-engineering and travels about the world lecturing on the subject.

They felt that the integrated approach is an absolute necessity and chose SAP because they feel it's state-of-the-art, and that since universities don't change often, it's important that when they do change, they go for the cutting edge, since they won't soon change again. Also because it's powerful, and Oracle-based.

They knew they would have to invent some of the pieces. Bought HR/Payroll and financials, then worked with SAP's 'alliance partners' to develop Research Tracking and Alumni Tracking Systems. Student Systems is another administrative arm and elected not to go with SAP, though they could have, so went with a different system that was Oracle-based. Then a reporting tool from IBT was bought to report across both platforms.

The decision was made three years ago. It was hard fought, and the VP for Computing left. (I'll leave out a lot of detail here). They brought Financial system up first. It was not on time and necessitated a mid-year implementation which was not pretty. This was NOT SAP's fault. He says they were a terrible client. He feels that the reason they had this problem was that they greatly underestimated the degree of cultural change entailed and the training needs that would arise. It changed the work for everyone in the system, gave people everywhere the ability to affect the system as a whole. HR changes affected budget, for instance, whereas before they were separate entities. It was very destabilizing. Academics were not happy to see this decentralization of power and authority.

They were sure that they didn't want to just automate what they were doing; they wanted to become more standard, keep the plain vanilla version of the product. They re-engineered business practices to do so (He's done 18 of these re-engineerings!). They decided to customize in two ways, reports and building WWW interfaces for academics. (They are also having to customize the payroll module somewhat because it's US and they're Canadian).

SAP is a Euro-company, becoming adept at public sector applications. They are probably still somewhat better for municipalities than for higher ed. For instance they had no way to deal with multiple appointments, but are working on incorporating that and similar higher ed needs into their package. They have been purchased by MIT and Harvard, and the European Common Market.

The estimated cost of hardware, software, and effort is $15M for all the SAP systems. The Oracle-based SIS is estimated at $6M.

He doesn't recommend consultants. He thinks you have to make these decisions yourselves and take responsibility for them.

They've been live for one year. His descriptions of user acceptance levels are quite colorful. We're talking ballistics here. They found that many people performing administrative tasks were not really trained, and didn't really know what is was that they were doing. A lot of education was required. He says they are slowly coming around as they see what is actually possible. Measures level of success by the fact that several departments have abandoned their shadow systems, a sure sign of comfort. These difficulties occurred in spite of heavy user involvement from the beginning. He felt that users involved themselves but didn't really think it would ever happen (they'd heard all that talk of change before) and were unprepared when they actually put the system up!

TIPS:

 

  1. Pilot new system in one faculty. It will reveal 95% of the problems. Flood the place with personnel and fix the problems. Six months later, bring the others on board. (They did not do this).
  2. You cannot communicate enough. It doesn't work just to communicate to the top and it doesn't work to communicate to everyone. You have to figure out who you really need to communicate with. And you must communicate from the top.
  3. ONE NECK, ONE BLOCK, ONE AXE. It is vital that there be a single person who leads the effort from both the business and the technological side. That person must be publicly identified as the person responsible, and his/her remuneration should depend on the success of the effort. (They did do this).

Chris emphasized a great deal that throughout all his re-engineering efforts, one thing is very clear to him. Universities are all alike, and they all think they are unique. But from the standpoint of what they need to do business, the differences are insignificant, even if they are in other countries. He sees no significant differences in the way Toronto operates, or MIT, or the University of South Africa. The business needs are all the same.

TULANE UNIVERSITY

I spoke with Rick McGinity (rsm@mailhost.tcs.tulane.edu). He is Assistant Director of Administrative Computing. (504) 862-8000 x2559.

Tulane has gone through a process similar to the one we're in now in searching for an integrated systems package that would meet their needs. They haven't found anything so far and have become skeptical of claims made by the major vendors of client-server systems. They have looked at TRG and PeopleSoft.

They are upgrading to OS390, which incorporates a UNIX partition. When looking at the UNIX based client server systems TRG and PeopleSoft were offering, they felt they could not supply what was needed for administrative applications. Integrity and security issues were primary, as well as response time. One major concern was upgrades. With their mainframe CICS system they have test and production regions and can test upgrades before they go into production. With the client-server systems the upgrade just goes on top of the existing version, and one would have to purchase and maintain duplicate hardware in order to have a test area. PeopleSoft has something which purports to address this issue, but they didn't trust it. Their UNIX machines are breached regularly by student hackers, so they don't feel that they can guarantee the security of the data on client-server systems. With regard to response time, he says that 15 seconds to 1 minute is considered good. He has seen 2 to 5 minutes be the norm on a site visit.

 

They are also concerned that client-server systems would require standardized desk tops, which they don't feel they can achieve or justify financially. They are also concerned about the increased need for network capacity, which they are also not sure they could meet.

IBM is restructuring their costs and coming out with improvements in the mainframe that compete effectively with what client-server systems offer. They are now looking toward Oracle mainframe applications using an Integrated Browser which can also issue direct calls to CICS regions and also give them the WEB connections that some feel is important. They are going to have a demo by SCT soon, but they know that they haven't covered the upgrade issue. They are also trying to determine how much of their client base is 'fat client' (power users who need all sorts of interfaces and data manipulations), and 'thin client' (who just mostly browse data).

His main piece of advice: SITE VISITS. (Don't believe anything until you see it yourselves). He says that they encountered an amazing degree of disparity between what vendors said was happening at a site and what they actually saw when they got there (systems said to be running successfully that weren't even operational yet, people who were said to be using their package who were only using a tiny piece of it, incredibly bad response times, implementation horror stories, etc.) He emphasized that site visits should be without vendor representatives present, and that there should be the opportunity to talk to IS people who were not protecting the decision they had made.

CARNEGIE MELLON UNIVERSITY

I spoke with John Bird (jb8g@andrew.cmu.edu). He is a Project Director in Administrative Services. (412)268-5861.

Carnegie Mellon has a combination approach (really sounds like best-of-breed). Their Student Information System is written in-house using Ingres and C. They began it in 1988 and it has been up and running for a while. There is a compatible purchased Academic Audit package. There is push to move more to the WEB and they will eventually do that.

For HR they have an in-house 1980's system which feeds a SYBORG payroll package. They find the SYBORG time-consuming and clumsy. Will outsource payroll soon to ADP.

Their Financial systems are the ones which are sending them out on the client-server trail. They are old COBOL programs which are not YR2000 compliant. A big review process began 2 years ago. RFI's went out to 30 vendors about a year ago, and the field was eventually narrowed to five -- PeopleSoft, Oracle, Computron, SCT, SAP. There were demos, etc., with user participation. SAP reps were rude and rigid and turned everyone off immediately. PeopleSoft was only interested in selling them an integrated system to do everything. SCT technology didn't look good to them, had a mainframe look and feel.

The field was narrowed to Computron and Oracle. The business managers really liked Computron. They had nice interfaces, were cooperative, responsive, and flexible. The IT people really liked Computron. It ran on different databases, including Ingres; they liked the Computron people, asked them to show them a lot of things and they were always responsive and it always performed beautifully. Oracle people were not responsive, and it will double their personnel requirement to use it. The Executive Steering Committee didn't like Computron because they aren't in the higher ed market and they questioned their financial stability. They decided to benchmark both until Spring and then make a decision. They are studying Stanford, which has Oracle, and CUNY, which has Computron.

Oracle and Computron were comparable in price (which was pretty low, but he asked not to be quoted on an actual guess), and in implementation time (18 months to 2 years). They would have to upgrade their hardware to some sort of UNIX machine in either case.

They think that because of the YR2000 pressure, they just have to make a decision and get it up, knowing that the users may not be entirely happy with it and will probably want a lot of customization later. In effect they will be asking users to change their business practices, but aren't optimistic that they will actually do so. They feel that they just have to go ahead now and deal with that problem when the time comes.

 

Summary of Conversation with SUNY -Buffalo 10/15/96

Susan Huston Director Administrative Computing Services
(716) 645-3587 (x7002)

Most of their current systems were locally developed. They are presently looking to replace several systems including Financials. They recently (last year) installed the BSR Alumni & Development system and purchased out of Miami University of Ohio a Degree Audit Package (GARRS?). The A&D system was purchased as a best-of-breed because at that time they only were considering the replacement of that one system and integration was not a concern. They also looked at BANNER (SCT) and ranked it second. Integration was also not an issue with the Degree Audit system as it stands on its own for the most part.

At this time, they are considering integrated, best-of-breed and local development as options for future systems purchases. The Financial system they are considering is ORACLE, but it does not have a Student package. They have also looked at PEOPLESOFT which is at their Stoney Brook location, but have not been overly impressed. Susan will provide names of the team at Stoney Brook should we wish to contact them.

The team looking at system approaches is primarily from the business side. They decided not to involve the student side until they first determine if ORACLE (best-of-breed) will be their recommendation.

Pros and Cons of Integrated and Best-of-breed:

Integrated
Pros
  • Maintenance is cheaper

Cons

  • Often the vendor consists of generalists versus specialists.
  • One component is often the star and updates to the others suffer.
  • Hard to find an integrated system that meets everyone's needs; someone gets short- changed.
  • Integrated solutions are not always as integrated as are claimed or would be if built locally; must look closely at the data structures to see if integrated only on the screen or deeper too.

Best-of-Breed

Pros
  • Often do not have to give up anything; no one gets short-changed
  • Vendor usually keeps more up-to-date than an integrated system
  • Often a better product

Cons- Can be difficult to integrate

  • More expensive to maintain

Solution taken by SUNY:

Do reporting and analysis through a data warehouse. This lessens the critical need for integration in production. They have spent two years so far on their warehouse and feel it has made the biggest impact in problem resolution. By spring they will have most summary analysis tables available. First dealt with operations issues through the warehouse (labels, printing....).

Summary:

Leaning toward best-of-breed

G. Questionnaire Responses

PROPOSED QUESTIONNAIRE

ISTF STUDY GROUP #3

NOTE 1: Definition for Best-of-Breed and Integrated Systems were given to respondents.

NOTE 2: Verbal responses or number of responses in each category are in bold and italicized.

1) For your core administrative systems, are you using an integrated system, a "best of breed" approach, or a combination of both?

(3) BEST Of BREED (proceed to Question 2)

(2) INTEGRATED (proceed to Question 5)

(2) BOTH INTEGRATED AND BEST Of BREED (proceed to Question 5)

2) If you are exlusively using a "best-of -breed" approach, have you considered the use of an integrated system?

(2) YES (proceed to Question 3) (1) NO (STOP -Return Questionnaire)

We typically build new systems in-house.

3) If you considered using an integrated system, what prompted you to explore changing from a "best-of-breed" system (check all that apply):

(2) O desire for increased efficiency

(1) O desire to decrease programming costs

(1) O desire for easier training of personnel

(1) O desire for a paperless system

(1) O need to handle additional volumes of information

(1) O Other (please describe):

Easier to use (common look and feel), Avoid redundancies, Get out of the programming/maintenance business.

4) If you are using a "best-of-breed" approach because your institution decided against an integrated system, what were your reasons (check all that apply)?

(2) O Could not find a system to meet our needs

O Cost was prohibitive

O Would have required too much training

O Could not find a company that we felt we could work with

(1) O Other (please describe):

I don't think integrated systems existed when our systems were designed.

*** STOP - Return Questionnaire***

5) If you are using an integrated system or a combination of integrated and best of breed, what prompted you to explore changing from an exclusive "best-of-breed" system (check all that apply):

(2) O desire for increased efficiency

(1) O desire to decrease programming costs

(1) O desire for easier training of personnel

(1) O desire for a paperless system

(1) O need to handle additional volumes of information

(1) O Other (please describe):

We started from integrated and moved to best-of-breed.

6) If you are using an integrated system, what applications are included in this system (check all that apply)?

(1) O Human Resources

(2) O Payroll

(4) O Financial Accounting

(1) O Management Accounting

(3) O Purchasing

(3) O Accounts Payable

(3) O Student Information/Grades

(3) O Student Billing

O Grant and Contract Management

O Facilities Scheduling

O Facilities Maintenance

O Investment Management

(1) O Development Records

(1) O Alumni Records

(1) O Other (please describe)

Student Housing

7) Which of these applications are currently operational under an integrated system at your institution (check all that apply):

(1) O Human Resources

(2) O Payroll

(4) O Financial Accounting

O Management Accounting

(3) O Purchasing

(3) O Accounts Payable

(3) O Student Information/Grades

(3) O Student Billing

O Grant and Contract Management

O Facilities Scheduling

O Facilities Maintenance

O Investment Management

(1) O Development Records

(1) O Alumni Records

O Other (please describe):

8) What vendors did you pursue before choosing your integrated system?

O Peoplesoft

O TRG

(2) O SCT

O CIS

(1) O Datatel

O Buzzeo

(3) O Other (please describe)

Chose SMS first then looked for other systems which ran on the same data base; Can't remember - players were different at the time; Computing Options in Fredericksburg, MD.

9) What vendor did you choose for your integrated system?

O Peoplesoft

O TRG

(2) O SCT

O CIS

O Datatel

O Buzzeo

(2) O Other (please describe)

Combination of Bay of Plenty Polytechnic & Applied Logic Systems; Computing Options of Fredericksburg, MD.

10) Why did you choose this vendor (check all that apply)?

(3) O Their system best met our needs

(3) O The price seemed reasonable

(2) O The training time to learn the system seemed reasonable

(1) O Their staff appeared to be interested in our concerns

(3) O The necessary modifications to their system seemed minimal

(1) O We were impressed by the track record of the company

(1) O There were certain "special features" that we thought we'd use

O We were dissatisfied with products we've used from other vendors

(1) O Other (please describe):

Has Been Here Since 1986.

11) What were the costs to implement an integrated system (round to the nearest thousand)?

SMS & FMS approx. $300,000-400,000

12) What percent of the total project costs could be attributed to customization of the package?

10%; 20%

13) What percent of the customization was done:

a) Internally, by your own employees _____%

0%; 75%

b) By the vendor _____%

100%;25%

c) By a consultant _____%

0%; 0%

14) What was the implementation time (in months) to get the package up and running from the day the vendor was selected?

6 months; 9-12 months; 36+ months

15) Did you experience any significant delays or unanticipated costs during the implementation of the project? If yes, please describe:

(1) ( ) YES (2) ( ) NO

We were switching their software to HPUX, a first-time UNIX implementation. This involved many unanticipatable changes.

16) Did you use a consultant during any portion of the decsion making process?

(1) YES (continue with Question 17) (2) NO (continue with Question 20)

17) During what portions of the decision making process did you use a consultant (check all that apply)?

O In deciding whether to go with an integrated system versus a "best-of- breed" approach.

(1) O In determining user requirements for the new system

(1) O In writing the specifications for the purchase

O In selecting the vendor

O In implementing the system

O In training our personnel

O Other (please describe):

18) If you used a consultant during the process, are there any suggestions you would make in how to best utilize the consultant?

Produce a "Terms of Reference" for your expectations of them.

19) If you had the process to do over again (using the consultant), would there be any changes you would make to the process?

20) What do you like best about your integrated system?

Encourages staff to take a more global look at their data; The integratedness of the system.

21) Are you satisfied with the vendor's performance?

Not entirely; Reasonably

22) Are there any unresolved problems in implementing the system?

Parts of the system not implemented. Ownership of data not formalized. Training still a bit of pot luck; Some but they will be resolved.

23) If you are using an integrated system and had the decision to make over again, is there anything you would have done differently?

Lower our expectations.

***STOP - Return Questionnaire***



IV. Diagrams

A. Interfaces Needed for Best of Breed Replacement

B. Interfaces Needed for Integrated Replacement

 

V. Glossary

Casual Users - Casual Users would be those persons who use the applications somewhat infrequently in their jobs. That is, the majority of their workload does not involve the use of any of the applications. They may never use some of the applications (e.g. they place orders for their department but never use ISIS).

Committed Costs - Committed Costs are those that are expected to be incurred regardless of which direction the University goes. For example, the number of ITC staff would be expected to remain the same but the tasks they are performing might change.

Fully Loaded Costs - Fully loaded costs are the costs of a person's productive hours after adding in things such as training time and expense, administrative time, vacation, etc. For example, an employee is paid for 2080 hours per year. However, after subtracting out the hours for vacations (80 hours), holidays (88 hours), training (40 hours) and administrative time (e.g. 200 hours), they really only have 1672 productive hours. By dividing their annual salary + fringe benefits + any additional expenses associated with that position (including department overhead) by the productive hours, you arrive at a fully-loaded cost per hour. That is, the cost to the organization for an hour of productive work.

Out of Pocket Costs - Out of pocket costs are those costs that are not currently being incurred, that is, additional costs that will be taken on. Examples would be the cost of new software, licenses for the new software, etc. They are costs that are not incurred if the institution maintains the status quo.

Power Users - Power Users are those persons who use the applications frequently in their jobs and probably use more than one application in performing their work. A majority of a power user's day would be spent inputting, reviewing or analyzing information from the major applications.

Workstation - A workstation consists of a computer and a monitor. It does not include the cost of a printer, furniture, or a fax machine.


Return to the Integrated Systems Project Home Page

Site maintained by kat4w@virginia.edu
Updated January 12, 1999