TechTalks Transcript
Budgeting for IT Infrastructure�
February 7, 2002
Audio
• Streaming
MP3
• Download
MP3 (Download
Tips)
 Judith Boettcher [JB] |
 Howard Strauss [HS] |
 David Lambert [DL] |
JB: Welcome to the CREN Tech Talk series for spring of 2002 and to this session on Budgeting for the IT Infrastructure. You are here because it�s time to discuss the core technologies for your future campus. This is Judith Boettcher, your CREN host for today, and our session is coming to you today with the support of the CREN member institutions. Let me welcome Howard Strauss of Princeton. Howard is a well-known web technology expert and portal expert and, of course, our technology anchor. Welcome, Howard.
HS: Thank you, Judith. In this webcast, I invite you to join Judith and me in a lively technical dialogue with our guest expert, David Lambert, that will answer the questions you�d like answered about budgeting for IT infrastructure and ask those very important follow-up questions. You can join in this dialogue by sending your questions via e-mail to expert@cren.net anytime during this webcast. If we don�t get to your questions during the webcast, we�ll get Dave to provide an answer in the webcast archive. It has been said that an army runs on its stomach. While food may be the lifeblood of an army, more important than planes, tanks and ammunition, money is the lifeblood of an IT organization. Without it, there would be no computers, networks, programmers, help desk or even CIO. One might think that an IT organization�s focus must be on obtaining a technological infrastructure that we make available to ourselves and our users, but without a reliable source of funding, information technology support on campus would grind to a halt. Willie Sutton, the infamous bank robber, understood the power of money. When asked why he robbed banks, he said, �I rob banks because that�s where the money is!� IT managers obviously need an alternative source of funds to pay their vendors, employees and consultants. Of course, getting the money solves only a small part of the problem. The real challenge is budgeting the money to have the most impact on your college or university. How a campus sets priorities for spending IT funds, where it uses cost recovery, how carefully it allocates costs to projects and what financial controls it establishes is vital for its success. As good IT professionals, we know the importance of following the flow of information through a system. In our role as money managers, our understanding of the financial side of the IT organization depends more upon our understanding of the flow of money over time than bits over the networks. During the Watergate scandal, Deep Throat told investigative reporters to follow the money if they wanted to unravel the puzzling events. That�s also good advice if you want to understand how an IT organization really works. Money flows through everything we do and our budgeting process shapes that flow like dams and levees shape the flow of a river. It is tempting to give budgeting short shrift and to focus on the far more interesting technological innovations that appear every day. After all, we are in the information technology business because we are fascinated with the technology. But a sound budgeting process is a key to making everything else possible. Amelia Earhart said that, �The most difficult thing is the decision to act. The rest is merely tenacity.� IT folks certainly have lots of tenacity. All we need is some help in getting started, so put your Linux or PeopleSoft manuals aside for a moment and let Dave Lambert get you energized about budgeting for your IT infrastructure on today�s webcast of Tech Talk. Judith?
JB: Thank you, Howard, and your comments make me remember when we talked about the idea of the infrastructure, the IT infrastructure being like the physical infrastructure on our campuses and as we have always been teaching and learning in physical classrooms, the IT infrastructure is what is going to be providing us with the digital classrooms as well. So it�s just a really important piece of what we have in our future. So with that, I�d like to go ahead and welcome today�s expert, David Lambert, from Georgetown University. David is the Vice President for Information Services and Chief Information Officer at Georgetown. Since coming to Georgetown about five years ago, he has overseen the creation of the UIS organization there which is the University Information Services, which sounds like a nice comprehensive group for the infrastructure. In addition, David has worked on major initiatives such as the Residence Hall Wiring Project, GU�s sponsorship of the Greater Washington Technology Showcase, the launch of the Georgetown University mail system and the development of a fundraising program for information technology. Perhaps some of our participants will be interested in hearing more about that. Welcome to Tech Talks, David.
DL: Well, thank you, Judith. Thank you, Howard. This is my first Tech Talk and I�ve listened in on a couple and I�m looking forward to a good challenging discussion.
JB: Well, good. And we hope it�s perhaps not going to be your first and your last and that you�ll come back!
HS: Okay, Dave, to make it challenging, let�s hop into the middle of this thing here. You might have expected us to start at the top. Why don�t you talk about the process by which universities set priorities for spending money?
DL: Well�I�m sorry, did I interrupt you, Howard?
HS: Nope.
DL: Okay, good. Sort of get used to the format here. I think, in all honesty, there are probably as many overall processes as there are universities and I think that because we are a community that has so much diversity�community colleges, state colleges, private institutions, institutions that technology is imperative for and has been around a long time and those that are just discovering it. And I think in general you find a sort of a different mix of factors involved at every university, but I think there are some underlying factors you can sort of look to. Clearly, as you try to set the priorities, I think the first and the foremost one that has to stay front and center that many people often forget is that the IT strategy has to be closely tied to the strategy of the institution as a whole and there are horror stories all over the landscape of IT strategies that have gotten divorced from the university strategy. If big sciences are the key to the university, then IT strategy has to be shaped by that. If distance learning is important to the university, IT strategy has to be shaped by that. And so the key sort of number one element, I think, really has to be to work very hard to discover what the vision and the strategies of the institution are. Sometimes that�s not very easy, incidentally. And to make that sort of the number one driver. I think there are others that then become very important. Clearly the needs and interests of the various constituencies in understanding that as well, the administrative constituents in the university, the faculty. But probably more importantly, that often-forgotten constituency, the students. And so I think those two things are critical to shaping the dialogue about priorities and because that�s the case, the people that have to participate in that process is virtually everyone to do it effectively, from the executive leadership of the university that has to shape that strategy and you�ve got to bring your IT strategy together with their vision all the way through the stakeholder communities who are the consumers of your services and the users of your infrastructure.
HS: When we talked before, David, you talked about decision rules. Could you tell us what decision rules are and what they�re like?
DL: You�re trapping me in a corner as a political scientist now, Howard, which is my background�which is one of the reasons why, I guess, I continue to find the business of doing IT in universities fascinating because it�s an extraordinary study in political science.
JB: You had mentioned that CIO�s, that one of the most important things for the budgeting process was to identify and determine what that decision rule is, David.
DL: Absolutely, [inaudible].
JB: Share with the folks what that means?
DL: Well, you know, actually what I�m sort of trying to do is apply a concept from political science which those of us who are political scientists consider to be the master science, you know, where politics is really a process that subsumes an enormous amount, but more importantly, the way decisions are made that allocate things that people want. And the key to being effective in any system�and, you know, all systems are political because all systems allocate values. They allocate things that people want and they use different processes to do that. The one key political scientists have learned for years and years and political activists to be effective is you�ve got to understand how the decision is made. So the first thing you have to do before you shape a strategy for getting any decision supported, particularly those involving dollars, is to make sure that you understand how these particular decisions are going to get made.
HS: Can you give us some examples of��
DL: Well, I�ll give you an example of some types of processes that are used to make decisions. In some universities, they�re very much driven by the president. President makes 80% of the decisions and if you�re going to get an IT plan approved, the obvious strategy is, by gosh, you�ve got to shape your argument so the president buys it. In other universities that tend to be very, very distributed in the way they make decisions, particularly those�many great universities that have very, very strong colleges and very, very strong campuses, particularly multi-campus systems�most of the decisions that get made at the enterprise level are a byproduct of some decisionmaking process where those communities have to come together. Sometimes it is so very difficult that no decision gets made that�s not made�that isn�t a unanimous decision and if you�re working in that sort of environment, then you�ve got to go find everyone. In other systems, bureaucracies make decisions. There�s a set of rules, people have very clearly defined roles and basically you figure out what the system is and you appeal to it. In other systems, you know, decisions get made by coalitions, that the eight most important people, you ought to get five or six of them on your side to get water sort of running downhill. So there are different kinds of decision processes running all the way from very authoritarian ones to ones that often depend on unanimity and all sorts of levels in between. And it is very difficult to say the strategy and get a decision made when you don�t what that decision rule is. If you prepare, for instance, to appeal to a bureaucratic decisionmaking process, you�d get yourself set up to do that and it turns out to be the case that there�s only one decisionmaker that makes a difference, you�ve wasted your time. If you spend a lot of time courting your president, pushing, getting the president on your side or the provost on your side and it turns out that it�s really the deans and a coalition of deans that create the consensus around which enterprise decisions are made, then you�ve wasted your time. You�re better off to go spend your time working the deans.
HS: Once you�ve done this, once you understand what the decision rule is and you�re going out and getting funding, it seems like there�s a couple ways to deal with the budget. One way is to use some kind of overhead tax and do some cost recovery or to fund things from central administration. Could you talk about both those schemes?
DL: Well, there are probably more schemes than that and to fund things from central administration and to fund things out of overhead tax actually turn out in practice to be very similar kind of mechanisms. And they�re really�they differ only on the basis of the way funding is derived in an institution. For example, in a state institution, a good portion of the operating budget comes from a state appropriation which means it comes to the president�s office and the president has the opportunity to distribute that money to the deans in order to execute their programs. At an institution like Georgetown and many of our private institutions, the deans are very autonomous and the academic program is driven by tuition. And the deans are the ones that set the goals for how many students they accept and so if you want to have central money to do a project, in effect, the deans have to be taxed. And I think either mechanism�you find both mechanisms in place actively in a number of universities and which mechanism you use, I think, and can be effective at really depends on where the money starts. You said earlier, �follow the money.� The most important thing to know when you�re following the money�or the two most important things�where does it start? How do you get it and where does it end up?
HS: What about end-user participation, that is, trying to distribute the costs to end users? Are there cases where that makes a lot of sense and other cases where the money really ought to come from some central source?
DL: Well, again, multiple approaches. I think the strategy that I have always sort of tried to take and been very effective here in a number of settings with is very clearly that the more ubiquitous the service--you know, the services that have to be available to everyone, the services that really require leverage, that are more cost-effective when they are done centrally, things like networks and big administrative systems are the sorts of things that it is imperative that the enterprise step up and fund because you�ve heard the old statement, �You can�t market infrastructure.� You really can�t. I mean, you can�t walk to the average faculty member and make an argument for them to put millions of dollars into an infrastructure network when their departmental meetings center around whether they want to spend $25 or $400 to pay to send the faculty to a meeting somewhere. And so it�s important for those kinds of issues to be dealt with at that level. On the other hand, sort of the micro purchase decisions�what kind of computer am I going to have on my desktop? Are we going to get a printer down here? Are we going have a student in our department who helps us? Are we going to hire departmental support specialists to make sure that we get good support here? Those are the kinds of things that my sense is, in most systems, the more responsibility gets taken at the user or departmental level, the more effectively those funds are deployed. It�s awfully hard to sort of plan centrally for the kind of support that varies so much from department to department, so I sort of have this continuum in my brain that sort of goes from the network and the core infrastructure all the way out to the individual, sort of end user application. And I think the level of distribution of control as well as the sort of source of funding kind of needs to follow that distribution.
JB: David, what about following that continuum�which sounds good, from the network infrastructure out to the end user�do you have any example of the kinds of services that might, in fact, be allocated for the deans on a per-person basis, something that would not be funded centrally but would be some level of cost recovery based on the number of people or usage? DL :Oh, I think there are several things and you could go at it from two ways. Number one are sort of specialized services, you know, the sorts of things where say, for instance, the demography department needs to have access to some very, very specialized GIS kind of equipment and they�ve got 15 faculty members who have to have very, very sophisticated skills. It would have to have very, very sophisticated skills in database and inquiry and management, where I think it�s appropriate for deans to step up to the issue of how they put the money on the table to fund hiring a support person. Coming at it from the other level, I think it�s also the case that one can draw a line between those things that are sort of uniform services that should be provided for everybody and specialized services and I think it�s entirely fair game to cost recover at the individual level for those specialized services. Say, from a strategic perspective you�d like to provide ten megabit Ethernet to every desktop on campus. You know, everyone should have at least that much networking capacity, when you really want to go after enterprise models for funding that. But there are some departments out there that need gigabit desktop access and I think it�s entirely fair game to say to those people, �This is a unique need, it�s not what I would sort of call a baseline service at this point, and you therefore should contribute extra for doing that.� Rather than losing control of the network by having that department go out and buy their own equipment, I think a good mechanism for that is a cost recovery on a per head basis.
HS: The next question I was going to ask actually has been asked much better by some e-mail that we have from Doug Emmons from the�oh, he says he�s a computer systems administrator at University of North Carolina in Chapel Hill.
DL: Hi, Doug!
HS: Doug says, �As a School of Education at a university suffering from major statewide budget cuts��which I�ll bet is not unusual!
DL: Not now!
HS: �What suggestions do you have for creating and sustaining life cycle replacement budgets for IT infrastructure?� He says he�s mainly concerned, as I�ll bet a lot of people are, with being able to replace desktop and laptop computers for his staff and faculty on a regular three year basis.
DL: Well, we�ve worked very hard at Georgetown�and I can certainly speak to my experience�to achieve a university-wide commitment to three-year replacement cycle. And some universities have been very successful at that. I know, for instance, that Indiana�a school I used to be at�has actually, I think, successfully deployed a university-wide approach to desktop replacement. We�ve found at Georgetown that the real stakeholders�that there are two real stakeholders in desktop replacement. One stakeholder is the IT service organization and you know, as I think we all learn, when every desktop is different and some are seven years old, some six, some five, that it gets harder and harder to support them. So uniformity is very important to us from a support perspective. The other stakeholders are the financial managers at the departmental and the campus level and I think my own experience has been that the ability to successfully deploy life cycle replacement lies in the development of the relationship and a compact between the campus service providers and the financial managers within the university. And if you can sell the financial managers on desktop replacement�and it�s not a hard sell. You have to approach them differently. You have to let them into the process, you have to educate them about the process, help them understand where dollars are really saved, that they will become extraordinary allies in helping develop desktop replacement cycles. And in point of fact, depending on how the financials are managed in your institution, you may have to go strike many deals. For instance, at Georgetown, we have succeeded in getting three year desktop replacement cycle for all of our main campus faculty and staff. We have succeeded in getting three year desktop replacement cycle for nearly all central administrative units except a couple and in both of those cases, you know, it�s really because we�ve not been able to forge the right partnership with the budget managers in those departments. So you�ve got two ways to go about this, I think. One is to try the strategic enterprise approach�and I will tell you, in my experience, that�s really hard. You�ve really got to have support at the highest levels. Everyone has to be committed to it. My own recommendation is the more practical way is go build your working relationship with the people who have the money and really focus on that, really focus on helping them understand the value of replacement cycle, make them part of the process. And one of the things we decided not to do at Georgetown, for instance, is the money that funds replacement cycle is not in the central organization budget. In some instances, it was already in the central organization budget. I gave it back and basically gave it back under the condition�and we�ve actually got this in writing�that it go into special accounts within the departments so it never really burdens the central budget. The departmental business managers feel like they still have control over it.
HS: But does it look like amortization? I mean, is that what you�re doing? You�re saying, �Take the machine and we�ll worry about a third of the cost of it each year?��
DL: Yes, well, I mean, how it reflects in your amortization schedule is something that the Finance Office has to deal with. But from our perspective, what we have done is to come up with a program whereby x number of departments sign up for this, we go on a three year replacement cycle. We�ve tried very hard to stagger them a bit so that we�re not doing one every year, and that�s proved to be a little difficult. We let some go four years and we actually replaced some sooner so we could get them on a staggered cycle. The central organization actually buys the machines, working closely with the business managers, working off of a standard system, and the goal is to get it turned over and out of here in three years. When it�s done in three years, we don�t keep it around and give it to graduate students. We have, in fact, we tried leasing for a while but the business was so irrational that we finally went to buying. But when it�s gone in three years, we have someone come and buy it from us and get it out of here because we don�t want the old technology sitting around the university driving our programmers crazy, trying to figure out how to support it.
JB: Let me just jump in here for a moment and remind all of our listeners that now is a very good time to send in questions to you, that you�re happy to answer them, and to send questions in to exper@cren.net. Let me do a follow-up question on that replacement cycle. Do you have any suggestions or ideas on what if an institution truly has gotten their budget cut, the budget that would in fact normally go to replacement of computers? Have you any hints or strategies for perhaps that kind of a situation as to what one might want to do?
DL: Well, I guess there are several things that I could recommend. Number one, I would�being in that situation, I would work very hard not to give up replacement cycle. I would look for all kinds of creative ways to try to do that. I would try to bring departmental money into play with a program. For instance, Syracuse University�and I haven�t checked in with them for a while, but I know when I was in New York and talked regularly to them�had a program whereby they supported the replacement cycle not by buying the whole machine but by contributing a set amount to the machine. And so they actually utilize not just departmental funds, but personal funds. Often they sort of tried to encourage faculty to put�they tried to get enough money in the system for a basic machine and then tried to encourage people to supplement that with private and personal money. The good think I always thought about the Syracuse program was you were in a little better institutional control of how much money you were spending. The thing I would really try to do everything I could to avoid sort of at this point in the technology curve is to not, therefore, go back to four years and five years and six years because catching back up once you�re on that cycle is a very, very difficult thing to do.
HS: David, what about using outsourcing to save money? Is that a good idea?
DL: Well, I will have to say that in the first place, I think, in our business, outsourcing can be successful in places. I think you have to do it very, very carefully and strategically. But my general experience with outsourcing in universities, and I�ve been engaged in various forms of it, both initiatives that I�ve started and initiatives that were rammed down my throat by people within the institution, and my general experience has been there are very few areas where you�re able to initiate a contract with an outsourcer that is really more cost-effective than you can provide the service. The exceptions that I see most regularly that are successful are certainly computer sales, computer repair, implementation of very unique administrative systems that tend to be black box kinds of things that take very specialized expertise that is difficult for you to get your hands on. I think sort of some fundamental things like wiring, for instance, every university I�ve been at we�ve been able to drastically reduce the cost of pulling wire by establishing a contract with a good outsourcer and trying to maintain that contract for a significant time period because they sort of take away from us all the overhead of having to deal with unions. If you find the right company, they really do training well. There are some real mistakes I�ve both done and been around in outsourcing. For instance, outsourcing desktop support. Throws the baby out with the bathwater. The second you�re not providing desktop support, you�ve lost your ears and therefore you have no clue about what it is people want. Outsourcing helpdesks. I�ve seen people do that. I�ve not seen any very successful instances of where that works. I�ve seen people try to outsource computer operations with some reasonable success, so I really think there are some places but you have to be very careful about those places. You have to be very careful about the partner you choose as an outsourcing partner and you have to write that contract very well. I think what most people discover is that in universities, we work at such low marginal rates that it�s very difficult for an outsourcer to deliver an equivalent service for the same cost.
HS: Okay, we have an interesting question from Mary Doyle at Washington State University and Mary asks, �Is there any advice that you can give about an attempt to move from a charge-back model to a centrally funded utility-like model for something like network services? Assume there is no central surplus that can create a new utility budget.��
JB: This is a good challenge for you, David!
DL: Whoa, how do you�this is an issue that�s going to be on everybody�s plate, particularly around telephone systems, for instance, which have in nearly all of our universities been set up as cost-recovered models. And right now, the cost structure underlying those is changing so radically that you don�t know whether you�re coming or going. I�ll give you an example. When I first came to Georgetown, our university long distance phone bill including the hospital was almost $3,000,000. This year, it�s going to end up roughly $420,000.
JB: That�s a lot of money saved.
DL: I spend more than $420,000 collecting $420,000 when I do that, so I think we�re going to see all kinds of businesses that we have set up as cost recovered businesses in the past begin to be impacted by financials in ways that we sort of have to step back from that and look for other unique models. The dormitory-based voice business, for example, I think is over with. I mean, 90% of students are bringing cell phones. Our ability to continue to make them pay for campus phones is going to get larger and larger, but yet we can�t walk away from them. And in many instances, universities have made the decision to use those revenues, the net revenues to supplement and augment other services. I know universities that are paying for network structure in the dorms by telephone charging and there�s going to be very little ability, I suspect, to do that kind of thing in the future. So I really think the key to this is�and in fact, right now at Georgetown, we are getting ready to launch a complete reexamination of all of our cost recovered programs and there are actually instances where we�re going to try very hard to get the university to buy into the notion that things that have been cost recovery need to be moved into the general allocation model. And what we�re going to try to do to balance that is to suggest that some other things might start to come out of the allocation model into a cost recovery model, so for instance, we would begin to move�we would begin to get out of the charging for long distance telephone business and we would begin on the network side, for instance, to begin to introduce charging for very high speed network pipes, particularly since many of the consumers of those are research teams who have the capability to access external funds. The key, again, I think to making that migration, at least from my experience and what I�m now trying to structure at Georgetown is you�ve got your partners, your most important partners in a university in being successful aren�t the president and the provost and the senior vice president and the deans. Your most important partners are the CFO�s and the departmental business managers. And if they�re on your side, if they understand these issues and they�re willing to work with you, 85 or 90% of those decisions about how you structure that can be made at that level.
JB: David, what about in this upcoming review of cost recovery and reallocation at Georgetown? Are you going to be talking about a service such as supporting wireless and PDA networking?
DL: We sure are.
JB: Is that part of what�s driving [inaudible]?
DL: Absolutely. And we want to be involved in those technologies. Some of them will become ubiquitous and we really need, when something�s becoming ubiquitous and really needs to be there for everybody, the first question you have to ask yourself is, �Boy! It�s going to be hard. I can�t just add a new layer of service on top of my old layer of service because I�m forcing the university to pay for two or three levels of technology.� The first question you have to ask yourself is, �If I�m going to absorb the cost of a strategic wireless deployment, how am I going to get rid of the costs of some of my old communications services?� Because it�s unfair to the institution to put them in a situation where, even if you�d get the money, they have to pay for multiple systems. And so again, that�s moving toward that model of an allocation process for those things that everybody uses and cost recover process for those things that not everybody uses, with some ability�very important, with some ability to perhaps augment and keep low the costs of some of those new technologies that you really, really want people to use, that you would hope if you�ve done your job correctly two years from now, they would become ubiquitous services.
HS: It seems from a user�s point of view, going from a charge back scheme to a centrally funded scheme, that�s wonderful! Users have no problem with that. But it seems the other direction gives users a great deal of problem. In fact, it seems like it�s an explosive thing for IT groups to do. If you�re taking something that you centrally fund today�let�s say backup�and tomorrow, you decide to use a charge back scheme for backup, don�t you think you�ll have your users up in arms, marching on your administration building?
DL: Well, my experience with that is shaped by several�knowledge of some things and several experiences I�ve had. Number one, a university for the most part is a closed financial system. And while it may resemble sort of an economy and one can certainly identify an economy in it, it�s closed. And so what goes up on this side comes down on this side. In all honesty, if you�re going to take a general service and move it from centrally funded to cost recovered, then you�d better be prepared to take the money you get from them now for those services or parallel services and move it back out to them. Otherwise, you won�t get very far, except when you�re doing things, I think, that are so compellingly useful that people will do it. I remember, for instance, when I was at Cornell. We moved out two services that were completely charged back. One was our first, our IP modem pool which many people have the alternative of outsourcing that, but in Ithaca, New York, that�s not a very likely outcome. And backup. And in both those cases, we introduced them as cost recovered services and did it successfully as a combination of two things. Number one, very flexible models for cost recovery so trying to find ways of packaging it that everybody could play so we had individual subscriptions, group subscriptions, etc. And secondly, it was a service that was so necessary that while they would grumble about it, they really didn�t have any choice. And backup was the same thing. I think the combination of those things worked effectively. But I will say one other thing. I will tell you about another lesson I�ve learned and I hate to not be as directive as I might like to be here in terms of saying there�s any one answer to this, and coming back to the notion that it�s all relative. Universities and the way universities are funded are very, very different. For instance, there were lots of things that I could charge for at Cornell because there was a significant amount of money at the departmental level, lots of grants, people got money, department chairs hung onto money. Whereas at Georgetown, most of the money gets caught at the campus level and there�s very little money out in the departments. In fact, there are departments all over the university�Georgetown has scores of departments that don�t even really have an administrative structure and don�t even have budgets attached to them. And trying to do cost recover for a service in that environment is something that you want to think very, very seriously about. So you have to ask where the money is. But I will also say that it�s often the case that any university, which is a very tightly budgeted organization, where everybody is screaming they�re broke, it is amazing sometimes what�s out there in the micro economy, what�s out there within the purview of an individual faculty member to make a decision about if it�s less than 150 bucks a year. And you can do backup for less than 150 bucks a year.
HS: Yeah. Isn�t there a problem that you run into�I mean, let�s say that you were funding backup centrally and you decide to do charge backs. Aren�t some people going to say, �It�s just too expensive so either I won�t do it or I�ll go outside and do it,� even though it�d be better for the university to have it done centrally? So aren�t you going to have a danger of encouraging behaviors you�d like not to encourage?
DL: Yeah, I think you are. I mean, whenever you move�I think that�s true. Whenever you make a decision to use the micro economy to fund a project at a university, you are making it much, much more difficult for that service to be applied strategically in the institution and, you know, often�and I�ve seen a lot of people do this very creatively, I know. I�ve seen a couple of masters practice this process where you can announce or develop�I would also like to say, the challenge of rolling out a new service that�s cost recovered is many times easier than moving the service from centrally funded to cost recovered.
HS: I can easily believe that.
DL: And my inclination when I try to do things like this is to do them at a turn in the technology so it doesn�t look like the same old service. It looks like a new service.
JB: That�s an awfully good idea.
DL: It�s just
HS: [inaudible] file somewhere.
DL: It�s just good marketing. You don�t charge for old modems but when you put out�so [inaudible] modems several years ago, people moved to charging and it wasn�t the same service, it was a different service.
JB: We had another�I�m sorry to interrupt there, David, let me let you finish what you were saying then.
DL: No, that�s fine. Just sort of, I think, in summary, you just have to be very�you have to be very flexible. You have to market a little bit. You have to understand what your market is and you have to send the message out there, and those are always tough decisions to make. But sort of harking back, Howard, to the quote earlier that you made, my sort of parallel of your quote is there�s no right decision. There are more importantly decisions you make right. And when you fight your way through one of these tough issues about how you�re going to fund something, it has a whole lot more to do with how you execute the strategy you�ve chosen than it does with the strategy you�ve chosen. Once you�ve separated the bad choices from the good choices, two or three strategies will get you there. But whether you get there most often depends on how well you execute it.
JB: Okay, at this point I would like to bring in another question that we do have from Ed Goray from the University of Illinois at Chicago, and it�s an area we haven�t talked much about yet. His question is, �Who should fund instructional grants?� And by that, he means giving money to faculty members interested in developing educational content and courseware so money for TA�s, hardware, software, time release, etc. What is your experience with that type of infrastructure grant provisioning?
DL: Well, my experience with it is that creating an environment for that to happen is the job of the CIO. Funding it is not the job of the CIO for two reasons�one because it takes money from getting the infrastructure in place and, two, because if you go very far down that path, all of a sudden you find yourself stepping all over the toes of provosts and deans and that�s not particularly a bright thing to do. And particularly, I think, you can co-opt the provosts and deans and I�ll give you an example of what we�ve done at Georgetown which I think has worked pretty successfully�and believe me, the situation at Georgetown was a catch-up situation. There was very little here when I came four years ago. And only in the last two or three years had there been any focus on providing support for faculty development, the technology, at all. And what I did was, we created a program to do that. We call it the Center for New Designs in Learning and Scholarship and I have a unit�we went out and recruited one of the very best, brightest faculty members to do that. I did this collaboratively with the provost. The head of the Center�s reporting line is to the provost, but he�s co-located with our staff. We staff that enterprise from multiple sources�the library has put some people into it, we�ve taken some of our instructional technology staff and put them into it, and then we�ve let him slowly develop a small budget to do that, including funding that�s come in agreements from the deans and provosts to fund internships, to fund people. We�ve also gone out�it�s one of the areas where we�ve been successful in fundraising. We�ve pulled in enough money from external sources to fund what we call the annual Candle Symposium where we bring on every semester 30 faculty members�between 25 and 30�into a semester long symposium where we pay release time for them to participate. We take them through a symposium where they learn about educational technologies in pedagogy. They think about a project they would like to do, and then at the end of the symposium, they have the option of moving on to a second semester in the symposium where they actually execute their project. And in general, what that will be, for instance, will be the redesign of an introductory biology class. And we actually give them a stipend. They get release time from their department chair to redevelop the course. We provide them with a TA to help them. We provide them with a place to come work where they have help and we actually pay them a stipend to do that and the money for that has come out of the direct commitment that the provosts and the deans have made to seeing this kind of change happen in their departments and the careful application of some corporate grants. We�ve worked very closely with our director of corporate development to nurture relationships with a couple of companies that have good relationships with Georgetown for whom this project looks very good on their resumes of support for higher education. So I really think it�s the sort of thing that you don�t run off and do by yourself as the CIO. There are too many pitfalls. It�s the kind of thing that you get a dean, two deans, three deans and the provost to buy into with you and you let them think�you know, you�ve got to support it. You�ve got to be the backbone. You have to make it happen. You have to put some of your money on the table but you do it in a way that they feel very much like the program�s theirs.
JB: Okay, Howard, did you want to ask the other question that�s come in on laptops?
HS: Actually, I have another question I�d like to sneak in here.
JB: Okay. You�ve been kind of quiet, so I figured��
HS: There�s just so many things here that I could�when we look at the infrastructure, Dave, what do you see as really the major infrastructure expenses and do you see that changing in the near future?
DL: Well, yeah, I think there have probably been several changes and there are likely to be more. When you look at where most institutional investments are made, when you look at it from a programmatic perspective, you�re going to see that a good portion of that goes into maintaining�once you�ve added up the cost of your network and your data center and the staffs that support that, you�re really eating up anywhere between 25 and 50% of the funds�most people are�of the funds that they�re going to have available. I think the next element of infrastructure that then becomes the next biggest bite in the process is your basic systems infrastructure, your administrative systems infrastructure. Those are very, very big ticket items and you see support costs normally and academic computing costs sort of falling in line somewhere behind those two big things that sit there and eat most of your money. So when someone asks me, �What�s my infrastructure?� it really is my network and its ability to get on and off the campus, what I call my network services which are my basic mail, authentication, authorization, calendaring, the sorts of things that tend not to be�some people think of them as end applications but they�re really closely interlaced into the infrastructure�and my core administrative systems. And that�s what I think about when I think about infrastructure. There is another one that always comes in place for us as CIO�s and that�s where do we live, how are we housed? And depending on where we are, we often find ourselves spending significant portions of our budget to make sure we have places to live and good offices and good support for our staff. If you then look at that budget on a sort of a natural budget category, you know, I think nearly every IT budget in the country, human resources is somewhere between 50 and 80%. I�ve worked in organizations where it�s been as high as 75 and 80%. Currently the situation at Georgetown now is it�s about 50, 51%. The only reason that�s really true is because we�ve been on a significant capital deployment and so the debt service portion of my budget for paying back the loans that the university has provided to undertake major capital initiatives like administrative replacement and network infrastructure replacement, you know, constitute about 25% of my budget here. So in terms of trends, Howard, and you asked that question, I think one trend is obvious and that is across the board, dealing with administrative systems is eating up increasingly high portions of university funding in information technology.
HS: The cost of software?
DL: Well, it�s not just the software. I mean, in an information systems project where you�re replacing basic business systems, the software�many universities thought that the total costs were software, but they�re really not. The software is cheap compared to the consultants you have to have, the amount of backfilling you have to do to get staff in the business offices to participate in the projects at the level you need to have them there and to change over server infrastructures and database infrastructures that go along with replacing legacy systems that are on second and third generation equipment and database architectures with systems that are on more modern architectures. Those projects are big, they have�there are scores of universities�all you have to do is look at the number of people that have licensed PeopleSoft or SAP or SET to realize that most all new spending�I think, Howard, there�s no documentation to support this. This is just from where I sit. Most new spending in information technology in universities is focused on administrative system replacement.
HS: You just talked about some really big, expensive things and we�ve seen a couple people talk about the fact that their budgets are tight, maybe even decreasing. If that happens to you�if you�re a university and your budgets are cut, how do you go about deciding what is it that you�re going to cut?
DL: Oh, that�s the $64,000 question! I mean, that�s the question that costs a lot of CIO�s their leadership mandate at the universities
HS: [inaudible] wrong thing.
DL: Well, you know, one of two ways. They cut the wrong thing, or two, they cut the thing long and I would just almost�my experience has been, and as I�ve watched other people work, it�s the latter that is the most damaging to leadership credibility. That is, to go through making those decisions in a way that appears to other constituencies in the university to have been abrupt, disconnected, intended to [inaudible] them. You know, for instance, the first reaction of any good CIO who wants to fight the budget battle is to say, �Sure, I�ll cut money. I�ll cut the helpdesk hours back 20 hours a week.��
HS: I�ll cut something that�s really important.
DL: I�ll cut something that�s really important.
HS: So you�ll restore the funds.
DL: And you know, my gut tells me�and I�ve always sort of worked off the principle that says, look, if a university is really in a situation where it�s forced to cut budget, the technology organization has to be a team player, has to be a good citizen, and you�re always better off to work very, very closely with your peers to do that in the least harmful way you can and assume that when the time comes that the budget begins to loosen up again that your good citizenship will be rewarded. Now, I will say, I�ve seen exceptions to that, but for the most part, I�ve been reasonably successful by playing by those rules.
HS: Okay, we have a question from Robert Brickley at Penn State, and Robert says, �What advice would you give for budgeting for laptop computers for students? What kind of models would be worth looking at?��
DL: Well, you know, I don�t think there�s any particularly easy answer for that.
JB: That�s probably why he asked it, huh?
DL: Again, my sense is that that question has a different answer at an elite private university where the student is paying�the parents are popping that $35,000 to $50,000 a year just to have their student there than it does at a community college where the funds�where the metrics are just completely different and the types of students are radically different. My own instinct has told me over the years to stay away from mandatory purchase programs because there are�and I feel this way about that because I always try never to get caught in churn of technology. And once you start putting institutional money in one way, shape or form into what represents a private purchase that turns over frequently, I think you run into something that just keeps eating and eating and eating and eating away at your budgets.
JB: What about the other half of this question, David, was, should the university purchase laptops and then sell them, basically, to the students or that concept of the computer store, large purchase. Do you��
DL: Well, you know, again, my instinct tells me that the time of computer resales and computer stores is past at universities, that these have become commodities. That doesn�t mean that you don�t want standards, that you don�t want volume discount contracts with suppliers, that you don�t want to connect your campus purchasing structure to their e-commerce website. All of those kinds of things, I think you want to do. In my mind�and I know some universities continue to own resale operations that are viable, but my general sense is the margins in that business have gotten so small that if you attempt to do that, the mere overhead causes you to put the prices up high enough that the students can go to Dell Online and get it themselves and save money, and they will.
JB: So that�s basically bottom line, your recommendation on that one.
DL: Yeah, that�s a personal bias and one that I�ve [inaudible] committed to.
JB: The other half of that is, you know, we used to be building all kinds of labs on campus for students. Where are you with Georgetown on that? What percentage of your students at Georgetown supply their own computing needs?
DL: Actually, the two questions aren�t all that disconnected. I mean, we had a belief some time ago that the individual purchase programs drove down demand for on campus labs and classrooms, and my experience has been that that�s just the opposite, that there may be some nirvana out there in the future where there is such extraordinary ubiquity that there is no value for on-campus facilities any longer. I haven�t found that to be the case. In fact, I�ve found it to be the case that the more students compute, the more they have their own machines in their dorm rooms, the more they want to be able to stop as they�re walking through the hallway in the library and log in and check mail. The more faculty want to require complicated software that needs to have specialized machines that run, the more people want Unix and special purpose things, the more disciplines want to have specialized facilities for the art production, for instance. And so I don�t think that there is an inverse relationship between computer ownership and the need to put stuff out. In fact, I would almost argue, my experience has been that it�s just about the opposite and I don�t think we�re close to the point where we can stop worrying about our public facilities because I think there continue to be good reasons for those kinds of service to be available. At Georgetown, using Georgetown as an example, we are actually in a public facility growth phase.
JB: Interesting!
DL: And we�re they�re because we�re catching up. This is not something that had ever been done very much at Georgetown in the past so we�re adding 100 7x24 hour computers in a period of less than a year. And again, who uses those? Well, people�students wanting to come work in groups and not hang out in their dorm rooms, students wanting access to specialized software, off campus students who come to campus and work and maybe the wireless campus with portable PC�s is going to fix that problem and maybe we�ll just simply replace these online labs with wireless hubs and wired lounges in student unions. And I think we�ll do some of that, but I don�t see the public computing lab or the technology centered classroom going away anytime soon. We have some deans who really wish it would. Of course, it�s costly. And it�s competitive, you know. Georgetown has gotten a lot of horror stories of really good students that we�ve lost because parents come here and don�t see those kinds of facilities that they see at other schools, so we�re on a sort of an up curve in terms of our investment there.
JB: Okay. Howard, how are we doing? Did you have a closing question for Dave?
HS: Yep, I certainly do! And we hope that he can answer it before we actually close here.
JB: Okay!
HS: David, in doing budgeting, could you tell us if there are some traps that you can think of that we should be sure to avoid?
DL: Well, I think number one trap is forgetting about the operating tail and not budgeting the�not recognizing that software maintenance and hardware maintenance are increasing eight to ten percent a year. That really, really comes back to haunt you. Number one problem.
JB: Okay, and before I wrap up, David, any final comment for our listeners? There are a number of questions we couldn�t get to so we�ll be sending those to you to answer offline.
DL: Well, you know, I think my only�the thing that I�ve learned over a long time is there�s some principles but there�s no single answer and the most important thing for everyone to do in this regard is learn as much about the institution as you can. What are the goals, what are the strategies, how�s the money spent, where�s it come in, where�s it stop, how are the decisions made? Understand the system you�re working in.
JB: Okay, so we begin and end with the decision rule. With that, I�d like to thank everyone for being with us here today and you can, in fact, still send follow-up questions to expert@cren.net. Thanks so much for being with us here today. Join us again in two weeks, on February 21, for an update on Copyright and Intellectual Property with Rodney Petersen from the University of Maryland. Steven Worona from EDUCAUSE will also be on hand, co-hosting with Howard Strauss that day. Many thanks to our CREN member institutions for support of today�s program; to our Tech Talk expert, Dave Lambert; to technology anchor, Howard Strauss; to Terry Calhoun, Tech Talk web guru; to Jason Russell, Gayle Terkeurst and the support team at Merit; to Susie Berneis; and finally, a thanks to all of you for being here. You were here because it�s time. Bye, David, Howard. Bye, everyone. We�ll see you on the 21st.
HS: Bye.
DL: Bye-bye. Thank you very much.
END OF WEBCAST
