Comments on the Chargeable Services Draft Proposal

Peter Marshall 1997-02-05
<http://www.uwo.ca/~peter/charge.html>

Key Points

  1. We need to get the perspective of the academic community into the paper.
  2. We need to play up the types of service that we'll continue to provide without charge.
  3. We need to be very clear about why we're charging and indeed if we should charge.
  4. We need to understand the changes in funding and budget models that ITS initiated changes will require.
  5. We should be classifying broad categories of service rather than specific services.

General

We should add to this document a spin that will appeal not only to the Administrative parts of the University but also the research and academic community. As it now stands, most faculty members would just assume that it doesn't apply to them. As I understand it, the proposal is that we should be writing a document that applies to all.

We should play-up the services that are provided from the base budget of ITS. If we are going to ask users to pay for some services it really helps that they understand that there are a lot of services that are provided without additional charge. I don't like the use of "FREE" since that is misleading and somehow indicates that they are less important. It is these services that are the most important, so important, in terms of infrastructure and all the other services, that they cannot be left to the vagaries of pay-per-use or pay-per-hour. Better descriptions might be "centrally funded", "base-budget funded" or "pre-paid".

Perhaps a survey of the best Universities in North America would provide some perspective on how we should handle the question of fairly providing services. Trying to apply the ITS Hardware Services model to all ITS services might not be the best or even a workable approach. We should examine why it has evolved in the way it has evolved.

On the Introduction

I have some trouble in relating the initial paragraphs to the addition of charging in some areas. How exactly does the addition of charge-back help with the work-load? Is it really the case that there are many things that are being done now that won't be done under the new scheme because they aren't important enough to the customer for them to pay? If that is the case, then why is ITS doing them now when we, by virtue of being in more complete control, could choose not to? I've probably missed something here, so it might help to add some bridging text that explains it.

Related to this is a question that came up at a meeting of SLB LAN users recently. If departments need to pay for services, they need to have extra budget to do it. It was suggested that the obvious place to go for that money is to extract it from the ITS budget, redistribute it to the departments with the idea that ITS can then "recover" it. This doesn't help with the shortage of resources! Perhaps there is a simpler way to go about finding new money since this seems rather convoluted. And how does this relate to the individual researcher or graduate student that needs our services?

Let's be clearer about the reasons for charging for our services. What is the mechanism or policy for choosing which services are chargeable and which aren't?

Limiting Services?

     Are they services that we don't think need to be done (or that we
     don't want to have to do) so we make it harder for customers to
     get them done?  Is it to encourage people to use outside services
     rather than ITS?

Generate Revenue?

     Is it to make some money to hire more people?  Where do we expect
     the money to be transferred from?  Would it be easier to go after
     the direct source instead?

     If we are trying to find money to hire people, is a fixed flat
     rate for a wide variety of services really the best way to bring
     in the right people to do the job> Or are we aiming only at the
     low-end of the market?  Much of the consulting done by R&I isn't
     $49/h work.  If there is an implied subsidy in the rates, then
     that should be stated clearly and up front.

     On the other hand, are we pricing ourselves out of the market?
     Are there good things that could be done that won't given the
     disincentive of charging?  Is there a fund for subsidizing good
     projects?

Is it part of a budget redistribution scheme? What are the goals for that redistribution?

As we currently understand it, the PeopleSoft project has or is about to hire about 8 new people. This and other recent hirings at ITS will not go unnoticed. So is ITS really down 15 people or have we just gotten rid of some people from various areas and re-hired into new positions -- effectively shuffling people to match priorities? Weak or misleading arguments shouldn't be included in this proposal.

On the Recommendations

The document states that we can no longer afford to subsidize any client area and yet services based on Citrix, mainframe and PeopleSoft are specifically listed as non-chargeable so we are presumably subsidizing the client groups that use these services. If these are strategic initiatives that are funded separately then rather than listing these as the services that are provided "FREE" then we should state the types of services that will be provided at no charge, give current examples and define a mechanism for getting items on this list. Basically the categories of service with some current examples.

There seems to be a significant skew towards making services for administrative units (especially anything to do with the PeopleSoft project) exempt from charging. What similar break are we offering academics? What criteria are being used to determine whether a service isn't charged for?

As an example of how the draft list gets too specific, consider the Citrix setup service: does that mean that we charge for X-window terminal set up service, what about NC setup service? These are all services in a similar vein and will appear to be inconsistent unless we can provide an umbrella that holds one and not another.

On the "FREE" and "CHARGEABLE" lists

We should have a category for helping people transition to new systems. Certainly the work to move people from the old Cray to the new Cray, off the VAX, from the MVS system or to the new e-mail system could then be lumped into this category. And that may allow us to use PeopleSoft as an example rather than a specific service. But if you take a wide brush, then it doesn't leave too much that can't be defined as some sort of transition to a new system.

There is no mention of the types of workstations that would be supported as standard. Have unix workstations been considered? If a researcher through our efforts purchases a Sparc Ultra (which we're very comfortable in supporting) rather than an machine from HP (which we have no experience with), then I'd like to be able to reward that researcher with a better level of support. Perhaps a cost break could be a part of that.

"Configuration of unique peripherals": I'm not sure why a scanner is a unique peripheral -- what isn't a unique peripheral? The first 22MB 5.25" disk drive on campus? A CDROM server? The example should be clearer.

Is it decided that dual boot workstations aren't part of the ITS standard workstation? Are unix and Mac workstations a part of this definition?

Why would we make password changes free? That sounds like exactly the type of service for which people almost expect to pay a nuisance fee. On the other hand, there is no mention of strategic consulting in the "free" list -- something that people probably won't pay for and ITS and the University will pay the price further down the road. (I don't think that initial consultation covers the area adequately, I'm thinking of working on committees to rebuild the computing infrastructure of the faculty of education or to help define a new program in Informatics in the Faculty of Medicine.) Many other in-depth consultations really do save us and the University as a whole time and money down the road. We don't want barriers in place that make it more difficult to start off a project on the right foot.

The pre-paid list should include web-hosting for official units, perhaps some limited help with preparing web sites, providing usage statistics, standard web information indexing and general support of campus Information Providers.

The whole category of network liaison services seems to have disappeared. Chargeable or not. There is a reference to the Network Handbook, but it doesn't show up as a category or example in either list.

ITS provides a usenet service and a facility to set up new local newsgroups at no charge. We also currently provide mailing lists for the campus community, again at no charge.

Where does consulting lie? I'm thinking of the sort of work that Val, Patrick, Bob and Debbie do to help people utilize campus computing resources effectively. Is this customized support? Are we intending to charge out for high performance computing consulting in the future? Have we considered the question of ability to pay especially for graduate students and non-science researchers?

"Local installation of software (not on the server)": What is "the" server? Don't we charge for installation of software on servers? Or are we talking about some specific set of servers?

We're very early on in the provision of services for teaching and I'd like to leave charging out of the picture for now. It only skews things and means that good ideas can't fly. Perhaps we could include this work under a pilot project category.

It could be argued that the elimination of charging should apply to all aspects of ITS's work -- i.e. use the library model rather than the ITS Hardware Services model. When we charge more than people are willing to pay for hpc or panther cycles then researchers don't do their work on those machines. A similar situation will exist for many people as we charge for things that they don't have budget for and they aren't interested in working to get that money because it takes a full budget cycle and lots of effort to get a $2000 charge worked in...

Could a survey of our "market" tell us for what people are willing or capable of paying? We have to remember that at a university we're not operating in a free market economy. How much reform of the funding processes will be required before it will be possible to realistically charge for services? If researchers don't have the funds to pay for our consulting services, then do we just turn them away? Perhaps we need a way to assign consulting resources to the most promising and best projects, but is "ability to pay" the right measure? (We could assume that the best projects are the best funded, but that assumes a very different world than the one we live in where many small projects are not funded at all and yet they get done. We need to learn more about these mechanisms before we reform the charging mechanisms.)

On Rates and Coverage

There seems to be a discrepancy between only handling visits to University owned property and several references to Affiliates. Does the argument about insurance apply to these separate entities? How about departmental machines located in hospitals and the research park?-- who is responsible for them? (Editorial note: Elborn College is a part of the University proper, not an affiliate.) If we don't go off campus, then why do we have an External rate? Will ITS Hardware Services also be restricted from going off campus for the sake of consistency?

Are we going to have a list of things that we consider emergencies or does the person paying get to decide? It seems that we should just set the rate and let the customer decide on this one.

It seems unfair to charge for travel time. Why should it cost an extra 15 minutes just because your office happens to be in Althouse College or some other south-campus building? Isn't that partially a result of ITS being based at the very north end of the campus and not the customer's fault?