Sunday, August 14, 2005

A Step Back...

Does it rain every day here in the "desert"? Based on my two weeks of observation, it sure looks that way to me. As far as my personal experience goes, seems we're going to run out of umbrellas way before we run out of water!!

Which tells you how easy it is to jump to the wrong conclusion when you've only been looking at an environment for a short time -- probably my principal occupational hazard at the moment.

Of course, on the other hand, there's the "go with your gut" school of thought, the rapid cognition crowd. Malcom Gladwell, author of the best selling "The Tipping Point", recently released a new book, "Blink: The Power of Thinking Without Thinking", which explores this decision mode -- what he calls "thin-slice cognition" -- basing a judgement or decision on a broad, but not deep, look at a situation.

I've met a lot of people at universities that didn't think much of the book. Trusting snap judgements is a notion that goes against the grain of an academic institution, founded as we are on deliberation and analysis. And Gladwell's book -- while entertaining -- in the end says little more than "sometime's snap judgements work, sometimes not" -- not exactly "American Journal of Psychology" material.

But it does say just a little more. It points out something we in "The Academy" don't like to think about very often -- i.e. that decision making is typically a harried, down and dirty process, driven by deadlines and competing interests, that doesn't always leave time to gather, rank and interpret all the data necessary for a well-reasoned analytical process. And that even when time permits, all too often the data won't stand still long enough for us to gather it. So, in the end, most decisions involve "thin-slice" cognition of one form or another, because that's how we do things in the messiness of the real world.

This is certainly true when it comes to running a University.

Which brings me to my comments on Nancy Lee's recent post on her blog about the ASU SIS system. Nancy asks a question that I think is probably in the minds of a lot of people in Computing Commons, from the "thin slice" I've taken so far.


It has been hard for me to fathom the apparent reluctance of our University administration to see the need for an institutional-level view of information management, since having that view means being able to perceive what's going on, what needs to happen, and when a crisis might be in the offing.

She goes on to suggest a couple of possible reasons to account for ASU Admin's strategy of delay:

  1. ... IT lacks glamor

  2. ... IT is little understood by lay people, and ... not all that well explained by the people who provide it.


That's not what my "thin slice" tells me. I understand that people have been warning of the impending failure of the SIS since 1993. That's more than a decade ago. It's tempting to think that ASU should have invested much more heavily in IT systems in the 90's than they decided to do. And one of the most tantalizing things about human experience is that we'll never know how doing it differently would have turned out.

But we do know how the choice they did make has turned out, and from my view they did pretty well. We serve a lot of students here, we have a growing research presence. Even if the heart of the system is held together with chewing gum and baling wire, we have pockets of excellence in the IT systems. The University survived the spectre of Y2K unscathed. As of this writing, no major IT collapse. So I think its not unreasonable to claim that people who recommended big expenditures to upgrade the IT systems in the 90's were wrong.

Not wrong that the IT systems would have been improved, because they probably would have been improved. Not wrong that the U would be running smoother now, because it might be. No, they were wrong because it turned out that despite the warnings, the systems were adequate to the mission, and the capital was available to invest elsewhere.

I don't think people didn't invest in IT because it lacks glamor. And I don't think it was a failure of understanding. I think University investment decisions, like all investment decisions, are a matter of scarce resources and competing priorities.


I assert that no one on campus can see the whole field both widely and deeply enough to KNOW FOR SURE what to do. We all see things from a unique and limited perspective. Those most familiar with IT see IT deeply -- they see its flaws, its risks, its limitations, its new possibilities. But they generally have a more limited view of the overall priorities and demands on the University as a whole.

By contrast, the university administration has a good view of the overall challenge, but a very limited view of the inner workings of any one system. It's the adminstration's job to listen to the advice of those who "dive deep", and then make tradeoffs based on resources and requirements.

So how'd they do? Well, lets look at Y2K@Asu. Again, from my "thin slice" vantage point, it looks like it worked out just about right. Several years ahead of time, those with deep knowledge began warning of potential trouble. While the U didn't respond to the first warnings, it eventually acted with time to spare, and the result was healthy, even though I'll bet the first and second times the warnings were "ignored" (i.e. projects were not funded) some people felt that the administration "just didn't get it." I think history shows they got it more right than wrong.

Maybe good decisions result from a good faith collaboration of "the wide" and "the deep". When it's working right, "the deep" provide early warnings and alternatives..."the wide" decide how to invest limited resources to do the things that are the most critical to the mission. At nearly all points in time there are more things to do than money to do them with. And that changes the threshold of what is "critical". That's where the alternatives come in. Cheap alternatives beat expensive ones almost every time. Incremental improvement trumps "run silent, run deep" or "trust us, we'll get back to you" nearly any day of the week.

At the end of the day, it's hard to argue that reworking the IT system was critical to the mission in the 90's, because ASU didn't do it, and it didn't smack the mountain. Didn't even come close.


"So what about now?", you ask. "We got Y2K by the skin of our teeth and this somehow means we never need to upgrade our IT capability?" Well, I agree with Nancy when she says that:


... help for ASU may lie in raising the profile of IT to its appropriate post-20th-century status. We're no longer pounding business calculators or keypunching cards in the basement. We are no longer an accounting adjunct to the University's real business. If the University's business is NOT information, in some form or other, I don't know what it is.


Giddy-up! Maybe IT doesn't matter anymore in manufacturing or retail (see "Does IT Matter? Information Technology and the Corrosion of Competitive Advantage" by Nicholas G. Carr), but it sure as shootin' matters to Universities. Fortunately, almost all the universities have been slow to take full advantage of everything tech can do to transform their operation, and there's plenty more innovation yet to come. So there's still time.

I agree that IT is now a core component of University success, and I'm confident that ASU and Dr. Crow are serious about developing it in the context of the New American University. After all, they hired me didn't they? :)

Success is a matter of identifying a comprehensive and implementable strategy, developing the institutional will to fund it, and organizing to accomplish it.

I'll tell you something else I agree with Nancy on, and that is that we need a map of the territory as it is and as we want it to be, and that map won't magically emerge if we just turn ourselves over to a batch of ERP consultants. We're going to have to figure it out ourselves. All of it.

Now Nancy assures me that:

...it's easy enough to confirm that our starship is covered with odd bolt-ons (which does make it easy to find in the parking lot!): just ask for a list of them... the gateways, feeder systems, duplicated data, departmental databases and applications precariously tethered to the Mother Ship.

Well I'd love to have that list. Maybe not a list exactly, maybe more like a map, or a flow diagram, or a UML model or something. I feel a little like the soldier from Stone Soup. You know, he comes to town during a famine, with a pot and a stone, says he's going to make soup. The people, reluctant at first, become intrigued with the crazy soldier and begin to bring things for the soup -- a bit of carrot, a chunk of beef, a block diagram of the linkages between the SIS, the Finance system, HR and payroll. Pretty soon we've got a feast.

So, if any of you out there want to take a crack on a whiteboard at lunch and make the beginnings of a system diagram, then take a picture of it with a digital camera and send it to me...well...I won't try to stop you... :)

5 comments:

dotguy August 16, 2005 at 8:39 AM  

New Media, Deep

I spend most of my energy diving in "the deep" spaces of new media, and therefore my "thin slices" tend to be focused on the significance of rich digital media and the impact it will have in ASUs developing technology vision throughout the transit...

Nancy August 17, 2005 at 9:12 AM  

Cool... I just hopped on to add a note and find Guy's note (click the "dotguy" signature to read his blog) about digital media. I wanted to express a wish for a "virtual whiteboard", one that we could use to collaborate on system diagrams. I haven't looked to see what's available on the Web yet, but I was thinking of something Wiki-like but graphic, something that would journal all additions with attribution. That way, if someone woke up in the middle of the night ("Eureka! I forgot to add the Data Warehouse extraction to the SIS diagram!"), they could hop onto the Web and add it. Well, maybe not in the middle of the night, but you get the picture.

Anyone know of something like that? It would HAVE to be non-proprietary and Web-based, to work. I'm sitting here not able to do several work-related things on my OS X machine at home. Switch? Never...

Nancy

Nancy August 17, 2005 at 10:15 AM  

Just Googled "graphic wiki" and didn't come up with much, tho I found two products - called "Ideagraph" (http://ideagraph.net/) and "Tinderbox" (http://www.eastgate.com/Tinderbox/, link) - that sorta do what I was talking about. But even if these aren't quite up to snuff, they have an important feature that hadn't occurred to me at first: their underlying data formats are XML. Whoopee! If you had a Web-based tool that allowed participants to modify diagrams graphically, WYSIWIG-style, and the underlying information was actually stored in XML format, you have got a LOT. With XML, you can have self-explanatory data structures that are convertible by any number of means later; you could use an XML document resulting from a system diagram to seed a data dictionary, for instance.

Nancy

Nancy August 23, 2005 at 10:50 AM  

Guy! Those products are cool, but I don't know if you can drag XML out of OmniGraffle, which was kind of my fantasy (as well as the ability to write Web-viewable diagrams with XML). It looks like Outliner uses OPML, though.

BUT! I Googled "diagram xml" and found the wierdest thing... Microsoft Officeworks Visio 2003 can export XML ("VDX") and MS will let you have the schema for it! I was stunned.... that should mean that you can make UML diagrams and workflow diagrams and all kind of good stuff in Visio, and be able to manipulate the data in a kabillion other ways. There's an useable artifact from the diagramming effort; you could seed a data dictionary or download a data dictionary (IDMS?) as XML, translate to VDX and be able to see and manipulate the components graphically. Wow. I can't believe that Microsoft is the source of such wonders. And a lot of us have desktop licenses for it.

See: link 1
and  link 2
Nancy, late lunch

Nancy August 23, 2005 at 12:43 PM  

Apologies, Guy - OmniGraffle does speak XML. I didn't see it as an export format on the vendor's Website, but some of the ins and outs are explained here

You can trade Visio XML documents back and forth with Graffle. I was really surprised to find that MS Word 2003 includes a pretty decent XML editor. My worldview quakes.

N