From Context to Core

• 12/01/08

At Campus Technology 2008, Arizona State University Technology Officer Adrian Sannier mesmerized audiences with his mandate to become more efficient by doing only the 'core' tech stuff-- and getting someone else to slog through the context.

The following is excerpted from Sannier's hour-long keynote address at Campus Technology '08.

IN THE PAST 12 MONTHS, we've moved from optimism to pessimism as oil prices soar, as the economy gets tighter, the credit crisis widens, and foreclosures continue to rise. It's like 1930 all of a sudden and we don't have any faith in politics. We have faith in one thing in this country: technology. Everybody knows we're going to solve climate change and they know it's going to be technology that's going to do it. Everybody knows we're going to solve the energy problem and they know it's going to be technology that's going to do it. And we also know that if we're going to extend education to a broader sector of people, if we're going to increase excellence and make it have an impact, technology will have to be part of that solution.

Technology is growing like wildfire. Yet for all the wikis and the blogs and the portfolios and the networks and the wireless and the yik and the yak, the learning process-- at least here in higher education-- still revolves around going to class, listening to a professor, and telling him what he said, on a test.

So we've got to get a little more serious about this whole thing. Frank Rhodes, president emeritus of Cornell University [NY], wrote on this subject, and I'm going to read you just a little bit of something he said:

The business of learning remains largely untouched by information technology. Most instruction is still a cottage industry, little influenced, as yet, by the benefits and support of modern technology. It's as though an industry had computerized its business and management activities but left its manufacturing operations and sales distribution essentially unchanged and unimproved. Universities have not yet applied new technologies even to reduce costs. In their basic business of teaching resident students, they have not diverged much from the methods of Socrates, except faculty members have come indoors.

In Forbes magazine, John Chambers, president of Cisco, wrote that he is very concerned about the state of education in the United States, and I think he has some really interesting things to say that are different than the things you hear in the academy. "Many agree that technology should play a greater role in our education system," he wrote.

I believe, though, that many don't fully understand the impact or extent of the role that technology has to play in our future. I also believe that the real question is whether we will be prepared when it becomes clear. We call this "catching a market transition"-- something you need to anticipate three, four, or even seven years in advance so you can position yourself properly to compete.

According to Chambers, what happens to a company that fails to anticipate a market transition? Good-bye. That's why, for academic institutions, anticipating these changes is so critical. At Arizona State, we believe a number of things must happen in order to succeed, and this is what I'm charging my team to do because you can only do so much.

Most of these things revolve around enabling transformation-- the most difficult thing for any institution to try to do, and the one that's the most critical because it's the one that liberates resources. Still, the approach is very controversial. What I'm getting at here is: Universities are the cottage industry of IT. Even big ones.

ASU is a big one. We have 65,000 students. That's 65,000 e-mail accounts. What did it cost us? It was costing us $400,000 a year to provision those for students. How many megabytes do we give them? Megabytes? Who measures e-mail space in megabytes? Universities still do! The question is: How much longer are we going to compete like this when other people are competing in gigabytes?

Which brings me to the theme here: We need to convert context spending to core spending. You have to change from taking 80 percent of your resources and spending them on context, and switch it so that 80 percent is spent on core. Think about what that actually means in practice. Look at your budget, see where it's all going and then ask yourself, "How am I going to switch it?"

Why is it important to do this? Because you pay attention to what you spend money on. Unless you figure out how to change, or leapfrog your competition by doing something fundamental, the time for just deploying the tools and getting the free advantage is gone.

This strategy revolves around the "concept of one": Go into your data center, and collapse everything. How many databases do you have? How many people make choices? How many different standards do you have? How many applications do you have for something? How many ERPs do you have? Collapse them! Then think about the "concept of zero." The concept of one is: "Do it once, do it well, and do it everywhere." But the concept of zero is: "Don't do it; get someone else to do it-- someone bigger than you, richer than you, and far more powerful than you." This breaks the heart of every technologist, but it makes sense.

Some examples: I'm used to shelling out a lot of e-mail addresses-- 65,000. How many do you suppose Google has? Well, they have a "googol." By going with Google for e-mail, I saved $400,000 a year. And that's irrelevant; it's not the savings, it's the curve. It's like they landed on our campus and they said, "We can take that thing you have legions of people doing, and we can do that." In general, you get on that curve, and suddenly, you're not doing anything-- they do it all.

We've got to start riding these curves and we've got to quit arguing about it. We've got to get moving here and it means we have to sleep with the devil. We have to sleep with commerce and get in bed with those guys. We absolutely have to. The most common thing you'll hear when you hear this strategy advanced, is this notion of, "Why are we doing this?"

You just have to get over it.

Compare some idealized universe to the commercial universe: Compare what you're really doing, to what they're really doing. Compare your business continuity strategies to their business continuity strategies. And compare your disaster recovery strategies with their disaster recovery strategies. Make all of these comparisons. No matter how well you're doing these things, economy of scale forces them to do theirs better.

Where do I get this strategy? I read what people are doing and I try to figure out how we are going to apply it. I have an RFP on the street right now to get out of the network business. That's right: I'm going to stop being an internet service provider. Why should I be an ISP? I have vendors doing that job already, and every one of them does stuff at a scale that I can't even approach.

Another example: We did our ERP in 18 months. That's unprecedented in higher education for a school our size. How'd we do it so quickly? We didn't try to do it at all. I didn't try to take my team, train them, figure out how to deploy machines, then configure them, then build all the different environments. I didn't do any of it. What I'm telling you is that if you're out there without partners, you're doing it wrong. Context to core. On your campuses, it's a revolution.

Now I'm going to tell you some of the things I think are core. The first one is administrative services. We have all these systems producing all this data and we're doing all this training. We're training all these people on PeopleSoft. We train them on this system, and we train them on that system, and every one of these people has to become an expert in all of these systems so that they, in a distributed way, can run all these reports to justify whatever position they need to justify. Admit it! If that's not happening at your campus, it's certainly happening at mine. We have fractured business intelligence, but we need to change it. At ASU, you can go to a new dashboard and look up any department, college, or center and ask a series of questions at any time. What's your research path? How many proposals do you have? How many awards? How many expenditures do you have? What are the trend lines? What we're doing is managing academia like a business. The bottom line: You've got to get to intelligence. You've got to run the same report every week. There's money to be had there. There's intelligence to be had there, too, because our people are capable of real analysis and much, much more.

Another core approach is to anticipate the way students are going to need to compute when they graduate and move on. For too long, we've said, "Students won't bring their laptops to campus," and "They won't do this," or "They won't do that." Hey, that ship has sailed and the cattle car has got to go! We've got to transform the way we approach this overall. We have to build students an environment that's going to be like the places they're going to go, and find economic value. That means scary stuff for any IT organization. It means you're going to have to support them: 65,000 kids packing virusladen laptops that are critical to their education. Some schools will say, "We'll standardize the laptops." Others will go straight to the iTouch or some other portable device. Whatever you do, this is what you should be spending your time and your energy on-- not on building the next e-mail system.

It's also important to make applications that we deploy available to everybody, everywhere, all the time. This means web delivery. We've got to put pressure on Adobe and Microsoft. We all should get Adobe Photoshop from the same hosting provider because I don't have any special expertise in building a giant server farm to serve up the same applications that you have. We all should be doing that.

We've got to go from cop to concierge. What are our students used to? Amazon. com? Well, I have a word: On his Comedy Central television show, Steven Colbert got famous with the word "truthiness." My word is: "Amazon.comification." That's what you must do. You must "Amazon.comify" your web presence. When somebody says to you, "That's on my website," you should say to them, "There is no my website. There is the website," and the website is the web page where the services are.

In order to accomplish "Amazon.comification," you need a heavy dose of prune concentrate. Prune concentrate is the notion of taking your website and pruning it because you've got too much there. At ASU, we've got more than 1 million web pages but, if you look at the distribution of hits, 50 pages get all the traffic. Yet, all of the best stuff needs to be concentrated in the core site. This is not a technical problem; it is a political problem. It took two years for us to build an Amazon-like site that we deployed in the fall; two years to get people to be willing to use the same thing at the top and the same thing at the bottom. You need to do this too. You have to start asking yourselves, "Where does this go? Where do we put this in the presence so it gets traffic?"

Another thought: You don't belong in the help desk business. I know we all invented help desks, and lots of us have help desks, but not many of us know how to run them. It's a massively data-driven business where you analyze every day. If you ask me, what you're doing is taking some of your most experienced and oldest employees and having them answer the phone from 8 am to 5 pm. Well here's a news flash: The next generation stays up all night long and when they want an answer, they want an answer immediately. Help desks must be a combination of having the wherewithal to make that happen and getting the political buy-in to outsource the answering of certain kinds of questions at 2 am, because there's no other way to achieve scale to do it.

Finally, I suggest you burn down the library. All the books in the world are already digitized! Burn the thing down. Change it into a gathering place; a digital commons. Stop air conditioning the books! None of us has the Alexandria Library; Michigan, Oxford [UK], and Stanford [CA] have digitized their collections. What do you have that they don't? Why are you buying new books? Buy digital and let's spend some more time making those things level, flat, and transparent so a single search turns up everything we have. This has to change, because it's clear that people want to find information digitally. They want to search for it, find it, have it, and then amalgamate search results into a précis.

The bottom line is that we're tool rich. We have all the tools in the world and we've got more tools coming every day. If we embrace core versus context, we can unleash these tools without even getting in the flow. Our faculty members believe that the only technology that they need is a death ray from their eyes to hit all the people with laptops in their classrooms. And this is a tough problem because I realize that what I've been sharing with you today is my intuition as a technologist who has an interest in higher education. I know that intuition doesn't really cut it in our enterprises. But I can hope...