Hurricane Preparation 101 dictates that, among other things, you make sure that your gas tank is full BEFORE the storm creates supply problems. Yet it appears that a lot of folks on or near the East Coast didn't tank up before Sandy, leading to panic buying and miles-long gas lines.
Similarly, IT leaders know that massive IT storms, among them consumerization and cloud, are heading their way. Yet it appears that most of them are ignoring the warnings, continuing to do business as usual, which begs the question: Will we survive the coming IT apocalypse? Survival will be hard. Even those of us who know better keep reverting to old practices. I've been guilty of it myself.
On a small scale, just last week I found myself admitting to staff that I messed up with the software we had picked to replace our department's outdated, magnet-sliding in/out board. I had allowed our team to bypass our normal analysis process: Define the business benefit, identify key stakeholders and requirements, and so on. What a mistake.
A systems administrator went ahead and installed the clunky software on our Exchange server and client workstations. Win64 problems crashed Outlook. The software didn't allow sign-in or sign-out remotely unless, in a grotesque workaround, users sent themselves an email with text commands to sign in or out. Mobility was limited. And we were perpetuating the high-maintenance IT practices of the early 2000s, if not the '90s.
After we went through this pain, a quick business analysis revealed another path: We could have simply subscribed to inexpensive software-as-a-service that not only was hosted by a third party, but also allowed for mobile access and came with an easy-to-use, modern Web interface. Doh!
We can't continue to be our own worst enemy. Our mouths say we believe in the new ways, but we won't survive if our actions don't match our words. Most IT organizations are under intense pressure to do more with fewer people, so if we want to survive, we must take at least five steps.
1. Make the IT organization live by the same IT rules business units must live by. My main failure in the in/out automation debacle was to let staff proceed without a business analysis and justification. Sometimes such analyses show that the proposed action is a waste of time and resources. I'm not saying that everything requires a stultifying process that stifles innovation; a quick but thoughtful assessment can keep you from running off the rails.
2. Embrace software-as-a-service. For certain kinds of applications -- where there's little downside to downtime, for instance, or where the data is temporary in nature -- it makes no sense to waste staff time on installation and maintenance. In/out boards are one example. Project management is another. Help desk SaaS such as ZenDesk and FrontRange is providing rich benefits for customers, yet IT organizations continue to buy on-premises help desk software -- provision a server, patch it, occupy a switch port, add it to the portfolio of monitored systems, add virus protection, install management software, and troubleshoot it when something breaks…instead of paying a few bucks a month?
3. Take a risk on a startup vendor. It's a question of risk management. Startups, hungry and eager to grow, can provide incredible value. Measure that value against the vendor viability risks and make a decision. CIOs who take no risks reap no rewards.
4. Move beyond the PC. When you default to doing what you've always done, there are cost implications. Instead of deploying our public in/out board display on a PC, with the associated costs of a Windows license, mini-tower and monitor, we could use a Wi-Fi-only iPad as a smartboard. This options cuts costs and lets employees interact with the app without a keyboard or mouse. At scale, decisions like this one will make a big difference.
5. Welcome selective sourcing. As with SaaS, sometimes it doesn't make sense to do something internally. But I see a xenophobia among my IT leader peers when it comes to outsourcing. The arguments? "We can do it better." "We don't have enough control over a contractor." "Retraining people is a pain." They're all nonsense.
We know there are downsides with internal staffers, too, but we don't lay everyone off. So what's the right mix? In my experience, there are core competencies -- typically high-frequency, high-value activities -- that staffers are really good at. Everything else is at least a candidate for selective sourcing.
If we want to survive the coming IT apocalypse, we'd better focus less on control and more on letting go of old school IT practices and belief systems.