CIOs often need to make high-stakes decisions regarding a vendor's financial and product sustainability. Those of us who've lived through Novell, Banyan, and early short-lived ERP providers know what I mean. Vendors that go belly up or discontinue key product lines will cost you significant disruption and expense, so it's critical that IT leaders regularly evaluate their vendors' health and modify how much they rely on them.

But this kind of due diligence shouldn't just be applied to the smallest, most vulnerable vendors. It's the big suppliers that present the biggest risk to organizations, because they tend to represent the largest portion of the budget. Dominant suppliers--IBM, Microsoft, and Cisco in their heydays--play upon this "nobody ever got fired for..." quasi-dependence. They also cater to customer devotees, those loyal to Websphere, .Net, and Catalyst switch lines, for instance.

Here's the problem. The cloud and mobility and commoditized hardware are changing the face of enterprise IT. Blindly sticking with dominant suppliers while their era of dominance is winding down is like ignoring the wind change in a yacht race. Those who adjust can succeed; those who don't, can't.

Prescription? Don't be afraid to diversify during a time when most of your staff still thinks the dominant supplier will be dominant forever. That kind of change can be hard, because staffers will naturally resist. They've spend considerable time and energy getting everything working just so and have finally operationalized the innovation that brought the supplier to the table in the first place. While they've been operationalizing, however, the market keeps changing. Meantime, staffers keep hoping that whatever appears to be replacing what they've worked so hard on turns out to be just a fad.

Remember the days before more rigorous change management, when maintenance windows were more loosey goosey? Technologists would be infinitely patient with whatever upgrade they were doing, instead of backing it out when things weren't going well. "Let me just try one more time. It'll work this time." The maintenance window would stretch out, and all of a sudden you were in unplanned outage land.

Global CIOs: A Site Just For You

Visit InformationWeek's Global CIO -- our online community and information resource for CIOs operating in the global economy.

So go ahead and vigorously diversify, even though your staff won't like it. You could try out new suppliers with low-risk, low-impact projects, but your staff won't be challenged to succeed with a new supplier until you throw down a mandate to use that supplier with a critical project. Yes, one where failure isn't an option. Staffers will be freaked out enough about failing that they'll give the supplier a fair chance, as opposed to taking the easy way out on a low-impact, failure-is-an-option project.

Just because you decide to discontinue using a product that you think doesn't offer your organization the agility you need--as one of my customers in my consulting days did with Lotus Notes--doesn't mean you won't continue to use that supplier in other areas. IBM, in particular, has shown a tremendous ability to reinvent itself over its 100 years, and its recent Smarter Cities initiative is proof that it can skate to where the puck is going.

Suppliers don't rise to dominance overnight, and they don't decline overnight either. Pay attention to the winds of change, and adjust your sails accordingly.

Jonathan Feldman is a contributing editor for InformationWeek and director of IT services for a rapidly growing city in North Carolina. Write to him at jf@feldman.org or at @_jfeldman.

The Optimized Enterprise, a unique virtual event, will feature presentations and discussions on the key topics related to creating a more competitive and efficient financial services organization. It happens June 23. Register now.