cio 1-1 Mark Settle has had his share of loving and loathing technology.
The chief information officer of BMC, who has over 20 years of experience as IT chief, has been around long enough to see the value of some technologies and tech visions depreciate over time. ERP systems, the veteran noted, are one such example of being extremely beneficial when first introduced, but which today do not seem to be as flexible in responding to the challenges of the business.
That said, technology and processes are not the only pieces of the puzzle--getting people to work collaboratively in an IT organization is also key, and a big challenge.
In an interview with ZDNet Asia earlier this week, Settle spoke about his CIO career and what BMC is doing to tap newer technologies such as cloud for software testing.
You've been CIO of BMC since June 2008. How has the organization's IT changed in the last two years?
BMC's value proposition revolves around what we call service management, or business service management. There is a series of best practices that kind of form service management, and we've people in those process roles, such as enterprise incident manager or enterprise configuration manager. When I first came to BMC, people in those roles were from all over the organizational chart. I put them together in a service management office, and I had the office report directly to me because I wanted to send a signal to the whole organization that these processes were not just processes--they were able to run datacenter operations or network operations or end-user computing.
Another change that we made was we formalized a job we call automation engineer. We introduced this to take a tremendous amount of day-to-day legwork out of the IT operations--the automation engineers actually study how people spend their time during the day and specifically try to identify highly repetitive tasks. Examples of these include labeling servers, resetting root access or passwords, and reimaging PCs. We find ways to automate these tasks that are repeated over and over. The answer we look for is to make low-value work go away altogether. If you want manual labor in IT, frankly, it's to correct or fix problems that occur.
Another area was change management. IT organizations always have a high sense of concern around putting change into a production environment. We want to minimize the number of changes that we put in on an emergency, expedited basis because our experiences have shown that the changes are not always well-thought through--you inadvertently create problems that never existed before.
And so, internally, we created a new term called templatized change. It represents a procedure that has been successfully executed in the past--it's like one of the lowest-risk things we can do because we've done it in a production environment before. Depending on the number of changes that are required, when we minimize the emergency ones and increase the templatized ones, the overall risk profile starts to go down and you just save a lot more time.
In terms of tapping technologies, what were some of the key highlights in your BMC journey?
We've become very extensive adopters of SaaS (software-as-a-service) applications.
If you think about it, BMC is a good candidate for SaaS. A lot of people develop anxiety about putting consumer-related data into an SaaS tool or in the hands of a third-party SaaS vendor or application…people do not want credit card information or personal data to go over to an SaaS environment. We're a B2B business--we don't really have that kind of exposure to retail-type issues. And also…what you read about SaaS is true. The time to market is short, initial upfront costs are less.
Right now, we use about 120 applications to run BMC--these would be systems at the enterprise or business level--and a quarter of those are SaaS tools.
Stepping back a little, how have your previous experiences helped you in your current role?
I've seen lots of change over the course of my career. I've seen different waves of technology come through from things like ERP systems, to the wave of innovation that occurred toward the end of the '90s to combat the Y2K problem. I've survived the database wars in the old days between Informix and Sybase before Oracle kind of phased that out.
Probably early in my career there were periods when I fell in love with specific technologies…but what I learned over time is that there's no one magic bullet in IT. There's no one set of technologies that is the solution to all problems in the business or all the efficiency issues within IT.
You hear stuff from "The mainframe is dead, we're not going to make it anymore" all the way to "Now all business will be conducted on Twitter and Facebook"--none of that is true. We need to find the right combination of technologies to support BMC, and what's right for us will be different from, say, a bank in China. We all have to figure out the right combination of technologies--there's the old saying: "If the only tool you have in your toolbox is a hammer, then every problem is a nail."
Can you share some advice on how to derive that right combination of technology?
That's a good question. You've to find some way to keep bringing new knowledge into your organization. You can do that by employing people who know a lot about new technologies. You can do that by partnering vendors to experiment, in some cases pilot small-scale projects where you can take a risk, and if it doesn't work out there will be no financial-operation consequences. Sometimes in larger shops, the experimenting exists through incubator-like environments.
The most effective way is hiring. But these days a lot of IT organizations are not growing headcount, so it's hard to achieve that.
What are your priorities for 2010 and how does your budget this year compare with 2009's?
Like a lot of other companies, we are trying to figure out what the cloud means to a company like BMC. We've already talked about the first step, which is tapping SaaS applications to simplify and add more value to the business.
We've also made big strides in using virtualization to improve the efficiency of our internal environment. We have development teams in North America, Israel and Pune that are currently using a distributed set of IT labs. We've about 6,500 servers and we've pretty much virtualized most of our servers at this point. So that's pretty much a success story. We've retired a whole lot of hardware, reduced datacenter floor space requirements and we've significantly reduced our power consumption. There's a lot of focus on continuing that journey and driving up the utilization of servers and storage.
The third leg is to start to use public cloud providers for some of our software testing activities. A lot of our products are used by large companies. You may have heard of a product called BladeLogic--before we purchased the company, its largest single customer was the Bank of America, which deployed BladeLogic on about 35,000 servers distributed around the world in [its] data centers.
Now, we upgrade and release new versions of BladeLogic all the time and ideally we'd like to be able to test each release on several thousand servers. It'd be very difficult for us to actually go and find several thousand servers within our existing in-house resources. So we'd like to be able to burst out to an Amazon.com-like cloud or some other third-party provider and actually conduct our scalability testing up there in a public cloud for a period of, maybe, couple of weeks. We've conducted pilots over the last 90 days when we demonstrated the ability to do that to ourselves, and we want to turn that into a production capability for our development teams next year.
In terms of budget, we're spending a little bit more this year than we did last year. Not a huge increase.