Nmedia - Fotolia
CIOs are under pressure to make the right technology-driven decisions that will accelerate their business in today's age of digital transformation. From automating business processes with software like RPA to save costs and boost productivity, to migrating to the cloud to increase business agility within an organization, CIOs and other IT leaders must constantly make sure they're keeping up with the competition. Safe to say, it's not an easy job.
As Red Hat CIO, Mike Kelly is no stranger to this kind of pressure. In addition to his day-to-day role as CIO, which primarily involves overseeing the company's IT department, he helped create a new data team at Red Hat that's allowing the company to get more insight from data at a faster pace. He also helped launch a data literacy program within the company to help employees better understand how best to use clean data.
He talked to SearchCIO about how these data initiatives are bringing all the data under the Red Hat CIO role, as well as his time as CIO at Oncology Therapeutics Network and McKesson Corporation and the skills he was able to transfer over to his current role.
You recently helped create a new data team at Red Hat. Can you talk a little about that and how it's now bringing all data under the Red Hat CIO?
Mike Kelly: We aspire to be more data-driven and leverage our enterprise data to help us run the business that we have in a more effective way, but also to see if we can have insights for data that helps us think about new businesses. We, like many companies, didn't build our existing system landscape around capturing the data that we get. It wasn't really designed that way, way back when.
We had a lot of the data quality and data literacy issues and it has been a long-standing problem -- we would have situations where people would come to meetings and say, 'Where do you use that information?' So, we decided that we were going to form an enterprise data analytics function. As part of that, the original remit of it was to put some governance around the data quality and put standards around data definitions and form some committees of experts and things of that nature -- build a common shared data analytics platform that the company could use to bring some sort of order to what we had put in place.
It was decided that the role would report into me [and], as we went on the journey together, if we felt that it would be more effective in another function of the company, that would be fine.
If you study the data analytics functions, in some cases [it] will tell you that it reports directly to the CEO, in some cases it reports to the CIO and in some cases reports to the CFO [or the] CMO. I think there's a real reckoning in companies today around the need that data and analytics is a real corporate function. And it's not something that people do off the side of their desk -- data science typically is a real profession and something that people can have a career progression against. So, we set that up. I wouldn't say that not all data and analytics live in the IT organization -- we have it set up where we enable analysis out in the lines of business and we encourage it.
You also helped launch a data literacy program at Red Hat. How did you go about implementing the program?
Kelly: We built a pretty robust data literacy program [where] we have a data rights and access [function] that we have to put in place for people so that sensitive information is not widespread. But we give a general data literacy program to many of our Red Hat associates and that program is doing really well.
What didn't Red Hat employees know about data literacy? If there's holes and gaps, was it mainly about security or a single source of truth?
Kelly: I think the knowledge was really tribal -- it just wasn't common. We're still in the midst of the program [as] it's a multi-year kind of thing. But I think people that were close to a particular domain, they maybe knew the data inside out [or] maybe people who weren't as close to that domain but needed to have that type of data to perform more holistic analysis were able to come up to speed and everybody can speak the same language.
It was really about getting a common vocabulary and taxonomy around the data itself. For example, you might have somebody in the products group who knows the product data inside and out, but maybe doesn't know the customer data inside and out, and yet they need those two things to perform a particular piece of analysis they want to do. So, we give them the ability to learn the pieces that they don't know. It is a level playing field.
For a while, it seemed like most companies were appointing a chief data officer. Did Red Hat go that route? Did it ever have one?
Kelly: We did not. We have the next best thing, which is a head of data and analytics that reports into IT. I think a lot of companies make the mistake of standing up a function like that -- that they don't take inventory of where data and analytics is happening in the rest of the company. And then you have these competing organizations and shadow organizations, and we had a little bit of that at the beginning. But we rightsized it -- we put together straightforward roles and responsibilities for people. The whole goal here is around enablement and commonality and trust in the data.
You were previously CIO of Oncology Therapeutics Network and then CIO at McKesson Corporation, as well as a few other senior leadership roles there -- what skills were you able to transfer over to your role as Red Hat CIO? What remains consistent amongst CIOs in industries across the board and what's different being a CIO at a software company?
Kelly: The companies that I had worked in prior were, in a lot of ways, rooted in operational excellence and dabbling in more of this innovative risk-taking type [strategy]. You have to have a good mix of both to do this job well, in my opinion. I think the skills that I was able to bring over were operational and execution-based [and] that I knew how to do well. And then also how to make sure that we create a culture around that. Bringing that mindset to people was something that I brought over and just how to work in a big company and how to scale ideas really quickly.
What insight do you have for CIOs and other IT leaders currently leading digital transformation initiatives within their organizations?
Kelly: I say they're really tricky. Talk to your peers. Everyone's sort of going through the same thing here and I think that there's a common misconception that there's a secret recipe. I think it's very situational to your particular company. And in my dealings with people, the first questions I ask are, are 'How tolerant for change is your organization?' and 'How do you pressure test the reply to that?' Because all too often we wave these fancy words around like 'digital transformation' and people don't really understand what you mean. And when the rubber hits the road and you actually have to change, that's where these things fall over.
Different companies have different audiences for change.
CIOs are under great pressure to make the right technology-driven decisions. Have you ever faced that kind of pressure of having to make sure that you're making the right decisions?
Kelly: Yes, I think there are a lot of technology decisions that are available. I'd be less concerned with the technical decision itself, and more of the decision about how quickly and how widespread the impact of the change is going to be to the company. That's where I get stressed because I think as CIOs, you can see the path clearly. But I put a couple articles out there about this concept I call 'leading from behind and finding your change agents to help you advocate and deploy change.'
To me, the technology is the technology, but if you choose technical solution A or technical solution B, and in neither case, there's a willingness of the organization to truly change, then both are going to fail. I don't think it comes down to picking the right technology or not. I think [it's] if you pick the right pace of change that the company is capable of sustaining.
Editor's note: This interview has been edited for length and clarity.