He’s a dog person who has a cat; saves his creative work for the mornings; and when asked what application he would manage if he had to dedicate his life’s work to one, he said “the one with the fewest users.”
He’s also a 7x CIO.
Mark Settle has a storied career leading IT at companies like Visa and Okta and has since published two books on the art of IT management and writes a monthly newsletter on the same topic.
But I learned so much more than just his pet preferences and favorite time to work when I played 20 Questions with Settle for the Systematic community earlier this year. I got a legacy player in the IT space’s unique insights on a broad range of topics related to everything from the challenges surrounding the proliferation of SaaS applications to how to gain executive buy-in.
Here are some insights I gleaned from playing 20 questions with a 7x CIO.
Data is huge.
“There’s a huge focus right now on data quality, data observability, the use of data in machine learning, and data security issues as well,” Settle said when I asked him what trends are taking hold of the industry right now. “And to top it off,” he continued, “there’s a whole slew of privacy regulations that are being developed pretty much on a piecemeal basis by different countries and even different states.”
One of the biggest challenges with hybrid work is that it breeds insularity.
The conversation around hybrid work is seemingly endless as employers experiment with how to pull off an effective balance of in-office and remote work.
“I think one of the biggest challenges to IT as an organization is insularity,” says Settle of hybrid work’s impact specifically on IT teams.
“So, unfortunately, if we really look in the mirror, IT groups can be pretty insular under any condition. They kind of talk a common tech language that no one else understands and most other business colleagues don’t really care about.” But what largely goes missing in a hybrid context is ambient learning and spontaneous sharing—that experience of sitting with someone from another team at lunch or having a quick catch up by the coffee pot.
“If somebody complains about something that’s happening in their world, they’re not going to schedule a zoom call with you as an IT person to tell you what’s upset them about the decision that was made in the marketing organization,” Settle notes. “So I do worry about IT being cast in the role of fast-food joints: All we do is get requests to do task-level tactical things, and we are kind of graded on that basis. And a lot of that exposure you have to the inner-workings of the business gets tougher. That’s one concern.”
Leverage both internal and external realities to gain executive buy-in.
Gaining executive buy-in is one of the top challenges Business Technolgoy leaders face. From getting the green light for needle-moving projects to making the case for additional headcount, securing what your team needs to reach its full potential can be difficult. As a longtime executive, Settle had some advice on how to approach these conversations. His first bit of advice? Flex everything you know about business processes that the executive may not.
According to Settle, there’s often an “icerberg of ignorance” that shields the leadership team from knowing what kinds of bottlenecks or inefficiencies are holding a process back: “I’ve seen situations where someone says, ‘Do you know that to get X thing done this quarter it took 17 approvals by like 20 people, and every hundredth time we do this, we repeat the process twice?’ and the exec was like ‘Are you kidding?’” Don’t assume leadership knows the nuts and bolts of what’s going on on every single team—make the case by illustrating the “objective reality” you’re ready to fix.
Another tip from Settle is to leverage your external networks. You can join a community like Systematic to participate in knowledge sharing with practitioners. By gathering use cases from other companies trying to solve the same problems, you can gather data to prove the potential ROI of a project.
Companies cracking MLOps are worth investing in.
In Settle’s monthly newsletter, The Modern CIO, there is a section dedicated to recent fundraising activity and his thoughts on some of the up-and-coming software taking the VC world by storm. An area to keep an eye on? MLOps.
“A lot of times, machine learning teams are very small, and what they rapidly discover is that once they have a model that’s working and being used in a business context, it requires a fair amount of care and feeding,” said Settle.
“Success breeds this problem of who is going to be taking care of these things—and I don’t just mean from a technology point of view, I mean from a business point of view. There are several tools that can help monitor the outcome of the models and another set of tools that can monitor the data that is coming in…if there are changes in the frequency of some data or the variability, just the manner in which different forms of data are being updated in the model can throw off the model.”
These unique challenges have created an opportunity for some major market disruption.
IT has to think more strategically to truly integrate an organization’s SaaS ecosystem.
The proliferation of SaaS applications—accelerated by Covid-19 and the adoption of remote and hybrid work—is in many ways a double-edged sword. While best-of-breed tools allow for more complex and bespoke solutions, the sheer volume of software now running at the average organization creates its own set of unique problems.
According to Settle, there are three well-known challenges created by this new SaaS storm: The administration of these applications; the cost (“Everyone talks about how they’re spending too much, and every survey I see shows people saying that they have licenses they’re not using.”); and security issues.
“IT’s ability to resolve those issues varies, from one domain to another,” says Settle. “A lot of times IT has primary responsibility for the security concerns but can’t always do much about the cost because those buying decisions are being made in a highly distributed kind of fashion. And even the administration may be highly distributed—which it should be. I think some of these age-old arguments about shadow IT are just a thing of the past.”
So, then, what’s the real challenge? Value realization. When you enlist a series of applications that all address niche areas of an enterprise-scale process like order-to-cash or hire-to-retire, you are then tasked with building an ERP functionality by cobbling together a series of disparate applications.
“So, IT’s responsibility,” says Settle, “is to look at the way data flows through that portfolio of apps that are addressing an enterprise-scale process and worry about data mastering, data quality, and the efficiency of the process. The real challenge of SaaS proliferation is value realization for these best-of-breed capabilities. If IT can think more strategically and use the best parts of the best tools to be best in class in the way the process functions, then you can make a real contribution.”