HBO’s emmy-nominated comedy series Silicon Valley offers a humorous behind-the-scenes take on the trials and tribulations of technology startups. Its techie-driven dialogue is unusual for mainstream entertainment, but refreshing for those of us in IT. Pulling back the curtains on life at Silicon Valley IT departments is a specialty of our guest on this episode, Mark Settle. As a 7-time CIO of companies such as Oxy, BMC Software, & Okta, Mark reveals quite a bit in his latest book “Truth from the Valley, A Practical Primer on IT Management for the Next Decade”.
Drawing upon his many experiences leading IT shops, Mark succinctly articulates both the problems with current approaches to IT, and his prescription for how people, processes, and technology must adapt to a digitally transformed future. Along the way we’ll learn what he thinks are the top three characteristics a CIO should look for when hiring an IT professional (other than intelligence), the key to overcoming robophobia among staff, and why a CIO’s success depends on them picking a handful of things to be Best-in-Class at.
Guy Nadivi: Welcome everyone. My name is Guy Nadivi, and I’m the host of Intelligent Automation Radio. Our guest on today’s episode is Mark Settle, author and former CIO of Oxy, BMC Software, and most recently Okta, the cloud based identity and access management software provider. Mark’s most recent book is titled Truth from the Valley, a Practical Primer on IT Management for the Next Decade. Now I’m not a CIO, nor have I ever been one, nor do I aspire to become one, nonetheless, I found this book riveting and a very enjoyable read because it articulates great insight for IT professionals, whether they’re working inside or outside Silicon Valley. And even within the very first few pages, you realize it really is the blunt unvarnished truth. And Mark, who is a seven-time former CIO pulls no punches in his acute observations and advice. And since many of Mark’s observations touch upon the topics this podcast focuses on, we’ve asked him to come on the show today so we can drill down a bit and tap into some deeper truths about the Valley. Mark, welcome to Intelligent Automation Radio.
Mark Settle: Thank you Guy. Thank you for that introduction. Any author loves to hear their books talked about in glowing terms. It’s sort of like your baby. So everybody likes to be congratulated on their children. So I wish I had more introductions of that nature.
Guy Nadivi: It was my blunt unvarnished truth. It was genuinely an enjoyable read. So you have a section in your book that talks about the five levels of CIO consciousness. And I won’t go through any of them because senior IT executives should really read that section for themselves. However, in order to get to level five, you cite that one of your keys to success and a principle you live by is based on something that Steve Jobs once said, “It doesn’t make sense to hire smart people and then tell them what to do. We hire smart people so they can tell us what to do.” Now it’s clearly understandable why a CIO would want smart people on his or her staff. But aside from intelligence, when you were a CIO, what were the top three characteristics you looked for when hiring an IT professional?
Mark Settle: So I think in a recruiting process, you have to look at the role. Some roles are very technical, IT centric, IT focused. Others are going to involve extensive interaction with your business partners. So the role can be somewhat different. And I’m also assuming in the question, that these people have the requisite skills, or else we wouldn’t be looking at them at all. But I think as a generalization, regardless of role, there are three things that I’ve always kind of focused on. So one is the breadth and variety of experience. And I’m always interested in people that have worked for different companies, or have assumed different roles in a single company. I think a lot of people are very successful in IT that come out of consulting backgrounds early in their career. And I would attribute that in many cases to the fact that they’ve had opportunities to work on different kinds of problems in different business models, and have to work with different kinds of people. So I think that type of experience can be pretty important. The ability to work with others, and of course, this is kind of a happy thing. Everybody would say this, but there are very few roles in any organization, IT or otherwise, where you can operate as a kind of a lone wolf. And again, whether it’s internal to IT and you have to win the internal hearts and minds around doing things differently, or new initiatives, or whether you need to be influential outside of the IT group itself, a record of successfully working with other people, and teams, and projects and maybe distributed remote team coordination, those are important capabilities. And then the third is a record of progressive accomplishment. And I think a lot of people, and it doesn’t, I’m not talking about becoming an astronaut or anything, but you can see through over the course of somebody’s prior experience whether they been progressively challenged with broader responsibilities, or more difficult assignments, and the presumption is they’ve learned some things along the way. They may not have succeeded at every project. And in some ways you can learn a lot more by failing than you can with succeeding. But I think people that have a track record of progressive accomplishment or stand out are exactly the kind of person that you want, because never lose sight of the fact that you’re really hiring somebody to make the whole group better. And if they can’t live up to that standard, then you kind of have to scratch your head about like, “Why are we doing this at all?”
Guy Nadivi: A number of years ago, I wrote an article for Forbes titled “Why you shouldn’t hire a perfect job candidate.” And I pointed out that nobody is perfect, that’s why pencils have erasers. You argue essentially the same thing, and use a great term to characterize what some hiring managers do, which is go on “unicorn hunts”. Very interestingly though, you also quantify what an IT organization should cap its tolerance point at for talent gaps, and that’s 20%. And you go on to write that, “Candidates with gaps in excess of 20% should never be hired.” Now I’m very curious Mark, in a hyper-competitive market for IT talent acquisition like Silicon Valley, why not instead revise the job descriptions for the position being filled so that a particular candidate’s talent gap drops below your 20% threshold, then hire and develop them further internally? In the meantime, perhaps you could plug their talent shortcomings with an outside contractor until they get up to speed. In the long run, wouldn’t that help grow the IT organization’s brand so to speak as a place where talented people can grow, thus making it a more attractive workplace for other talent?
Mark Settle: That’s certainly in certain situations that might be a desirable thing to do, particularly in a difficult to obtain skillset. Let’s say you, your team has no dev ops experience whatsoever, can barely spell the term. You would most likely want to bring in a contractor that was real knowledgeable in terms of those kinds of processes, and go find somebody either internally or externally that could step up to approximate leadership in that area. But in my comment about the 20% gap that needs to be closed, that isn’t theoretical. I mean, that’s a very practical and pragmatic assessment on my part because I’ve seen many cases where managers have that good intention that you just described. “I’ll operate outside that 20% gap. I’ll find somebody that satisfies about two thirds of my requirements or three quarters of my requirements. And with a little care and feeding, I’m perfectly confident in my own management capabilities to close the gap.” Now that would be true in many cases, if that was sort of the only thing that the manager had to worry about. It’s quite likely that he already has other members on the team who want developmental coaching as well, and so they’re going to place demands on his or her time. And again, I’ve seen time and time again where the gap just didn’t close. Take a major application platform like Workday or NetSuite, people that would come in because those skills are in such high demand with some degree of experience on those platforms, but if they’re asked to do something that far exceeds what they’ve done in the past, in many cases, they come up short, even though they have this kind of initial, maybe apprentice level understanding of how the platform operates. So, obviously the remarks weren’t made to denigrate the importance of development. It’s just really to denigrate maybe the over-optimism of managers who kind of wade into those situations and feel that they can make a big difference or actually close the gap in the end.
Guy Nadivi: Fair enough. On the employee development front, you have a great section titled “Secrets of Developing Really Talented People”. And in it you write about how talented people have a “tendency to value ideas over practical business results.” And then further down, you continue “Talented individuals who suffer from implementation deficit disorder need to be counseled that ideas alone, regardless of their potential significance or theoretical value are not a source of competitive business advantage. A relentless stream of great ideas that never get implemented is actually a sign of career failure, not career success.” Okay, that’s a good point. But what if the talented individual is reporting to a manager who lacks the vision to recognize a talented individual’s ideas should be implemented because of the competitive business advantage it could create. And this could be especially true in an era when talented automation, or AI, or machine learning professionals are proposing things that have never been done before to a manager who lacks the technical insight to understand what’s being proposed.
Mark Settle: Great question. My experience is every organization is drowning in good ideas, right? So there are lots of ways in which information technology can contribute to improving revenue or improving profitability within a commercial organization. And I think in many cases, when staff members come up with ideas, which are very technical ideas, sometimes the manager has to play the role of realist. So that doesn’t directly address your question, but I want to come back to it in a minute. In many cases, the manager understands that there are budget limitations, that they may or may not be able to go find a business partner that’s going to join forces with them so they can lead the crusade on this kind of idea. Well, there may be an understanding that there are other priorities right now. The bandwidth of the organization is tied down because there’s already a commitment to finish certain projects on time. You’re not going to go back to existing business partners and change any of those commitments. And then there may be just kind of overall organizational risk aversion, maybe you’re not working in a company that’s known for being early adopters of some new kinds of things. And so, I think it’s unfair for the staff member who has this epiphany to sort of unload all of those practical issues onto the manager and say, “You go solve all that. You develop a political consensus. You go find the budget dollars. You go talk to somebody about why their project end date’s going to get delayed” or whatever. “If you can sort all that out, then we can join forces.” So I think in many cases, that’s the more predominant problem that you run into. Now to go directly back to your specific question, I mean, if there are people in the management role that grew up in a different set of technologies and just can’t appreciate potential importance of some of these new things that are emerging, you have to look around the organization as a whole and see, “Well, is that just sort of like an anomaly, this is kind of a one off case with my manager, or in fact is this the whole company sort of this way, a pretty comfortable mainstream adopter, not wanting to do things too soon, or too early, or assume too much risk?” And then I think if you’re the kind of person that really loves to be at the cutting edge, with all due respect, it’s kind of time for you to look around and maybe look for some alternative form of employment, because you’re just kind of fighting a cultural problem, not just in the form of a single manager. And if it’s a single manager, and then I think the only other thing that over time, if you tried several ideas and they didn’t work out, you can probably jockey around and try to find another team or group to join. So you represent a very specific case that has to be dealt with, but I think in the more general case of why don’t more good ideas get adopted, there are many practical considerations that the technical people just don’t even weigh in.
Guy Nadivi: Could be the basis for your next book.
Mark Settle:Could be.
Guy Nadivi: And in the process section of this book, you write that “Data management is no longer an unavoidable hygienic activity performed within the bowels of the IT organization. It has a critical impact on employee productivity, accurate financial reporting, business agility, and customer satisfaction. Consequently, it should be at the forefront of the process agenda of every IT leadership team. It requires their immediate and sustained attention.” Mark, how will an IT leadership team’s failure to heed your advice, impact things like machine learning, which relies on clean data to achieve many of the advanced capabilities critical for digital transformation?
Mark Settle: Well, obviously no, and I think that’s an easy question to answer because it has a tremendous impact on the ability of those tools to be effective. And as Guy you and I have talked about in the past, many times people develop data science skills, they want to go off and develop predictive models, and what they end up spending 80% of their time doing is data wrangling, trying to clean data, and normalize it, and get it into some kind of usable form. IT tools and IT teams can go long ways towards making data more consistent and normalized, et cetera. But at the end of the day, you really can’t fight bad data collection habits, or a lack of data collection hygiene. Good example is on the Salesforce platform where you have a large extended sales team. It’s asked to supply all different kinds of information through dropdown menus and many cases with very little training, very little enforcement of definitions or standards, and the way data is supposed to be collected. And so when you start to query the database for simple things like why did we lose these kind of deals and these kind of industries during the last 18 months, the quality of data that’s in that database to actually come back with the answers can’t support the question. So all the hard work that IT can do to try to take all these different points of collection and try to harmonize things is a benefit obviously, and critical to the success of ML & AI capabilities. But bad data in, no results out at the end of the day. It becomes insurmountable after a certain point in time.
Guy Nadivi: There’s a fascinating segment of your book titled “Overcoming Automation Indifference”. In it, you recount an intriguing experience about building an automation center of excellence at one of your prior employers that focused initially on using automation to improve the service desk and data center operations. And you go onto write, “The results of these initial automation projects as measured in time savings were astounding. They far exceeded our intuitive expectations. We actually discounted the results of our time savings analyses because we didn’t believe that executives outside IT would believe our results.” I’d never considered skepticism or outright disbelief about automation’s deliverables to be a possible hindrance to its adoption, but I’ll ask anyways. Do you think that incredulity about automation in this era of fake news is somehow curtailing its deployment and adoption?
Mark Settle: No, no, not at all. I think what I talk about it in the book, it’s always beneficial to try to quantify the impact of new tools, new automation tools in this case specifically. But in a lot of cases, these automation capabilities lead to very little slivers of savings, maybe multiple times per day, or over a large number of employees. And when you start to do simple multiplication and addition, it would appear that there’s just dozens of man years of effort that are going to be saved at the end of the day, after we’ve deployed all these robots, or scripts, or whatever the specific tool actually is at the end of the day. And I think this is a case where actually the business execs are very receptive to arguments for more automation. I don’t think this is a hard sell. I think what I’m trying to apply here sometimes maybe some of these calculations actually work against you. You might look as if you drank the Kool-Aid somehow, like you believe that the whole workforce could be eliminated through the use of robots or something in the future. So, so, no I don’t think there’s incredulity there. What I do encourage a lot of people though that try to implement these automation capabilities is to try to be careful in this clarification process or business case process. And I would encourage people to look for proof points. And what I mean by that is to go and look in the back office operation. And if you can say, “Well, here’s a case where staff that’s half our size is processing three times the number of transactions. Wouldn’t you like to be like that company” or “We purposely grew the company, added new accounts, acquired another company, and we were able to step up to the challenge of doubling our workflow through some kind of a process, and we did it finally having increased our head count by 10%” or whatever … So I think proof point arguments can be very persuasive if you really get into very detailed, finely focused set of calculations about getting back five minutes every other day from persons, a person in this role in saving three minutes per day every day from a person in this role. And you start to roll those things up. I think it’s almost self-defeating in terms of signing up business executives to invest in these kinds of tools.
Guy Nadivi: Mark, in the technology section of your book, you talk about how proliferation of SaaS-based resources will change the role of IT. And you state that they’ll “need to operate in more of an oversight and advisory capacity since they will no longer have direct command and control responsibilities for many of the computing resources supporting daily business operations. They will need to ensure security policies are being enforced, data governance practices are being observed, and cloud computing expenses are being prudently managed across the enterprise.” You also argued that “IT leaders need to stop managing construction costs and start managing consumption costs. Many have not developed the organizational competencies to fully exploit the pricing models employed by cloud vendors.” Mark, is the ideal IT professional of the future going to be a hybrid blending of a cost accountant, financial analyst, and perhaps a data architect?
Mark Settle: If you add industrial psychologist to that list, you might just have the perfect IT employee of the future. So I think you’re onto something with the question. Again, we’re all coming out of this command and control world in which IT was the middleman between new forms of information technology and the needs of the business. And one of the things I’ve talked about a lot recently, and truthfully don’t remember the extent to which this theme is developed in the book, is that of stewardship. So with all the functional groups going off and buying their SaaS tools, and different data services, and sometimes without the knowledge or forewarning for IT, you end up in a situation where it would be very easy to step back and say, “I don’t have any responsibility for what they did. They went off and did those things, never told me” or “I was consulted at the 11th hour and treated like a procurement agent.” “They cut a deal, they got their pricing, they scheduled the professional service implementation team, then they came to me and said, ‘This is what we want. Go get us a contract with this particular vendor.'” So I’m saying that’s the easy, and maybe spiteful thing to do is to say that “Now that’s your problem. I’m really not part of this,” but that could be career limiting, and I think it’s very unprofessional. So even though you don’t have the budget to be paying for these tools, you may not control some of the customizations that are going to be made, et cetera, I think you still have stewardship responsibilities for the entire portfolio with respect to things like total spend across the organization, security issues that may exist, ensuring that integrations are being made so that the company is getting the full benefit of these dozens or even hundreds of SaaS applications that have been brought in. And so I guess, do we just all need to probably to learn how to win friends and influence enemies over what in the future, maybe that’s the perfect person because you’re not going to be able to come in and say, “I’m the General and Chief here. My budget contains the money for these things. And therefore I have the ultimate control.” Those days are fading fast.
Guy Nadivi: You’re a big advocate for automation. And towards the end of your book, you discuss the many benefits it provides with a particular focus on elimination of rework, often resulting from human error, and you implore IT leaders to adopt an automation first approach to daily operational processes. But you also touch upon the potential resistance to automation that might be encountered from employees. Now, here on the podcast, we refer to that resistance as robophobia, and it can prove to be a potent cultural obstacle for enterprises deploying automation. Mark, I’m sure a lot of the IT leaders in our audience would love to know what tactics have been most successful for you in overcoming robophobia?
Mark Settle: So, almost every time you bring an automation tool in, you tell the employees that we are going to free you up from busy work so that you can work on higher value activities. And if you want to win their hearts to the automation cause, you really have to go find that high value work that you want them to do that in effect replaces the time and energy that they’ve put into the busy work in the past. It shouldn’t be left to them to go figure out what to do with whatever kind of time savings or efficiency improvements that are realized through the use of these technologies. And I think all too often, managers, directors, and leaders just think “We’ll get around to that later. Let’s see how this thing’s really going to work. Let’s see how great the time savings actually is. We’ll sort it out kind of after the fact.” I think as long as the folks who are having their work lives surgically altered are told very bluntly and directly about what new opportunities are, the reskilling that’s going to take place, the projects that they’re going to get assigned to that they would have never had time or just step up to in the past, I think you’d very rapidly find folks embracing the tools. The other thing I counsel people was to start small. So if you had, and what I mean by that is go find some very limited processes, and just make everybody’s workday just a little more efficient. And I think when people see their own personal productivity improve with some really kind of small personal wins, I mean, that’s another way I think the hearts can be converted pretty rapidly.
Guy Nadivi: Mark, your epilogue describes the book as thought provoking and scary because of how much it gives the reader to consider. Nevertheless, I’m going to ask you a question that I think all the CIOs, CTOs and other IT executives listening in want to know. What is the one big must have piece of advice you’d like them to take away from our discussion with regards to managing IT over the next decade?
Mark Settle: So what I tried to do in the book was to kind of paint a landscape of all the secular changes taking place in the industry, and the book as you know Guy it’s divided into three sections around people, process and technology, nothing terribly novel there. We all know that we have talent issues within our organizations, and recruiting challenges, and huge demands for personal development that need to be addressed on the people’s side. In terms of internal operational practices, we touched on several things in the course of this conversation around how do you manage SaaS applications, how automation can be brought in to not only transform IT but different kinds of business processes, and other, other significant operational practices around security cultures, tech debt reduction, et cetera, so that’s on the process side. And then in terms of new technologies, and again, we’ve touched on some of this, things like machine learning and AI predictive tools, the ability to actually get out of the data center business altogether and use the cloud infrastructure as a service, it’s a very popular thing, and the lurch that we’ve made towards almost a completely virtual workplace through the COVID crisis and people are going to undoubtedly want to continue to virtualize their work environment, and their work practices as well. And then there are other challenges. So my advice to the listeners that I try to present in the book, no IT organization is going to be best in class at all of those things. But I think you have to look at your own business model, and the challenges facing your company, and as the CIO or a leader at any level of the organization, you have to place your bets. And you have to choose, intentionally choose a handful of these topics and really double down on “We are going to be best in class at SaaS management, and security, and compliance. That’s the key of success for us as a business. And maybe getting our data center closed down that’ll happen over time. That’s not something I want to get done in the next two years, that’s going to have that huge set of dividends and pay offs to my organization.” And I could make the same point with regard to some of the other topics that I just listed. So, I mean, we talk about a lot of these things all the time. And it’s another point I make in the book, if you don’t seize these opportunities, you’ll be victimized by them, and you’ll get buffeted around and expect to do all these things. And you’ll be challenged by your business team, and your business peers. “Well, how did that security problem emerge? Why weren’t we on top of that? Why is my buddy at this other company deployed a thousand robots, and we don’t? We have three that are working on the backend of the HR department.” and whatever those questions are. If you don’t step up to the challenge and show some leadership, all of these problems and issues are going to confront you to one degree or another. And without leadership, you’ll just get buffeted around and really end up not being good at any of them, which is the worst possible outcome.
Guy Nadivi: Sage advice. All right, looks like that’s all the time we have for on this episode of Intelligent Automation Radio. Mark, what a pleasure it’s been having you on the show today and hearing the insights that only a seven-time CIO could provide. Thank you for joining us, and congratulations on publication of your second book.
Mark Settle: Thank you Guy. I really appreciate the opportunity. Thanks for the invitation.
Guy Nadivi: Mark Settle, author of “Truth from the Valley, a Practical Primer on IT Management for the Next Decade” now available on amazon.com and other book retailers. Thank you for listening everyone, and remember – don’t hesitate, automate.