Norm Fjeldheim is Senior Vice President and Chief Information Officer for Illumina. Norm is responsible for evolving a Global Information Systems (GIS) strategy that will support the scale and future growth of Illumina, including supporting product and market development strategies from an Information Technology perspective. In addition, Illumina is investing significantly in global business processes, and Norm will be instrumental in the success of this program, including implementation and sustaining support for ERP and related ecosystem business applications, integration services, business intelligence, master data management and business process management.
In addition to his IT responsibilities, Norm is also Head of Global Facilities Management Services. This includes Real Estate, Security, Engineering and Operations as well as Health Safety and Environment.
Prior to joining Illumina in 2016, Norm was Senior Vice President and CIO at Qualcomm since 1987. He served as Manager, Director and Vice President of Information Technology. He was instrumental in the creation and implementation of systems to support Qualcomm’s growing and diverse corporate needs. Norm and the IT department guided the selection and implementation of technology to link Qualcomm’s corporate sites across six continents.
Norm began his career as a systems analyst at Unisys Corporation and was a programmer analyst at M/A-COM Corporation.
Norm serves on CIO Advisory Boards for Salesforce.com, Hewlett Packard Enterprise and Amazon Web Services and has presented at numerous industry conferences on challenges currently facing CIOs including CTIA, Oracle World, The Economist CIO Agenda and IDC’s Data Center and IT Forums.
Norm holds a Bachelor of Science degree in business, information systems from San Diego State University. He also completed the Wharton Executive Development Program at the University of Pennsylvania.
Illumina is improving human health by unlocking the power of the genome. Our focus on innovation has established us as the global leader in DNA sequencing and array-based technologies, serving customers in the research, clinical and applied markets. Our products are used for applications in the life sciences, oncology, reproductive health, agriculture and other emerging segments. To learn more, visit www.illumina.comand follow @illumina.
As the CIO of Illumina, what’s the number one area that you’re focused on right now?
Everything starts with the organization. Since I got here, I’ve been here almost two years now, and I’ve been focused on what is the organization good at and not good at. Who are the people that I can count on? Building up a leadership team. Then building up the strength of the organization.
I measure that in a number of different ways. Customer service, what is our customer service score, and how does the organization view us? Having a good customer service attitude, we get stuff done. I look at the capacity of the organization. One of my goals when I first got here was I wanted to triple the number of projects we were doing in a year. We accomplished that in first year.
I also look at what’s the morale of the organization. Do people want to be here? What’s the turnover rate? Couple different ways they can measure that. Obviously, you can look at turnover, and people want quality candidates coming in that want to work here, but I also like to go out and kind of get an outside survey. Computer World runs a Best Places to Work in IT survey, and it’s a pretty extensive one. It actually has meat behind it, and so I did that the first year that we were eligible to do that, and we made the list the first year out of the gate that we applied, so that was gratifying that employees viewed us as a good place to work because that’s where it comes from.
How is IT structured? Do you have a centralized IT, where you’re servicing internal employees within Illumina, or does IT also align to the business units and your built-in solutions that are customer facing?
One of the things that I found when I got here was that the organizational structure of IT was very confusing. It was confusing for the employees, and it was confusing for the people in IT, so I simplified the structure and built an organizational structure. It’s a matrix structure, but from an outside-in perspective, it mimics the organizational structure of the company.
We have functional organizations. We have a finance organization. We have a manufacturing, and so on. I have a finance IT organization that lines up with finance, and one with operations, and so on. That makes it very easy for people to understand, who they work with. People on the operations side, they know who their IT folks are that they work with.
Two advantages. It does clarify everything so people know who their customers are, and who the IT person is that they need to get work from, but it also breaks up into multiple channels of projects. I can run multiple finance projects independent of multiple operations projects, which are independent of multiple commercial sales projects. That’s partly how you get this big increase in project capacity. You don’t try and run everything through a central team. You break it up and have essentially multiple swim lanes, multiple project tracks, and then each group can operate relatively independently of each other, and different groups can have a different cadence.
Finance tends to be a little bit more conservative and risk adverse, so they want their projects to be a little bit slower, a lot more testing, more of a waterfall approach from the project management. On the ops and commercial side, particularly commercial, they’re much more comfortable with an agile methodology. They like fast, iterative projects, lots of little projects that add up to a lot, and ops is kind of somewhere in between. We can have different cadences set up by having not one central project management office, but actually break up the project management, and each team has a project manager, they have testers, they have developers, they have all the resources that they need to get the projects done.
It’s an organizational structure that’s optimized for speed and capacity and customer service because it aligns directly with how the customer, each functional group, wants to work with IT.
How do you bring that back, in all of the different infrastructure that’s being built?
The other half of the organization is essentially the enterprise side of that. There is an enterprise architecture team that governs the technology roadmap for all the organizations. We maintain technology roadmaps, and we all get redone systems, and they maintain centers of excellence. The technology gets deployed the same way into different groups, and then they run the kind of common or the enterprise-wide systems there, so learning management, identity management, applications services, and microservices come out of that group.
Then, beneath that is IT infrastructure. That’s servers, data centers, network, customer service. Security is in that group, so those are enterprise functions. It is a hybrid organization. The bottom layers are horizontal, cut across the infrastructure and cut across the entire organization, and then the applications layers are essentially vertical and align vertically with the customers on the business side.
It’s a pretty effective model in a high-growth company, where you don’t need the IT to be super cost efficient. This is not cost optimized. It’s still pretty effective cost-wise, but it’s really optimized for speed of delivery and keeping up with the business when the business is growing as fast as Illumina’s is.
What’s your segue with scalability growth and all that good stuff? How do you deal with, how do you take on public cloud, scaling resources?
We’re actually one of Amazon’s largest customers already. I’m on their advisory board. They have a CIO advisory board of about 15 CIOs. I’m one of those. That gets me kind of executive connections with AWS, but we’re just a big customer. I’m a big believer in the cloud and have been for years, but I’m not an “everything has to be in the cloud” or “everything has to be on prem.” I don’t believe in one size fits all for anything. You’ve got to look at each business problem and determine what the right solution is.
Right now, we run an on prem kind of an engineering support infrastructure that handles a lot of the data from our instruments. Our instruments produce a ton of data. That infrastructure, on a per storage, per terabyte, per compute cycle, is actually cheaper than Amazon, so it doesn’t make sense for us to move to Amazon just on a pure cost basis. But there may be features or functionality that we want that would make it worthwhile. We evaluate that.
Do you look at Amazon as a long-term storage solution to potentially development, or how are you leveraging Amazon?
We use it for several types of things. We’re doing some AI work, machine learning type of work with them as well. We’re taking advantage of Amazon’s capabilities. I don’t think of Amazon strictly as storage and compute. If you think of them that way, then it’s just a cost equation, and often, you can do it better than them from a cost perspective, and we certainly do. But when you start thinking about them as a platform, and there’s all these capabilities that they have that you can tap into, now you can start looking at different value propositions. The graphic adoption, more capabilities, more services around your application that you don’t have to build.
I can give you an example. We had a situation where an engineer in our development environment, a fat-fingered setting, opened up the server to the public. It was completely exposed. Amazon told us about that situation 15 minutes after it happened. There’s no way I would have been able to monitor the environment that closely to be able to do that. If you need that kind of monitoring capability, you want to take advantage of Amazon, but that’s not in the base offering. Some of it is, but some of it you have to kind of move up the stack and add those value-added services and determine whether that makes sense for you to do. Sometimes, a lot of times, it absolutely does.
They were already using Amazon when I got here, so it was not a matter of convincing anyone that we needed to go to the cloud, and in fact, for the most part, the business leaves those technical decisions to us. They don’t try and get in the middle of those decisions. They trust us to be the experts, and it’s more about just being able to do this in the business. We need to enable this kind of business capability. “Okay IT, go figure what’s the best solution.” If we come back with, “We’re going to do this in the cloud,” they’re like, “Okay.” And if we say we’re going to do it on prem, they go, “Okay,” as long as it works.
I don’t have to get into technical decisions typically with the business. There’s some exceptions. There’s different groups that like to get into the middle of the technical discussions, but for the most part, they’re looking for us to provide a solution, and they would just want to give us the requirements.
When I got here, I realized that we had very, very little reporting capability. I couldn’t believe it, how little we had. We had just implemented SAP, and it really hadn’t built up the reporting side during that implementation. We also didn’t have an enterprise architecture function. We didn’t have any, so I built that, and then we ended up implementing an analytics platform and a reporting platform. We actually deployed that in the cloud. We ended up purchasing HANA and ended up initially starting to, and had SAP run it in the SAP cloud. So we’re running SAP the application on prem, migrating the data into SAP’s cloud, and then we found out that SAP doesn’t really do that very well. What they know how to do is SAP on top of HANA in their cloud. What we were doing is HANA as a data warehouse infrastructure and engine, so we took that over and did that ourselves and just migrated it recently to AWS.
Our analytics platform is now at AWS, and what I like is to enable it to something I call self-service IT. Other people call that shadow IT. I call it self-service IT because you give them the right tools, you give people outside the business the right accesses to the data, and then say, “Go ahead and have fun. Write your own reports. Do your own dashboards. Do your own analytics,” and we give them the tools to do that, we standardize on Tableau for that. HANA’s the backend. 20x improvement in performance going off SQL server to HANA. Direct translation of the data from SAP right in the system. We control the access, and now we’ve got hundreds of people writing a report, and that was the only way I was going to be able to get through the backlog of reports. If IT was going to have to write them all, we never would have got them done.
It’s a model I’ve used in the past, different technology this time, being able to go to HANA. What used to happen, and it still does to a degree, but we’re in the process of consolidating, was operations had its own data warehouse and its own IT team running and doing reporting off that data warehouse. Then supply chain had one, and commercial had one, each region in the commercial division had one, and finance one. Data spread out everywhere. Who had the right answer? All kinds of duplication of effort.
It’s all now centralized. It’s all in one place, so now we can control one data point. Let’s say backlog. You define what your backlog number is. You built that into the HANA database. It’s calculated automatically every time you update the data, and then everybody can report on it, and everybody’s reporting on the same network. Pretty amazing.
What’s your take as we see this all unfold, and we see economies of scale, the Amazon’s, the Microsoft’s, the Google’s, how do you weigh your options between those players that are probably top tier in market?
That’s an interesting debate that you see take place within CIOs, and some CIOs will say, “Well, I want to avoid lock-in,” and so they will develop to the least common denominator across those platforms. They’ll invest heavily in containers or whatever so that they can, in theory, migrate their workload from Amazon to Microsoft to Google, and then they could negotiate pricing on that.
I actually don’t believe in that. When you get large data sets, we’re talking petabytes of data, we’re already petabytes and petabytes of information, you just don’t move that. That’s just not easy, but then going back to my earlier comments about, here’s all this functionality that Amazon’s developing. They’re not developing that. Microsoft’s not developing that same capability, so if you say, “Well, I’m going to not take advantage of that so I can port,” now you’re losing out on all that innovation that these companies are putting into their platforms.
We, as an IT industry, have been dealing with vendor lock-in our entire careers. Whether it was Oracle or SAP or IBM or whoever, that’s just a fact of life. This idea that we have to avoid vendor lock-in just because it’s the Cloud makes no sense to me. What you want to do is you want to have good negotiations. You have good contracts in place. You stay on top of that, and then you want to maximize the capabilities of what you value. You can extract from that investment with that particular supplier, whether that’s an SAP or whether that’s an Amazon or whether that’s an HP or whoever you’re buying that equipment from. You wouldn’t go in and say, “Oh, I’m going to not take advantage of a bunch of capabilities with an HP server just in case I want to go buy a Dell one.” Now you’re losing the value of that particular server.
I look at it very much as, again, what is the right solution. Google, for example, is doing a ton of work, as is Microsoft, around AI and machine learning. What they’re developing is different than what Amazon is developing, and in some cases, it might be even a little bit further ahead than Amazon. It’s possible that we would be saying, “All right, for that workload, we’re going to take advantage of that at Microsoft or Google rather than Amazon.” There’s no reason why we couldn’t do that.
I don’t believe in that, “Oh, we’re going to pick one partner and that’s it,” or “We’re going to not pick a partner.” We’re going to try and manage this, and we can pick any at any time. That’s my philosophy on the cloud. Then, you look at your own on prem, and that’s a decision. I’ve never believed in segregating my teams and saying, “Okay, here are the cloud guys and here are the on prem guys.” It’s the same team, and the cloud and your data center, they’re just tools in your toolbox, and they have to understand what’s the right tool for the right job. I always have done that very early on in my career because I wanted the people to own those decisions, and say, “Okay, yeah, I’m the expert. I know whether a data center, on prem, or Cloud is the right solution for this particular business plan.”
Everybody has the same challenges being in San Diego. What type of hiring challenges do you face?
That kind of goes back to where I started. You want to make this a great organization. I like to build world-class IT organizations. I pride myself on that, so you’ve got to make it a good place to work and all those things. There’s really three factors in that.
It’s the type of work the people get to do, the people they get to work with, and the manager that they work for, and then the environment that they work in. If you can make all those good, then you generally get top talent wanting to come to you, and you get that top talent wanting to stay. First and foremost, you see more people quit because they don’t like the manager. They don’t like their boss. That’s number one.
I started right away on assessing my management team. It was the first thing I did when I got here. Could they do the job? Could they attract the talent? Could they build the team? Could they manage a team? Could they make all these decisions that I’m talking about?
That’s your starting point, and then, you build out your culture from there. I developed Qualcomm, a large, growing company, where a lot of people in San Diego were interested in joining that. For many years, I said I had the best job in San Diego, and now, I can say I have the best job in San Diego because I’m at Illumina, and we’re literally changing the world. It’s interesting, coming from a tech company, at Qualcomm, we were being able to change the world through mobile devices, and it certainly had a great impact on the world. I feel better about the mission statement at this company.
There’s not a person here at this company that doesn’t talk about something personal. Whether it’s a disease or a family history, or they have family, or they have friends that haven’t been impacted by something that we’re working on to try and cure, cancer being kind of the one that’s foremost. We’re involved in so many different things. Genetic, the genome, and personalized medicine. Personalized medicine starts with your DNA.
How far do you think we are from each of us in this room having our own genome sequence?
I’ve had mine done.
At a cost that’s affordable to everyone?
There’s two ways to do that. You can do the partial genome sequencing, so that’s Ancestry and 23 & Me, and that gives you some information. That’s useful, and they can tell you a number of factors. I know from that analysis that I have that I’m a better sprinter than an endurance runner, so I’ve been able to adjust my exercise routine appropriately. My wife has sensitivity genetically to certain foods that she’s learned because of that.
When you do the whole genome sequencing, now you can get even more information. We’re doing everything we can to bring that price point down so that it’s as cost effective to do the whole genome as it is to do partial. Once we get to that point, now you’re almost like, “Why do we even bother with partial. You just do the whole thing and get all that additional information.” I learned from the whole genome, there’s a particular drug-thinning agent that they would give people that’s having a stroke. It doesn’t work on me. That’s really good to know if you’re ever having a stroke, they don’t give you the drug in the hospital that isn’t going to work, and then three hours later, they realize it doesn’t work, which is what would normally have happened. Now I know that.
I know what my predispositions are on certain kinds of diseases. One of the most interesting one is when you’re going through kind of the prep, they’re talking about, “Well, what do you want to know? We can learn a whole bunch of things.” One of the questions was, there’s a particular genome combination. If you have it, you will get Alzheimer’s disease. It’s not a 50% chance, no, it is 100%, you just don’t know when and you don’t know how badly. Do you want to know that? That’s an interesting question to think about.
Is that available today? Did you want to know?
Yeah, that was one of the questions they asked me. Yeah, absolutely. I figured I’d forget it anyway.
How did you transition from Qualcomm and get all of this new knowledge?
For me, it was a blast. It’s been so much fun. Illumina is a high-tech manufacturer. We produce an extremely complicated instrument, and then we produce the enzymes that are used in the DNA sequencing process. In that aspect, it was very familiar. Qualcomm ran a very complex supply chain. It operated a scale, so a lot of the same capability, and even earlier in my career, Qualcomm was a manufacturing company. We manufactured cell phones, base stations. There were a lot of skills that transferred directly over.
Then there was the whole new industry, learning all the terminology around genomics and how do we sell, what’s the different selling models. There was enough of the familiar that I was able to come in and hit the ground running and feel like I could make a difference right away on the IT side, and at the same time, there was all this new, fascinating, just wonderful stuff that I got a chance to learn about. Then, to top it all off, they gave me facilities, too, and I was like, “Oh my god, I’m in heaven,” because now I can work on building and do that. I had actually started off in construction as a kid, so it was like coming home. It was enough of the known coupled with enough of the unknown to make it just fascinating.
What did you want to grow up to be when you were a kid?
I was going to be a contractor like my dad. General contractor, building buildings. I was fourth generation construction, and I was the first one in generations that didn’t actually go into construction. My grandfather built the Hoover Dam, worked on the Hoover Dam. It was crazy stuff like that, and somehow, I ended up in IT, and it turned out to be the greatest move I ever made.
Where did you grow up?
I grew up in Northern California, right in the heart of Silicon Valley, so I knew all the names. Hewlett Packard and Intel. They were all neighbors. I came down to San Diego to go to school, and I was taking business classes and it was actually business management. I wanted to learn how to run a construction company, and oh my god, it was so boring. I was just dying in these management theory classes. There’s 800 people in the lecture hall, and one of the required classes was a programming class. I was scared to death. I though, “Oh my god, I don’t know what this is. I don’t know how to do it.” I took the class and it just clicked. It was just like, “Wow! This is just great!” I got an A in the class, and I said, “I can get a degree in this. This is fun enough. Management stuff is horrible,” and at that point, I really just wanted to get the degree and then go into construction.
I had one interview, just to kind of get some interviewing experience, and they offered me a job. $25,000 a year, I could stay in San Diego where my girlfriend was, and I didn’t have to work in the dirt. It was like, “Why wouldn’t I want to take this job?”. I took the job. I did a couple of different jobs, bouncing around, getting some experience, and then I joined this little startup called Qualcomm, which nobody had ever heard of. It was 90 people when I joined the company. It was just crazy, and I was the first IT guy they hired. Right place, right time, and the right combination of skills that they were looking for. Boom! I couldn’t have done this job if it hadn’t been for Qualcomm because I learned so much going through all of Qualcomm’s various iterations and business models that they did.
What are any big mistakes that you’ve learned from?
Let’s see. God, there’s so many. There were times when I put loyalty over kind of the person who isn’t really getting the job done, and when you’re in a high-growth company, it is very difficult for every single person in the organization to grow as fast as the company’s growing, personally. There were times when the organization grew faster than the individuals in it, sometimes the leaders. We’d have people that were very successful and great leaders at a certain size, and once you got to the next level it became much more challenging.
I struggled with that, being able to recognize that and trying to make the right organizational decisions. Some of those became painful. I let them go too far, and then it became really challenging and emotional, whereas if I had acted on it sooner, I could have put the person into a different role in the organization that was better suited for their skillset versus ultimately, a couple cases where I had to terminate people, and they were people that I had been friends with for a long time. That became difficult.
It really is back to people issues and organizational issues that you need to address and be timely about that. When I took over the entire IT organization at Qualcomm, there were a lot of problems with the organization, and there were morale issues and a lot of things, and I realized that I needed to do a lot of housecleaning. I ended up doing some layoffs, and what was interesting about that was when I did the layoff, morale went up. People started looking up going, “Oh wow! Management’s finally dealing with the folks that are not carrying their weight,” and the performance went up because the top performers were getting rewarded, and they weren’t having to carry people along.
It was a revelation, where you really saw this complete shift in the organization by actually focusing on the bottom portion of the organization. You’ve got to take care of your top performers. A lot of cases, top performers take care of themselves because that’s why they’re top performers. It’s the bottom performers that end up taking 80% of management’s time. Actually pulling the trigger and acting decisively on that makes a huge difference for the organization because now they see management being decisive and making good decisions and acting on the problems.
I did that, and it became kind of a mantra within the organization, and just became something that we did collectively and culturally. People ended up getting a sense of pride being there. “Hey, we’re the elites. We’re part of a really strong organization. Not everybody can make it here.” The few, the proud, the Marines, whatever.
One of the nicest compliments people would say, “Norm’s really tough, but he’s fair.” I could care less whether they’re purple or from Mars or whatever, as long as they got the job done, then that was what mattered. Early on, there was a quote that I heard from one of the managers, a really smart guy at Qualcomm, one of the senior guys. He said, “Never confuse effort with results.” I’ve had some of those conversations. “I’m working 60-70 hours a week.” But you’re not getting the job done. Now you’re just not working effectively.
We’d kind of rank, we force ranked kind of the bottom 10%, and I never said you had to get them out to my management team. It was, “You have to have a plan. Either improve their performance, get them into the right job if they’re in the wrong job, or exit,” but it wasn’t, “You have to exit.” It was, “Fix it. Fix the problem.” Whatever the best way to fix it was. There were plenty of times when I thought, “That person, there’s no way they’re going to make it,” and we shifted the roles, and they took off. There were times when I would have bet, this person’s going to go and just knock it out of the park over here, and they would crash and burn. I learned not to judge, prejudge, and that was another one of the mistakes I made. There were situations where I kind of prejudged and thought, “Yeah, I know what’s going to happen.”
If you were on a deserted island, what three things would you bring and why?
As long as I have Internet connectivity and a computer. You said I could bring anything I wanted. A satellite communication, whatever it is. One of my hobbies is I still kind of have that architecture, design gene in me, so I do 3D graphics, 3D art. Pixar, DreamWorks kind of stuff. I have all that software. I do animation and people and superhero characters and animals and landscapes and buildings. Whatever I just kind of get into, so I would just have a lot of fun. That’s the creative outlet that I have. I do digital sculpting with a program called Zebrush. I learned lots of different tools that they use in the entertainment industry, and then some cheap ones too. I don’t buy the really high-end stuff, but anyway, that’s my hobby, so as long as I have that, I could probably live anywhere. I read all the time. I don’t have to have a book. I like books, but I don’t have to have a book, so I can read on my computer. I’d just order food from Amazon.
You need a solar cell to power the computer, and then you need a satellite uplink to pick up your internet off the satellite. I’d actually probably include my wife because I would want to have somebody to talk to. I can only stay buried in my computer for so long. She and I have been 35 years together, so she’s not sick of me yet.
If you could give guidance to the world of CIOs and VPs of IT, what would you give them in terms of how to look at their careers and skillsets?
I would say the main thing is you’ve got to get a diversity of experience. You can still do that within IT, staying in the IT field, but different companies, different jobs. Qualcomm was a unique place because every few years, the job just completely was different just because of the company, but I would never recommend anybody stay on in a company 28 years, or even 10 or 15. It’s actually a negative on a resume now to see that because you don’t show that you can succeed in multiple places with multiple cultures. You can only show that you’ve succeeded in one place. Now you’re a much more risky hire to move to another company. I don’t know that you could succeed there if you’ve only succeeded at one place.
The culture here is different here than Qualcomm’s. The decision making process is different. I definitely had to adjust. My boss talks about how he watched me adjust my first year, and he said he was actually more pleased with that than the accomplishments that I made because it showed that I could actually modify my behavior to fit Illumina.
It wasn’t that the decisions I was making were wrong. I ended up implementing the decisions that I made based on my initial assessment. I did the 100 days. First hundred days, you do that evaluation, and at the end of that 100 days, I was ready to make a bunch of organizational changes. I hit a wall. The organization wasn’t ready for me to move that quickly, and it took me another three, four months to actually then evangelize the changes I was making, convince people that these are the things that made sense and here’s why, and then implement those changes. I had to adjust to the decision making and the cultural style and the speed at which the company was comfortable operating at.
I couldn’t expect the company to conform to me. I had to conform to the company and the culture. I learned, but I was open minded enough to say, “It’s not Illumina, it’s me. I’ve got to go adjust,” and then figure out how to get it done within Illumina’s structure.
Now that I’ve been here a while, and I’ve built up trust and people are getting more comfortable with me, the organization’s much more comfortable with me, that decision making process has sped up. It was that first year, I was the new kid. I had to get people to kind of rally around me, and put the right leadership in place that was able to deal with that kind of speed. Now, we’re way faster, three times, four times faster. I hope to get another 25-40% more capacity out of the organization.