This week I’m joined by Mark Grimse, the Vice President of IT at Rambus. Rambus is an engineering and intellectual property company with many different revenue models.
In our discussion, Mark takes a pragmatic look at how new technology such as RPA, ML and AI play a role in a 25 year old company. The answer is not as simple as you might think. We also discuss the role startups play in scouting the horizon. As Mark put it: We didn’t have a taxi cab problem until Uber arrived.
Mark Grimse Twitter: https://twitter.com/markgrimse
Mark Grimse LinkedIn: https://www.linkedin.com/in/markjgrimse/
Tim Crawford: Hello and welcome to the CIO In The Know podcast, where I take a provocative but pragmatic look at the intersection between business and technology. I’m your host, Tim Crawford, a CIO and strategic advisor at AVOA.
Tim Crawford: This week I’m joined by Mark Grimse, the vice president of IT at Rambus. Rambus is an engineering and intellectual property company with many different revenue models. In our discussion, Mark takes a pragmatic look at how new technologies such as RPA, ML, and AI play a role in a 25-year-old company. The answer is not as simple as you might think. We also discuss the role startups play in scouting the horizon. As Mark put it, we didn’t have a taxi cab problem until Uber arrived. Mark, welcome to the program today.
Mark Grimse: Thanks, Tim. Glad to be here.
Tim Crawford: So, Mark, you are the CIO for Rambus, and I was hoping you could kind of start off by telling us a little bit about who Rambus is, and then a little bit about your role as CIO at Rambus.
Mark Grimse: Sure, Tim. I think maybe a lot of people are knowledgeable about Rambus’s past. We have been a, traditionally an IP licensing company. About six or seven years ago, however, we started to shift ourselves and become more of a product and a licensing company, and try to balance those two. It’s a big shift to do that, and we can talk about that. And my role specifically here is covering all of IT. I cover the worldwide IT, support business systems, applications, as well as the engineering infrastructure, and we do spend quite a bit of time supporting our engineers, because we are basically an engineering company.
Tim Crawford: That’s great, and I want to dig into that a little bit further, because it’s not just an operational role that you’re filling, but you’re also taking on some different aspects. How have you seen that CIO role and your CIO role at Rambus? You’ve been at Rambus for a number of years. How have you seen it evolving over time?
Mark Grimse: When I first started, I was IT guy number 14 here at the company. We did not have any business applications whatsoever, we were almost 100% focused on engineering support. My role has been to shift some of those focuses to include business applications, and now we have the full suite of things that you might imagine any other company have. However, when I started, as I said, we didn’t have business systems, we didn’t have an order management system, and the CEO would come and say, “Hey, when am I going to get an order management system?” And I was kind enough to think about things and say, “Would you like an order management system for both orders you got this quarter?” Because we got very large licensing orders every quarter, but there was not the volume, there weren’t parts, there weren’t items. So we, you know, 25-year-old company, started to implement these systems, and so now I’m… Obviously we moved out of just supporting engineering, and now we support order management just like anyone else.
Mark Grimse: But my role hasn’t stopped just at doing what you would normally think as an IT guy or a CIO doing, in doing the business systems and the infrastructure. I have gotten to play a significant role in some of the acquisitions we made. Because IT is very knowledgeable across the board, I’m able to bring the right types of questions into a due diligence aspect of a deal, as well as I led the integration efforts that Rambus has done in some acquisitions, not just for IT, but the global integration leader.
Tim Crawford: And it seems like there would be a bit of complexity, because on one hand, you’re licensing IP, but on the other hand, you’re also a product company too, right?
Mark Grimse: Absolutely. There is a balance that needs to be made between the two, but we think them as being very complementary to our customers, and that they, some of them are very strong and are able to take our technology that they’ve licensed from us, and they’re able to take that, run with it, and deliver products with it. However, there are others that maybe have expertise in other areas, and they need that technology to deliver their products, but they aren’t specialists in that. And then we can come in and say, “Okay, here’s some products that instantiate that technology that you can then leverage to help deliver your products,” and they don’t need to be specialists, even though they need that technology to be delivered.
Tim Crawford: Hmm. I find it just fascinating, because this is a common problem, is when you have different models to your business. Whether it’s a product and service or IP and product, it creates a bit of consternation, doesn’t it, in terms of how you think about systems and systems thinking, and how they come together, and everything from sales models all the way through to marketing, and then the underlying components.
Mark Grimse: Yeah, it absolutely does. In trying to prioritize activities, who is the lead dog? You know, all the other dogs, the view doesn’t change. However, the lead dog in this particular activity, which does it need to be? Let me give you an example: revenue recognition. For our chip business, revenue recognition is relatively easy because we build chips, we’re a fabless semiconductor company, and then someone buys the chips, we ship them, and revenue recognition is fairly straightforward in Oracle, and easy because you ship it, you invoice it, you can recognize that revenue.
Mark Grimse: However, when you think about doing different business models in which maybe you’re selling some services as subscription, or you’re doing maintenance on some of your products, or you’re doing a NRE type or nonrecurring expense type project, in other words, you’re just building it for someone, or you’re doing a percentage of completion project or a milestone-based, you might have lots of different triggers to recognize revenue. And recently, we’ve changed along with the rest of the world from 605 to 606 revenue accounting. The SEC put out those new rules, and that was a fairly significant change for us. So, all of a sudden the patent revenue had to be significantly changed in how we did that based on those rules. And we’re still going through, and we figure we have seven to nine different revenue models within a company, and that’s relatively hard when you’re not of scale in a lot of them. So you have to figure out, again, how do you prioritize?
Tim Crawford: Yeah. So, as you kind of go through this process, and I’m thinking about how you engage with other parts of the organization at Rambus, where are you finding that the CIO plays the best role or plays the strongest role in the organization? I mean, there are obviously a lot of roles that the CIO can play, but if you were to kind of put your thumb on, “Okay, you know, that is true, but here are the aspects that tend to be the best place to be,” where do you find that to be?
Mark Grimse: I think the strongest point or best role that the CIOs and IT in general can play is at the weakest point of the businesses, right? You’re part of a team, and each team player has their specific role. You know, the marketing guy is really good at driving and knowing about demand and brand awareness, and different types of sentiment analysis, or whatever it happens to be. That’s their specialty. And the sales guy doesn’t know that well, but he knows he needs it, but he knows how to close a deal. So I think the CIO role is best at the weak point of the two, and that is the joining of where sales and marketing meet, and that’s where I think we can play as technologists and solution providers. I think that’s where we play, is how do we get the company to be glued together, because otherwise it would just be silos in each of these different areas.
Tim Crawford: That’s a fascinating way to look at it, look at the weakest point between these strengths and find ways to be that connective tissue.
Mark Grimse: Yeah, I think that’s what we… You know, I’m really, if you think about it, I kind of get upset when we think about ourselves as information technology, because that’s not… That is part of what we do, but in reality, what we should be aspiring to do is we provide solutions, and it just happens to be our toolkit is mostly filled with technology, but we also bring solutions that are part of change management, because change management is part of our job, or organization. Yes, we partner with HR when we’re doing a big project and saying, “Listen, you need to do this. If you’re going to implement this system, it’s going to change processes. It’s also going to change the organization. So, how are we going to do org design around it?” So I think that we should just be, you know, a solution provider and not just technologists.
Tim Crawford: Yeah, and the other piece that I’ve brought up in past conversations before, and it sounds like you’re thinking along the same lines, is you also are seeing across the organization. So when HR wants to run that project, you have the ability to see how that might impact other aspects of the organization that HR may not have visibility into.
Mark Grimse: Yeah, absolutely. You know, if we think about some of our core innermost things in technology that we used to do around… Think about ITSM, you know, the CMDB. We needed to know all of our assets, and we need to know how they all play together, and we can tell you which router and which server and which storage unit all work together. The business doesn’t really think about how their processes all work together, and so you need to be able to have that dependency model for your business processes. Right now, most businesses don’t have that level of documentation that they’re able to figure it out. However, by knowing what goes on in HR, what goes on in finance, what goes on in legal, what goes on in marketing, you as an organization have kind of pieced this all together, and you can bring that dependency model to the changes that are required, and so that people then can see, “Oh, now I see when I make this change, all the things within the world, you know, my business world, they need to think about it.”
Tim Crawford: That’s awesome. And I have to imagine that one of the pieces that you can also bring to the table, in addition to that technology piece, is innovation. You know, there are many ways that people get sources, or pull in sources of innovation into the conversation. How do you bring innovation into this conversation as you start talking about the different groups, you start working with them as part of a cross-organizational effort? And then also, where do startups play in this for you?
Mark Grimse: First, I think as a CIO, you always have to be on the lookout for new technology. And it’s not like “I have a problem, I’m going to look for a solution for XYZ.” Sure, you do that as part of your job, but I think you always have to know what’s going on on the horizon to be able to know what problem you have yet to see that you can apply this technology when required. So you’ve got to have that, I guess, scouting effort, whether it’s you or someone within your organization, out there looking at the horizon and figuring this out.
Mark Grimse: And startups, to me, are a great way to be able to see on the horizon, because that’s where the VCs see the market going, and that’s where they’re funding, so that’s where the startups are going to, so it’s a great… You know, startups are built by people who see problems and are trying to solve them, and so you may not know you have that problem, but a startup will, and a VC will fund, and then you’ve got to be able to see that on the horizon.
Tim Crawford: That’s actually the next question I was going to ask you, just for clarification, is do they necessarily just focus on the problems that you have in front of you, or is it a great way to see problems that you haven’t necessarily encountered yet, or don’t know that you might have, or might be coming down the pike? You know, as a way to kind of open your eyes to, “Ah, so others are dealing with this, maybe I should be looking at it too.”
Mark Grimse: Yeah, there’s always the thought about, “Wow, I didn’t realize I had that problem until it arrived,” right? I mean, we didn’t know we had a taxi cab problem until Uber showed up.
Tim Crawford: Let’s pause there for a second. “We didn’t have a taxi cab problem until Uber showed up.” I love that line. That’s classic.
Mark Grimse: But it’s true, right? I didn’t go to New York on vacation, wanting to go to a show, thinking, “Wow, I’m standing on a corner with my hand raised up. Why isn’t no one showing up? This is a problem.” So, there are people who have different minds, and that’s why diversity is really important, to have these different thoughts. And somebody said, “Yep, there’s got to be a better way to do this. I have a device in my hand that I can call a taxi in an instant. Let’s do that.”
Mark Grimse: So I think about the startups as having people with different thoughts who can bring a problem to me and say, “Gee, we should think about this differently.” Here’s an example that I think just popped in my head the other day. We do expense reports, and it’s… Everybody will complain about expense reports. “I have to do a receipt, and I have to put it in, and I have to get approved.” And people have solution providers, ISV has come in to make that easier for us; however, no one has really changed the paradigm around expense reports.
Tim Crawford: Mm-hmm (affirmative).
Mark Grimse: We all think about doing expense reports because that’s the model that has been used from, you know, probably the ’50s or even before that. But in reality, I don’t need to do an expense report. I have an expense, so why can’t I expense and get approved each item as I go through, and as I take a picture of it on my phone, just send it off? Why do I have to wait for a report to do it? If I do it individually, does it matter that they’re grouped or not? We can group that through metadata on the back end. So, why do I do expense reports? It should just be a bunch of expense items.
Tim Crawford: Yeah, and there are so many of those processes that I know both you and I have had to contend with for not just years, but decades, that create a lot of friction and frustration within the organization. And it just seems like there should be something that can disrupt that process and make those experiences more frictionless, because at the end of the day, the expense report is not something that’s that interesting. You know, it’s more of a protectionism, making sure that someone hasn’t gone off the rails, per se. But there are ways that you can automate that.
Mark Grimse: Absolutely, and we just have to think through or think differently about them, or be on the horizon and see what other people think that that solution might be looking like.
Tim Crawford: That’s actually probably a good place to bring in newer emerging technologies like RPA and ML and AI into this. And I know that’s something that you’re passionate about, is in that same kind of genre of RPA, ML, AI, if you look at it as a continuum. But how are you leveraging those emerging technologies within your organization in different ways?
Mark Grimse: I think that the first way that we’ve been able to find some value in this space is by serving our employees, not our customers, because I don’t think it’s… We’re not mature enough, and the technology may not be mature enough to be able to use this in our products and our customers. But within IT, we can be kind of the test bed, and look and see how this works. And with most of these solutions, put RPA off to the side for a minute, it’s really about how well you are testing or training the system, and so if you have lots of data that you can train the system, then you can start to utilize it more effectively.
Mark Grimse: As I mentioned previously, we didn’t have a lot of orders. We still don’t have a lot of orders within our order management system, so that’s not a great place to think about training an ML or deep learning system. The system where we had a lot of, unfortunately, we had a lot of data was our incidents and service requests within IT, so we started there, and how could we learn about that, and how could we then apply these techniques to help our customers.
Mark Grimse: So we took a solution, we found a startup that was able to help us solve one of our problems. We didn’t know it was a problem, but we realized that when a user comes to us and gives us a service request or an incident, in the ITSM or ITIL type language, we need to bifurcate that and say, “Yes, it’s a problem, an incident,” or “No, it’s not a problem, it’s just a service request.” “I need a mouse. I never had a mouse, I need a mouse,” or “I need a new distribution list.” And so, by using this technique, we had an SLA that my service desk would look at it, and within 15 minutes they had to automatically put them into one of those buckets, so that then the queue management would take over, and it would get kicked off to one of those groups.
Mark Grimse: Well, by using these techniques, we’re able to do that 85% of the time through ML and AI stuff, and I don’t have to spend my service desk time doing that. By taking that automation off their plates, they’re now able to increase the number of incidents that they’re able to close. And so we’re just going to chip away at this problem through more ML/AI until I can reduce the amount of effort I have to service my customers through the basic levels, and then my service desk level one, level two people can move up to higher-level problems.
Tim Crawford: Got it. So you’re leveraging the technology, and there are different ways, mind you, that I hear folks leveraging ML and AI specifically. And I do want to get to RPA, but it seems like you’re focused on specific functions and trying to automate those as a means to increase the time to value, take the error rate out, increase the level of service. Is that resonating with… Sounds like it’s along the same lines as what you’re focused on.
Mark Grimse: Yes, absolutely. And I think we’re doing this that direction because I think we’re… It’s still a new technology for Rambus, and it’s a new technology in our space. Other people may be doing it in different ways. I could see… I see places in which I would like to try it in engineering, and it’s going to take some of the CIO change management activity to work with them, to have them see the light, and then to apply some of their brainpower and some of their time to see if they can make the strides in error reduction, increased in productivity, and those types of things.
Tim Crawford: But you mentioned, Mark, where you see RPA is different than ML and AI. Can you maybe briefly talk about how you differentiate between them?
Mark Grimse: Sure. I think about, and maybe wrongly, so I’ll put that out there, is that I’m never, we’re never perfect, and we’re always interested in learning more. I think about RPA as more of a automation capability. There may be RPA solutions that have ML/AI behind them, but the core part of RPA is all around automation, and it’s about automating things or systems that don’t necessarily have what I would call systemic APIs, APIs that are constantly in use. Right? When we think about communicating between, take the Uber example, your Uber client on your phone and the server has a very distinct API that goes back and forth. However, RPA is, how do you take a look at and automating those things that don’t have that perfect level of API, and that you need some other intelligence to help bridge the systems, and that’s where I think RPA can come into play. And I think one of the big things that RPA needs as part of their ROI analysis is what’s the error reduction, and what’s the transaction volume? If you don’t have kind of a couple of different parameters, then RPA isn’t suited for everything.
Tim Crawford: And that’s probably true for any technology, including ML and AI. Like you mentioned, looking at the volume of orders, you don’t have the sheer amount of data that would really serve a learning model well.
Mark Grimse: Right. But I think most ML/AI applications today have probably a very systemic and deep-level API that needs to be exercised.
Tim Crawford: Sure. So as we kind of wrap on this episode, and I have to say I love the way you’re taking a very pragmatic approach to both the conversation, but then also each of these challenges that we all face as CIOs, as IT leaders. You’re taking a very regimented and pragmatic approach to it. I think that’s commendable, because a lot of times, you know, there’s a lot of aspiration that comes into the conversation, but you got to take care of the basics too, and I think that’s very refreshing. So, let me close with this question. As you kind of think about where things are going, and you think about Mark Grimse, and you think about the role of the CIO, what excites you most about the role of the CIO both today, but then also where it’s headed?
Mark Grimse: So, I think there’s two different parts of the role that we have to think about, and that is the business partner role. And to me, it’s never been stronger, and it’s never been more needed to have a CIO that can do that close business partnership role, because every organization needs technology. Marketing needs it more than ever, you know? HR and finance were the first, but I think every part of the organization exudes a need for technology to help, so you need to be the best business partner you can be. And striving to do that excites me because that’s an area in which I need to continue to grow my muscles.
Mark Grimse: But the other part is your technology tool belt has never been richer, and never been able to deliver value as quick as it is today. And I think that’s what excites me, is that I could deliver systems 20 years ago, and people would be excited and they would do great things, but it would take me forever to create those systems. And now, using off-the-shelf components, using the cloud, using different techniques, APIs, all of these, you know, the agile mentality, allows you to come to value so much quicker that you can then bring value to the business in a much easier way. And then you’re no longer a partner that says, “Yeah, I can fix that. I’ll come back in six years and you’ll have an order management system.” By being a great business partner, you’re going to generate demand, and you need a way to fulfill that demand, and now we have kind of the agility and the tools to be able to do that.
Tim Crawford: That’s great. Mark, thanks so much for being part of the program today and sharing your thoughts on both the role of the CIO, but also what you’re up to at Rambus.
Mark Grimse: Thank you, Tim, appreciate the time. It’s always good to spend time with you.
Tim Crawford: Thanks, Mark.
Tim Crawford: For more information on the CIO In The Know podcast series, visit us online at CIOitk.com, or you can find us on iTunes, Google Play, Stitcher, Spotify, and SoundCloud. Don’t forget to subscribe, and thank you for listening.