The State of Enterprise Technology
A fireside chat with startup investor and Fortune 50 Board Member Gary Reiner, the former CIO of GE.
If there is such a thing as a “Rockstar CIO,” Gary Reiner is surely it. After a stint as a star consultant at BCG, the long-time GE CIO was Jack Welch’s right hand man in designing and implementing the “Six Sigma” system that transformed GE and revolutionized American manufacturing. Now an investment advisor at General Atlantic, Reiner sits on the boards of Hewlett-Packard, Citigroup, Box, and other firms.
I recently had a chance to spend an hour with Gary to discuss the future of enterprise technology before a room full of C-level enterprise customers, service providers, and enabling technology players at Actifio’s eCloud Summit in Austin, Texas. Here’s what he had to say…
♪ [music] ♪
Mike Troiano: Welcome, Gary. Thanks for coming to Austin.
Gary Reiner: Delighted to be here.
MT: Very glad to have you. So what’s the big deal with the cloud, Gary? Why is it worth getting all these people to Austin, Texas, in February to talk about the cloud.
GR: One, to get out of Boston… but two, there’s a big change going on. From a customer’s point of view, from an enterprise point of view, by far the most important thing they care about is the cost of running infrastructure and the responsiveness and the flexibility with which you can get capacity. Really the migration that’s going on is the migration from hardware to software. Data centers are basically in the midst of migrating from hardware to software as everyone knows and VMware really took the first step at that by virtualizing servers and allowing servers to go from sort of a 20% capacity utilization to a 95% capacity utilization. I think what software running a cloud does is it really allows you to migrate from expensive hardware to the world’s cheapest hardware and thereby allowing you to have much more scalable infrastructure at a much lower cost. So right now, as you all know having software run a data center basically allows you to have the world’s cheapest hardware basically processors and the software allocates the work to whatever processor is running and if the processor doesn’t run because it’s broken then the software just reallocates it to somewhere else and so it allows cloud companies, service providers, to basically have much cheaper hardware than they otherwise did and I can tell you sort of from an HP perspective what it looks like which is it’s tough. I mean because when we were selling servers to enterprises it was a good business. I mean it was a 30, 40% gross margin business. When you’re trying to sell servers to a cloud guy who doesn’t care about reliability because the software can reallocate the work, doesn’t care about all the other things that are put into a high-end server, the margins go way down and so from a supplier’s point of view it’s very different but from a customer’s point of view it is an all win. It is a terrific thing.
MT: How about the business impact beyond the technology? Right? The economics, the utilization rates on the hardware and software itself obviously dramatic efficiencies there but what have you seen in terms of the impact on the business itself? In terms of resiliency, agility, business process change, those kinds of ideas.
GR: Sure. Sure, it’s a great question, Mike and the obvious one that we all know is that if you’re an entrepreneur and you’re starting up a company it is dramatically easier to start on the cloud than building your own data center. Most new companies in the Valley have no data centers at all. Everything is on the cloud. Most of them start with Amazon, or Azure, or now IBM, HP and others. So the B2C side of building companies is all of the growth that we’ve seen in all these companies, Uber, all these other companies, it all from a cloud infrastructure. From an Enterprise point of view I think what we won’t know is how much of the success of the Salesforces is in the Workdays of the world, and other SaaS providers are driven by the fact that they have great software? And how much of it is driven by the fact that they’ve got very low infrastructure? So, but you know, there are a lot of great SaaS companies that are growing and they’re growing at a much more efficient infrastructure than they otherwise would had they had traditional data centers.
MT: It’s interesting, you know. We were talking about this at breakfast that it’s almost like the Facebooks and the Googles of the world have defined an architecture that’s highly scalable, highly efficient and Enterprise is kind of following them in a way, the ground that they’ve broken.
GR: And there’s one fundamental difference which is that the cloud infrastructure for companies that basically run one application are going to be a little different than the cloud companies that are providing infrastructure for lots of different customers.
GR: So a Google infrastructure is going to look…and a Salesforce infrastructure which is basically running one very big application is going to be different than an Amazon, Azure and others that are running many, many applications yet still controlled very much by software.
GR: Yeah, yeah.
MT: So if you’re one of the guys tasked with providing access to all those applications you’re running, maybe you’re the CIO of a company, a big complex, global organization…
MT: …you get it, just hypothetically.
MT: Where do I start, Gary? What do I do first?
GR: Okay. The first thing you want to figure out is where do you want to go SaaS. You had a great chart up there that had there’s SaaS. There’s private cloud, the public cloud and hybrid. So the first thing to do is, “Okay, what software am I dissatisfied with internally? That I want to go to SaaS,” and a lot of the reason why SaaS, I should say this, a lot of the reason why SaaS is better than On-Prem is it’s newer. Okay. So you know one of the reasons why Salesforce is successful is that it’s on a cloud infrastructure. The other reason it’s just newer CRM software than what it replaced and so there. So I would first thing you do is you look at, “What are the software solutions that I can start up right now? I can rent a few seats, pilot it and make the migration.” The next bit is I would look at the applications that are most easily moved to the cloud. Okay and they tend to be applications that are more horizontal more common and then the toughest stuff to move and the stuff where you’re going to have the biggest economic constraints is in stuff that you’ve written yourself okay that you wrote for On-Prem and it’s going to be really expensive and hard to justify the economics of migrating them to a cloud infrastructure. I mean it would be cheaper but the upfront cost in rewriting it would be quite significant.
MT: So the data center doesn’t go away. It’s like paper.
GR: You know, I’d still say… I mean, I don’t know what… I haven’t seen the stats but I’d guess 90% of a large Enterprise’s infrastructure right now is still on traditional data center.
GR: Okay and then you’ve probably got 6% that’s private cloud and only 4% that’s public cloud. I mean I’d bet if you kind of did a survey it would be something like that. I’ve done it.
MT: Right, right.
GR: I’ve done kind of informal surveys with my friends and that’s kind of where it’s at you know. So they see the excitement. They see the growth but it’s off of a very small base so far.
MT: All right. So you kind of figure out, “Here’s a portfolio of applications. Here are my priorities. This is the continuum of architecture options.” What are the three biggest challenges you’re going to face in making that change?
GR: Security, security, and security. Okay. There is this belief by many people that moving to the cloud opens you up to security concerns that you don’t have internally and as just one observer of the situation, it is exactly the opposite. Okay. I’ve often said that a CEO should bring in a CIO and a CSO of their company and then bring in a Salesforce, a Workday, a Box any of the SaaS application providers and have a little competition as to who’s been broken into more. Okay and they would turn to the SaaS company and they would say, “Well, how many times you been broken into?” And the answer is in most cases, “Zero.” Zero and then if they were honest, they’d turn to the CIO and CSO, I admit I was one of them. “How many times have you been broken into?” And the right answer is “How many times a day are you broken into?” Because let’s face it, enterprises, large enterprises are broken into all the time. We’re easy targets. We’ve got a lot of uneducated employees that hit the wrong link and allow bad guys to get in. You know when I was doing this I spent every quarter reporting out to the audit committee of the board on the latest break-ins. It’s only gotten worse. It hasn’t gotten better.
GR: Yeah. So I think what we as a function need to do a better job of communicating is that SaaS is actually more secure, not less.
MT: That sort of “Come to Jesus” meeting in the office of the CIO.
MT: There’s a dynamic at play when we’re talking about an infrastructure change as strategic as moving towards the cloud. You know lots of people have received a note from their CIO saying, “Got to get me more of that cloud.” Talk a little bit about the internal dynamics.
GR: “I want to be able to say I’m on the cloud.” Right?
MT: Yeah, yeah. You know there’s a dynamic between kind of the bridge and the engine room to use a Star Trek analogy.
MT: …in this movement and it’s politically charged and there’s lots of stakeholders. Talk a little bit about the real world challenges of struggling with this transition inside a big company.
GR: So the biggest challenge that I’ve seen is that you’re trying to convince your customers to outsource something that had been in-sourced and in the old days when we had big data centers that we moved out to IBM, and CSC, and EDS they would basically buy the people. I mean they would just take all the people. They would re-badge them and everybody would be just fine. I mean you would convert your costs from capital costs to operating costs, maybe get a little benefit, certainly in the first three years, if not the last seven and for those that have been there know that and everything would be fine. I think what is true about the cloud environment is it is less people-intensive than a traditional data center and so that when you go from managing your own data center or managing particular functions of a data center, to outsourcing it, you are going to be telling a lot of people that you’ve known for a long time that they’re out of a job and so I think that’s the toughest social issue with moving to the cloud.
GR: Yeah, and it’s not easy. There’s no easy answers there.
GR: You know hopefully the best of the lot will get hired by the cloud provider just because they’re talented people but you don’t have the same kind of engagement between out-sourcer and in-sourcer that we used to have.
GR: So it’s a big change.
MT: The people stuff. It’s all..
GR: That people stuff. I hate that.
MT: …people stuff.
MT: So you know a lot of vendors in the room, service providers, technology providers. What’s the right way for us to engage with that process inside a big Enterprise account?
MT: Is it important to you to find your sponsor inside the technology organization? And then look to get folks on board. Do you need to sell down? What’s your advice to people in the room on that question?
GR: So there’s two ways to go. One, you need to get to the CIO because you need to get to somebody above the whole social issues. Okay. So that’s one way to go. The other way to go is to find somebody that’s quite entrepreneurial, lower in the organization, do a proof case, okay. Prove that you can actually save money with higher quality and more responsiveness and then the two of you go to the CIO and show them what you’ve done. Okay but it’s one or the other. You’re going to find a lot of resistance in middle management. A lot of resistance. Yeah.
MT: And what’s the determinate of which of those strategies you should pursue?
GR: Oh, I think you should try the former. Sorry. You should try the one where you find the entrepreneurial guy in the function.
GR: Just so you can develop a proof case that… Because it’s much better to go to a CIO, or anybody, any executive having demonstrated in their organization real productivity.
GR: Okay. The way we used to do it at GE was we would pick a business, you know. If we were blessed and cursed with a lot of different businesses, okay.
GR: And so we would say, “Why don’t you try it in lighting? And if it works in lighting then we’re going to roll it out to the rest of the company.”
GR: Okay. That was a very common strategy of us and it wouldn’t be like we’re picking lighting as much as lighting would volunteer.
GR: And we would say, “There’s our laboratory and we’re going to take that. We’re going to use it and we’re going to see what comes of it and if it works then we’ll standardize on it.”
MT: Got it.
GR: Yeah, yeah.
MT: So you’re bringing new tools to bear, new technologies, you know. How should Enterprise customers think about the role of the various forms of service provider? Cloud service providers, MSPs, that type of company.
GR: So I think the obvious answers are price and flexibility but there’s another one, which I think is out there, which is switching costs and as enterprises are looking at service providers the question is, “Okay. Give me a price. Give me your up time, your SLAs in terms of up time but also in terms of responsiveness, in terms of availability, of capacity and scalability of capacity but also I want to know if things don’t work out how easy is it for me to get out?” And then I think the fourth thing which is really, really important is, “How do you differentiate yourself?” Okay. What’s going to be different? because, as you well know, you’re competing against some very big boys out there. You know with Amazon, and Google, and Azure, and IBM, and HP. I mean there’s some guys with a lot of capacity and a lot of cash to put behind this opportunity and in some cases I won’t you know some of those guys are perfectly willing to lose money as a way to gain share, as we know and so it’s tough and so the question really is, “How do you differentiate yourself from them?” And so we could brainstorm about ways to differentiate yourself. One of them clearly is security and I’ve run into a couple of companies that define themselves by being extraordinarily capable from a security point of view. There are others that offer co-location because it’s for high-speed trading and the closer you are every millisecond matters but there’s got to be some way to differentiate yourself maybe by using Actifio, maybe that gives you the lower enough cost that…
MT: Somebody tweet that. Can you get a tweet?
GR: …that you can be lower cost and therefore differentiate yourself that way but it’s otherwise potentially a pretty tough business with low margins unless you can figure out a way to say, “We will do something that no one else can do.”
MT: You know it is challenging, the service provider business.
MT: You’re kind of capped at the low end by guys that are happy with retailer margins and you know climbing the food chain is a capital-intensive process. I was reading in Gigaom that there are 5000 service providers in North America and 10,000 around the world. Ten years from now what’s the over-under? Is there going to be a winnowing in that business? And if so what will define the winners?
GR: So my answer and who knows, right? but my answer is, “Yes, there will be a big winnowing.” There will be benefits to scale and so the larger you are, the more economic you are and you’re going to get into a virtuous cycle as a large player and getting more and more scale and it’s going to be harder and harder and with network speeds getting faster, and faster, and faster where you are matters less and less, and less. So it’s really who’s going to have the biggest scale, the lowest power and the lowest heating and cooling that’s going to matter more than anything else and I was only half kidding, Mike, over breakfast that the right place, I beg you guys to consider this, the right place to think about a data center is Iceland. Okay? Iceland has the lowest power on earth. You want to cool the place you open the door. Okay. Land is basically free and the reason power is so cheap is because both hydroelectric and geothermal power are virtually free there.
MT: Got it.
GR: And so, but anyway, it’s just a… But we’ve got to figure out ways to differentiate yourself. That’s the key thing you know and you’ve got to be able to say it in an elevator speech. I mean, that would be my advice.
MT: eCloud Summit 2016, Reykjavik. All right.
MT: So I’m going to put you on the spot a little bit here.
GR: All right.
MT: So you’re on the boards of HP and Citi as well as a bunch of smaller technology scrappy companies. You really have a kind of a unique perspective. The old guard of IT, the hegemony of the last 30 years, are those guys screwed or not?
GR: It’s tough. Every one of them has a big, strategic, existential challenge. They’re different. Okay. So we can take them one at a time you know. EMC is challenged by the world going to software defined and making hardware much less important as we talked about before. Cisco same, same exact thing and you’ve got new players out there that start from a software-only concept where the hardware’s much cheaper and it becomes much cheaper, more capable functionality and they can either copy them and give credibility to them or fight against them and so it’s tough. It’s tough either way. Oracle’s got a different kind of challenge and by the way I’m sure everybody will figure something out here but I actually think that NoSQL is going to be a major game-changer in the database business. Not only is it much, much lower cost but it takes about a third of the time to develop applications on it a NoSQL database than on a SQL database and the tough thing that I think Oracle faces but I’m sure they’ll figure something out, is that even if they want to roll out and they do have a NoSQL database it’s at a much lower margin base than the SQL. So I think from that perspective it’s going to be a tough one to get through that. HP as you know is splitting up and you know they’re going to have the same challenges of moving toward software-defined everything. HP is rolling out a cloud strategy based on OpenStack and we’ll see how that goes and then there’s IBM and IBM you know has had an interesting strategy over the past 12, 13 years of buying up a lot of software companies, mid-level software companies and also on these long-term service agreements and we would go much deeper on that.
GR: But how they continue to grow with that strategy is going to be tough and they’re finding it very tough to grow right now and it’s reflected in their stock price. So they’re all challenged in different ways. There are no easy answers.
GR: No easy answers.
MT: I mean there’s smart people at all those companies.
MT: And they’ve all read “The Innovator’s Dillema, “
MT: They get it.
MT: Sitting on a public company boards, that dynamic of either moving away from higher-margin hardware or weaning yourself from capex purchases towards subscription models.
MT: Is there a sense at the top of companies like that that, “At some point we’ve got to bite the bullet?”
MT: Is there an appetite for it?
GR: The answer is, “I think so.”
GR: Okay. Some are easier than others.
GR: I think it’s, you know, I think EMC can get its head around a more software-oriented storage approach. You know one person asked me a long time ago, “If HP had owned VMware, rather than EMC had owned VMware would it have been as success?” Who knows? Because then there’s truly an innovator’s dilemma there, right?
GR: Because you’re buying much fewer servers if you have VMware but look I won’t at all deny, nor would you believe me if I did, that the quarterly pressures are all-consuming okay in all these public companies.
GR: And so they’re doing everything they can to keep their current product lines successful. Okay but they also know about innovator’s dilemma.
GR: And so, they’re going to be trying to do both. They’re going to try to keep their current line as strong as possible. I mean Oracle is going to keep trying to sell SQL databases and there’s going to be a big market for it but they’re going to have to play in NoSQL as well. It’s just too economic from a customer’s point of view to ignore NoSQL and it’s going to get more and more so.
MT: Should EMC divest from VMware at this point? And will they?
GR: They’re very different businesses. Sold often to different customers. They don’t necessarily conflict. There’s only one small place where they conflict. So I don’t think it’s a big negative that they’re together. I don’t think there’s many positives that they’re together either.
GR: You know. So I don’t actually think it’d be a big event one way or the other if they were together or separate would be my answer.
GR: You know. Just like I mean I’ll tell you. Look, HP’s gone through it and I can give you the logic for it, which is quite public, which is it’s just easier for senior leadership to focus if you’re, you know… And so if you are more focused, you do have benefits and so that’s why HP split up.
MT: So put your customer hat back on.
MT: I’m a brand guy and you’re talking about some pretty venerable brands.
MT: In the list of companies that are struggling with some of these challenges, you know. Famously no one ever got fired for hiring IBM. As a customer, how do you navigate the shifting landscape? How do you figure out what to put with these more trusted old guard folks? And what to maybe look for new partners on?
GR: So here’s what a lot of companies do. Whether they should or not, here’s what they do. Is anything that’s going to be long term, anything where they’re actually handing off with relatively high switching costs, which I think is the key dimension, they want to make sure the balance sheet of that supplier is strong? Okay. So if you’re going to be outsourcing any with any kind of long-term relationship or even if you’re going to a technology where once you go, you’re there, one of the things that a customer really cares about is, “Do I need to worry about your financial liability?” And so that becomes really, really important. Beyond that you know what I see and everybody else sees as well is that most large companies are experimenting with a lot of the Silicon Valley technology companies and 128, Route 128 technology companies as well but they’re trying it in small pockets.
GR: And I mean it’s no different than my GE Lighting example and if and when it’s successful then it becomes a company-wide rollout assuming that the balance sheet is strong.
GR: Yeah, but pilots are a very big deal. I mean you know and taking your time with them is a really big deal as well so that’s what we did a lot of and I find a lot of… That’s what Citigroup does and that’s what HP does. Everybody does that. Yeah.
MT: You’re plugged into that ecosystem as well as an investor and I know you provide strategic counsel to more of the up and coming guys. When you look out at the landscape of the next generation, who do you see that you’re excited about?
GR: Of the next generation of companies. Boy, there’s a lot but everybody knows them. I mean I’m excited about Box, you know. I’m biased. I’m on the board of Box but there are a lot of great…
MT: Why? What’s there? What’s magic about Box?
GR: So there’s a real need for better collaboration amongst big companies okay. There’s just there’s too many silos of information too many ways in which information gets not connected to the right people in your organization, one. Two, obviously you know mobile is taking over…
GR: …everything, okay and so from a personal productivity point of view as well as from a company productivity point of view, when you have a document you want access to it whenever you are, wherever you are okay and so taking a document and putting it in there and saying, “Mike has access to it but no one else has access to it if Mike has access to it but he can only read it. He can’t edit it.” Okay. “And everybody else can edit it, but Mike can’t.”
GR: Or, “Mike can edit it but no one else can.” I mean and all of those functionalities and the workflow around it is something that every company needs. A GE we built our own because there was nothing out there in the early 2000s and it was getting 30 million hits a day. It became such a core part of the company.
GR: GE as you has gone to Box and I envision that that’s what a lot of Enterprise companies are going to be doing is moving to something like that, yeah.
MT: You know as a consumer one of the things that pisses me off is that I can’t edit my Microsoft apps and put them into my Google Drive. So we’re a Google shop and we really use Google Drive in exactly the way that you described.
MT: And now I have a Dropbox account I have a Google Drive account but there’s Windows Live.
MT: You know, Box is unique.
GR: One drive is one drive.
GR: Yeah, yeah.
MT: In some sense that it’s sort of disconnected from the applications and is the…
GR: Stay tuned.
MT: What’s that?
GR: Stay tuned.
MT: Ah, got it. Got it. You heard it here.
GR: But you’re right. Right now it is.
MT: Yeah, got it okay. All right. Tricky. We’ll leave that right where it is. So let’s talk beyond the sales process.
MT: You know at the end of the day I think with the flow of information you know we’re very focused on selling like every company but we’re very focused on making our customer successful. At the end of the day, you’ve got to show them the money. That’s the franchise. That’s the only way to get new customers today you know. I think everyone in the room is really in touch with that, that it goes beyond sales and you’ve got to focus on customer success.
MT: How do you know the technology players and the service providers in the room, how do we engage with our customers post-sale to deliver on our promise of business impact? What’s the right way for us to think about that as vendors and partners? And what’s the dynamic inside the company as we try to engage?
GR: So you know, look, there are a lot of dashboards that CIOs, CTOs in particular have that look at cost and capability up time and everything else and I think getting yourselves put into those dashboards and demonstrating that you are meeting and beating the metrics that the CTO or CIO is evaluating you on is the key. So demonstrating that you’re continually taking costs out not just one-time cost out but continually taking cost out, demonstrating that your up time is superior to what they had in anywhere else. I mean just basically beating, figuring out what the metrics are that matter most to the CTO and beating them but getting in the dashboard. Being front and center there is I think key.
MT: Well, I want to protect some time for questions from the audience…I’m going to close by going through each of the core constituencies in the room…
MT: Let’s start with Enterprise Customers. You’re here…. You’re wrestling with these issues. What’s the thought you want to leave those folks with? What’s your best piece of advice for them?
GR: Don’t go big first. Try it out. Try it on you know. I would start with SaaS for an Enterprise customer. I’d get comfortable with the idea that you’re not managing the infrastructure and I think the big debate for an Enterprise customer is private cloud, public cloud and it depends on the organizational comfort with private cloud versus public cloud. There are a lot of companies that just for whatever reason, regulation whatever, just are scared of putting their data in someone else’s hands and so they’ll build a private cloud. They’ll get most of the cost benefits of that by putting in a private cloud because you’ll still have a software-defined data center but you won’t have to face the organizational hysteria around outsourcing and in some cases legitimate regulatory but only in some cases legitimate regulatory concerns about that. All else equal going down the road I’d prefer a public cloud versus private cloud. Better you save your capital for more strategic uses in the business and make this an operating expense rather than a capital expense would be my advice.
MT: Good stuff. Same question for service providers in the room, C-level execs, owners of both. We have representatives of I think three of the world’s top ten largest.
MT: And a bunch of regional and vertical guys as well. What’s your advice to them?
GR: So this is a very capital-intensive as you all know probably a tough ROI business right now because it is so capital-intensive and so competitive at the same time and so I would, I mean, again I’m being totally redundant and I apologize for that but you’ve got to figure out how to differentiate yourself either in some way that you get a higher price because you’re offering something that no one else can like security or you know you’re nearer them in an era where milliseconds matter and so you can get a higher price that way or you know you’ve got to use Actifio and get lower cost.
MT: Good segue. For the enabling tech players. A lot of Actifians in the room but representatives of a lot of the other virtualization players VMware, Arista, Citrix, you know.
MT: Advice to us as enabling technology companies, both on the service-provider side and on the Enterprise side.
GR: Well, I think one thing that strikes me is the synergies between the two. I mean the better you do in enterprises the better the service providers are going to feel about you.
GR: And the better the service providers do with you the better the enterprise is going to feel about you and so you’ve got to keep feeding both.
MT: It’s interesting. We run these executive briefing center visits and I think one of the turning points that will go down in company history is we had the CIO of a big bank.
MT: And we were in this sort of after-meeting. We were talking about our business and we said, “You know about half our business is service providers and about half is enterprise customers,” and he said, “Guys, I’m a service provider. I do not want to be the VP of the data center. That’s not a great career option.” Right?
MT: So I’m providing services to my team and I think that conversation, the penny really dropped that there was convergence.
GR: Yeah, yeah.
MT: That’s even on the enterprise side, we’re all service providers now.
GR: That’s true. That’s true but from a customer’s point of view if you’re an Enterprise customer there’s credibility that the service providers give you by you being big service providers.
GR: And vice versa.
GR: A service provider is going to feel more confident knowing that you’ve succeeded with Enterprise as well. So I think feeding both of them, don’t focus on one versus the other keep feeding them both and that would be advice for Arista. It would be advice for anybody, yeah.
MT: There you go. Thanks, Gary. Now let’s get some questions from the audience…
Check out the video of this interview and the entire eCloud Summit conference at Actifio.com.