Managing Your Cloud: Expert Advice

In 2018, cloud computing is fully established technology, with many companies managing robust cloud deployments. Yet cloud’s new-found maturity is not without its problems. Many companies struggle with a long list of challenges as they oversee their growing cloud deployments.

To address these concerns, I spoke with two top experts in cloud computing:

Michael Liebow, Global Managing Director, Accenture Cloud

Bernard Golden, Vice President of Cloud Strategy, Capital One

Among the topics we covered in a wide-ranging 25-minute conversation:

1) What are the pain points you see with cloud users? What advice would you give to address these challenges?

2) How has the cost factor for cloud worked out? How do you think companies are viewing/handling their monthly cloud bills?

3) How can companies prepare for the cloud of the future? Say 2-5 years out — what can they do now to stay ahead of the curve?

INTERVIEW TRANSCRIPT: Managing Your Cloud

Addressing Cloud  Challenges 

J. Maguire:
Cloud computing in 2018, obviously it’s no longer emerging technology–its core, critical, essential technology for enterprise-IT. But, I think a lot of businesses are floundering. They don’t know exactly what they’re doing, there’s pain points out there.

Michael, what would you say if, among your cloud customers, what would be a key pain points and, most importantly, how would you solve it–which might be a five-hour question, but I’ll ask it anyway. How would you fix that pain point that that company is having? 

M. Liebow:
That’s a loaded question. In any transformation, which is what this is about, if it was just about technology it would be easy, I think. But, there are three major challenges. First, it has to do with the technology: how to use it, who to use it from, do you have the right sets of partnerships and do you have a vision for where you think this market is going, and how it’s going to benefit your business? That’s number one.

Number two is, commercially, people are struggling with how to acquire this technology, particularly certain industries where they’re somewhat challenged, policy wise, or with regulation or laws. That encumbers that drive. Then the third thing that comes into play, which is probably the largest, is culture. You have a lot of people who have been happy doing things the way they’ve been doing them, and now you’re asking them to them very differently. Also, their roles are evolving faster than, perhaps, they acknowledged.

So, those three things, between the technology, commercials and the culture, tend to slow the evolution of cloud in an organization. Particularly the kinds of organizations that we sell to and partner with. 

J. Maguire:
Is the culture a piece of that? Is it about of losing control? We used to have this in house and we had it under control, and now it’s going partially out of house, and we don’t like that because it lessens our profile? What is the culture problem there? 

M. Liebow:
There’s a major convergence going on, between the infrastructure and the application. People give lip service to the notion of DevOps, so it’s combining those functions, so that devs have to operate). That’s a cultural change. Organizations who have traditionally kept those siloed, can’t quite deal with the notions of continuous development, continuous deployment, continuous integration. These are new, foreign concepts, and some people call them bi-mobile. You know, how you treat legacy and how you treat cloud.

The biggest red flag there, is if people think of it as just another data center. As soon as you do that, you don’t have cloud. All the attributes of cloud go out the window. If you have approvals work flow, and you have a committee that reviews all requests, you don’t have cloud. There are a bunch of indicators for an organization, where they really aren’t evolving. They aren’t committing or going all in on cloud. 

J. Maguire:
I love those committees that have to approve everything. That really just heightens efficiency, to the utmost. 

M. Liebow:
Yeah, before the application has to scale, it has to request a service ticket.

J. Maguire:
Bernard, I know you’ve talked with plenty of cloud users–if anyone has their finger on the pulse of cloud it’s certainly you. What would you say in terms of pain point out there? And how are you going to magically solve that pain point for this company? 

B. Golden:
I would agree with what’s been put forward here. And, I would say that there’s a couple of things. One is, the inertia, the culture, whatever you want to call it, of the organization. Let’s just be frank about it, shifting to cloud is a change in the relationships and the power relationships within IT groups. Traditionally, the infrastructure group had huge amounts of clout because that’s where most of the spend was.

Now, all of a sudden, you’re going to that group and saying, “All that clout you used to have, that’s going to be lessened because we’re going to start spending it. So, you’re not going to have these huge capital budgets, etcetera etcetera.” That plays out in people going, “Wait a second. I don’t want this change, I want it the way it was.” Just to give you an example, this was so stuck in my mind. We were invited to a kind of technology retreat with a customer, and we were outlining all the magic, whizz-bangy things you could do with cloud, and this guy puts his hand up and says, “You know, my job is installing web sphere into virtual machines. That’s what I do. Eight hours a day,” 

J. Maguire:
Cool job.

B. Golden:
“And I plan to do that for the rest of my career.” 

J. Maguire:
Oh wow, okay.

B. Golden:
“Thank you very much,” So that’s the kind of thing you address culturally, or whatever you want to call it–organizationally, politically. A step beyond that is, there are people who are like, “Oh, I’m going to use cloud but it’s going to be a data center at the end of a wire,” and they just say, “I’m going to use the same practice, the same process, the same tooling, and all that,” And that’s, kind of like buying a Ferrari and then saying, “But my biggest goal is to haul groceries around in it, the way I’ve done with my station wagon.” I mean, can you do it? I guess.

And then, let’s just say you make those kinds of changes. You then have to confront the stuff that I think Michael is talking about, the process, architecture, tooling, practices that go along with the ability that you can have when you accelerate infrastructure availability, when you can use all the services and portfolio as tinker toys to construct applications. That’s a very different world than it was in the old days, when it was like, “it’ll take me a year to get a server so that’s my cadence.” And you confront those things. 

J. Maguire:
What is the answer, if there is one, to that question of, “It’s just going to be a datacenter at the end of a string”, how do you combat that mindset? 

B. Golden:
Well, I’m just going to be honest with you, I think it has to come as a mandate from above that says, “We’re going to go do this.” You have to have the kind of people who aren’t like the person who put his hand up and said, “I’m just going to install web sphere for the rest of my career.” You have to have people who are like, “I’m ready to embrace this.” Without those two things, you’re going to end up with a, “We’d love to be able to do something new and exciting and different, but we just can’t seem to get there.”

The Costs of Cloud Computing

J. Maguire:
What about the costs question? I think early on, cloud was sold as a big cost savings, sometimes it has worked that way, sometimes it hasn’t worked that way, and it gets more about competitive advantage, scaleability, flexibility.

Michael, how do you feel that companies are feeling about their monthly cloud bill? Do they look at them and go, “Oh my God, why is it so big?” Or do they look at them and go, “Hey, we’re relieved, we don’t have to build out our enterprise data center as much.” What’s your sense of the feeling about cloud costs out there? 

M. Liebow:
There are different ends of that spectrum, that you really have to dissect, or unpack. So, start with the business case. Most people thought cloud was a cure-all for cost, so they did the business case and they couldn’t split the business case because the business case requires that you get rid of your legacy footprint. In order for cloud to be dynamic, you have to move away from cement, you have to move away from tin, you have to retrain and put them into new roles. You know, that person who was loading web sphere or that D.B.A., or the guy that was addressing incidents and running to trade out a defective blade. Those things change dramatically in the cloud.

The level of automation, the level of self-service or shared-service, the level of standardization changes the game–that’s cloud. That’s the definition of cloud. When you start to think through that, you build the business case. The business case can be dramatic. But the thing that really makes the business case flip is speed. You can’t drag this out for ten years. The faster you make this change, commit to it, the better the cost profit I think companies like Capital One have experienced this, and others, that the faster you evolve to the “new”, the better. That will get you on your journey. And, once you’re on your journey, you’ve shifted–and Bernard mentioned this–you’ve moved from capital expense to operating expense.

So, you do get that monthly bill, and you are in for bill shock if you don’t have the policies and the governance and the analytics around how you’re using cloud. Who has permission to do what? And instead of approvals management, you want to put something in place called Quota Management–that’s how you manage. The notion of cloud is that you were constrained before in terms of capacity, now you’re constrained in terms of budget.

You can end up with somebody lighting up a whole bunch of resources, and not knowing in the month what they did or how they did it. You have to know in real time, with this notion of active management, what people are doing so that you can enforce your policies and your governance around that and know where your costs are at any given day. 

J. Maguire:
Well there are all those divisions of the company that have their own credit card, or they can spin up their own instances on AWS and it’s really easy. Are they supposed to pay attention to the larger company goals? I don’t know. It was so easy for them to get started. 

M. Liebow:
It was. When I started, five years ago doing this, we had one thousand individual Amazon accounts, tied to an individual corporate card. So, I became the parent of those thousand accounts, and I had to put the governance around that. 

J. Maguire:
A thousand accounts in one corporation? 

M. Liebow:
Yes. 

J. Maguire:
Oh my goodness. Okay, there’s a story there, but alright. 

M. Liebow:
To give you a sense of it, our Amazon bill has grown significantly, so every month I process the bill. Today, from those initial thousand accounts, our bill is well over three hundred million lines a month. 

J. Maguire:
Jeez, okay. And you’re sitting there, processing that every month? 

M. Liebow:
Yes. Absolutely.

J Maguire:
Okay. Sounds like it’ll keep you busy. Bernard, your thoughts? Cloud and cost–I guess it didn’t turn out to be the magic bullet of cost savings, but there are enormous competitive savings there, so maybe we shouldn’t feel bad if the bill is larger than we thought. General thoughts about cloud and cost? 

B. Golden:
I think part of it is the expenses now fall into different groups than they used to. New people are seeing the numbers that didn’t used to see them. It used to be that somebody built a datacenter. Well that generally wasn’t a business unit, or some application group, that was facilities, right? And then you would just use that. And maybe you got a charge back, but that was probably at the divisional CIO or the CIO level–you never saw that.

Now, you’ve got an account for your application, and you get a bill, and you go, “Oh.” And then you get the next bill. I’ll give you an example of this. It was the CIO of a big media company, and he said, “We got started with AWS. We had a developer who was immediately able to be productive, it was magic. It was, like, four hundred bucks the first month. It was great.” He goes, “The next month, the bill was ten thousand dollars. Now, we’re a big company. We can afford ten thousand dollars. But you kind of want to know what happened.”

So, I think, in part, because the bills fall into different people now, and they see with immediacy what the cost is associated with, and they see the deltas, I think there’s a lot more awareness of that. I hold, I guess maybe somewhat controversially, I think cloud is a lot cheaper for what you do and what you get. But, you have to manage it judiciously and you have to know what you’re comparing it to. It used to be, some group said, “We buy laptops for developers. And there’s something out there that does the computing.” Well now that something out there comes with a bill, and they go, “Oh, it’s so expensive.” Well, it’s expensive compared to never having to pay anything in the past. It’s not expensive compared to what the overall company was paying. 

J. Maguire:
I guess the real comparison is it cheaper or more expensive than maintaining an in-house data center. And can your in-house data center even do what the cloud does anyways? 

B. Golden:
Yeah, the answer to that is no.

The Near-Term Future of Cloud Computing

J. Maguire:
And I agree. The other big question is the future. Companies are here, they have one or both feet in the cloud, they want to know how to get ready for what’s coming down the pike two, three, four years from now. Michael, what’s your idea? Your crystal ball–what is the future of cloud and how can companies get ready for that right now? 

M. Liebow:
A few years ago, Accenture decided to go all in on this journey. I am happy to report that we just surpassed 90% of our workload in the public cloud. Close to a forty-billion-dollar entity with four hundred and thirty-five thousand employees. For us to make that statement is essentially saying, “We eat our own dog food. You can’t sell this stuff unless you live in this stuff.” Most organizations aren’t anywhere close to that. We’ve built assets, skills, capabilities around that journey to help an organization learn from our mistakes–I’ll just say it that way.

I can tell you all the things not to do, but I could also help you on your journey to get there, and get there with speed, because that’s the value. The value is transforming your operation, so you are all in the cloud. And if you look at the level of innovation, that’s happening in public cloud, the level of capital investment, because cloud is capital intensive, it’s just not your capital. It’s somebody else’s. You’re leveraging all that innovation, all that security, all of those new capabilities. My platform is serverless. We’re leveraging all these technologies from a performance standpoint, from a cost standpoint, flexibility, and that’s where we’re finding the value.

There’s a clear line of sight for how you go from where you are today, to this new place. You can leverage that in your organization, you can converge applications and infrastructure, that’s where this is going, and you can manage–to Bernard’s point–you were never able to do show back at project level. Now you can do charge back at a project level. Not only can you do charge back, but you can give developers a line of sight to every tweak they make, what that looks like from a cost profile.

You can make people responsible, and it’s exciting. It’s exciting for talent to be able to now manage their application. And not just code and run, but actually code and run. So, they can live with that application, they can push it into production, and they can manage what that user experience is. And that’s what it’s all about. That’s what cloud enables. 

J. Maguire:
That makes perfect sense. Is there one tip or two tips that you would say, “Hey, if you’re going to get ready for a couple years from now, make sure you’re doing this and this?” Are there specific action items you would tell people? 

M. Liebow:
I think, talent is hard to come by, particularly in this area. You want to attract the right talent, and perhaps, find the right partner for this journey who has the right assets, capabilities, and skills, who can apply. The sooner you commit, the better off you are in terms of finding and leveraging that talent. Organizations like Capital One are very early into this in the financial services.

Well now, financial services is following suit. You’ve got banks and other financial institutions across the planet jumping in. You see it in insurance and health care. If you’re not leaning edge, or fast following, you’re going to get left behind. You have to commit to the journey, and you have to do that now. 

J. Maguire:
Bernard, if you’re not leading edge you’re going to get left behind, agree or disagree? That’s a pretty radical piece of advice. That might scare a few people out there, but I love how forward looking it is. Agree when it comes to what’s coming in your crystal ball in the cloud? 

B. Golden:
Michael, with your passion, you could be the preacher at some sort of royal wedding.

J. Maguire:
You could be the cloud Evangelist, or both of you could. 

B. Golden:
I’d like to return to the cost side. Michael had mentioned, briefly, the digital or innovation side, I think that’s what is important to think about if you’re coming to cloud. My assertion is, ten years from now, we won’t even talk about cloud, it’ll just be the way computing is done. And the reason is, cloud enables innovation. It’s what I call an innovation platform. 

All you have to do is look at companies like Spotify, AirBnB, Netflix is the poster child for that, who take the capabilities that they get out of this new platform, and say, “How can I redesign business offerings? How can I go after incumbents?” If you’re not someone who is taking advantage of those things, your challenge is that you’re competing with organizations that are.

What’s the old phrase? Bringing a knife to a gun fight, or something like that? I am sorry to be so militaristic in this, but you’re sort of handicapping yourself in the competitive frontier that will be the basis of being competitive in your industry, I think, if you don’t do that. And, if you continue to carry forward the attitude that’s like, “Well, cloud computing is just a data center at the end of a wire,” you’re going to be stuck in the old way of doing things. That is a competitive disadvantage from a business perspective, not from a technology perspective. You asked for a couple of tips? 

J. Maguire:
Yes, please.

B. Golden:
I think you have to pick out a use case that’s most amenable to the capabilities of cloud building. Typically, those are around digital marketing or digital initiatives, mobile. Those are a good place. And you have to go all in with that. You have to say, “How do I adopt what the leading companies in this space are doing, how do I emulate them, how do I adopt their practices.”

It’s going to take skills, let me go higher, you need some really capable people. You should be educating and bringing up to speed your internal staff. They have domain expertise in your industry. You need that as well as the cloud technology people. Then, pick out a project and go for it full board. That would be my suggestion and recommendation. 

J. Maguire:
That’s great. There is so much more we could speak about, but we said an enormous amount in a fairly short amount of time, and I think we’re good today. I appreciate the two of you sharing your expertise, thank you very much.