Why IT must drive the next wave of cloud innovation
Transcript
>> Welcome to our session here on why IT must drive the next wave of cloud innovation. I'm Nick Brackney from Dell Technologies, and with me today is Mark Bowker from ESG. And we'll get right into it. >> Sounds good. >> Yeah sounds good yeah. So one of the biggest things that we were tracking for 2020 last year was that technology has reshaped every industry and you know we've been seeing all these data points that show that the economy is increasingly going more and more digital.
And as such, data and software is going to play a massive role in delivering differentiation and competitive advantage for companies. But with the changing and the challenging business clamp of 2020 that we've seen so far, Mark is that getting even more pronounced and more important? >> Yeah I mean absolutely, there's no question. I think that it's exciting times if you're in IT, right? If you're inside an IT organization, inside IT ops, or if you're a developer, or maybe information security team, these are exciting times because things are changing.
A lot of times those folks spend so much time just kind of maintaining their current environment. But what I see is so exciting now, Nick, is due to kind of big changes in the market around cloud consumption, big changes around changing the security posture of the company, big changes about how applications are built, really makes an exciting times for IT organizations. >> For sure and you know we've been seeing is that, really it's coming down to running your apps in multiple clouds. >> True.
>> People have to really innovate anywhere. And repatriation has been one of those things that we keep hearing about. Now, the cloud providers do a great job. And these clouds, there's a lot of great technology and things you can tap into and benefit from, but we keep seeing repatriation because I don't think everyone always thinks critically about how they should be approaching cloud. And defining a multi-cloud strategy really is dependent upon having a central voice, somebody in IT taking the reigns and saying this is how we're going to go and do this.
So here are four considerations, outside of security and outside of other regulatory compliance, that kind of considerations, these are four things we think about the workloads themselves, that can really determine your success when moving to the cloud. So the first one is providing a better application and developer experience. That's really what you're first place. What you're trying to accomplish. And have you seen this, Mark, as somebody putting in a workload in the cloud and then realizing oh my developers don't know how to use the tools?
Haven't trained them up or now I've got all these applications left in my own data center that don't get as good of value as the ones that are in the cloud. What are you seeing there? >> Yeah I've seen those scenarios, I've seen worse. I've seen CAOs lose their jobs because they've gone off and taken something, and placed it in a cloud for all the right reasons to be honest, but then only to realize they opened up themselves to unnecessary risks. And that risk caused them to lose their job.
So it's crazy. At the end of the day, I mean if you're a developer and an IT person, you love these new tools and these new services, so you're anxious to use them and you want to be able to apply them to I'll say, current book print of IT investment, but it's a challenge because what risk does it open up? Can I take current investment, and are they kind of consistent across new tools that I may adopt? And it's a challenge. And also by the way, this doesn't all happen in the four walls of an IT data center, it's shadow IT, a lot of business owners.
>> Yeah and so when we talk about the next one here, getting the most value from data. As we saw in the last slide, data is becoming the key differentiator for organizations. And cloud it operates differently, and I think data's the one thing that sometimes will break cloud economics wide open or will create a lot of challenges in how it's being used. What are you seeing with data? >> I think the thing with data that I find fascinating is basically applying intelligence to it, right?
So if I can take data, I typically done things with data inside a data center to run queries, run applications, store and all that great things. But now, wait a second, what if I can take data that's sitting in a cloud, and being able to process things faster, get more real time information, be able to make more of a proactive type of decision? The other thing that's really interesting is this idea of data at the edge. Data's at the edge, how can I make quicker inference?
How can I basically take action on that data without having to ship it all the way back into the data center? So I think there's some fascinating things just to watch out how people are taking that intelligence applying across all three places. >> And then the next thing here is really around the network and delivering services anywhere. So what we've seen is we've seen an explosion at the edge. Data being generated at the edge that needs to be processed, that needs to be sent somewhere.
We've also seen that most organizations now are really spending a lot of time thinking about applications that run outside the land, that are in the cloud. How are organizations able to stay in control when so much of their application and data footprint is residing in areas that are really difficult to deploy technology solutions that they can manage? >> They aren't. That's the issue. I think that the innovation is there to take advantage of the capabilities that you guys talk to, Nick, but the question is is that's a challenge for them, is how can, there's certain information and data that I can certainly I'll say is less risk.
So if I'm capturing information from some manufacturing floor, that's talking kind of I'll say production of those machines or kind of just input from those machines, sure I can kind of move things across with a little bit less risk but if there's personal information out there, if there's company kind of IP out there, I can't be just shipping that across the network with just no standard good IT practices in place. That's what I see is people really need to take those IT practices they know and apply it outside of their data centers these other consumption models use.
>> Now the cloud, a lot of people mistake availability zones for data protection. Hey I've got it in two places, it's good, and then all a sudden (mumbles) gets replicated across at like a millisecond speed. So how should people be thinking about protecting their data across all these different heterogeneous environments? >> I think the biggest thing there is data protection isn't anything new, that's been around as long as data centers have been around. But in many cases, companies have mastered it inside their data center.
So they can understand what's being protected, how to recover that data, what should happen in the event of even I've seen like ransomware attacks. How do I actually (mumbles), how do I recover from that? So I think the initial thing is look at the investments you have and understand how those investments around business continuity or data protection itself, can really stretch and extend into these other operating models. So if I do consume from cloud provider X or cloud provider B, I can have that same consistency to be able to protect that data or in some cases it's don't put the data down in the first place, let's find a different model to be able to look at.
>> Right. So we've talked a little bit about this, multi-cloud. (talking over each other) >> In case you're wondering, it's real. >> Yeah it is but we have all these lovely buzzwords all the time and I like to think about it from a standpoint of just the reality that you're faced with. Multi-cloud isn't a strategy necessarily, it's not a posture, it just the reality of doing business right now, you really can't escape it. And so, most people thought hey I'm going to get into this, have a public cloud maybe, have a private cloud, hopefully they're connected.
But they realize that's not the case, they've got more than one. And I think the most common reasons we've laid out here why this might be happening, or why it is happening is it falls into two camps. One's strategic, either Google has good AI and Amazon's great at packaging maps, and you decided to be in both of them. It could be a risk mitigation, you worry that one of these cloud providers may become, at some point, competition for your line of business or maybe you're worried that I can't see around the corner, I don't know what's coming.
And one, you can't bet on any one of these guys without betting against another one. And two, even if you picked the right company who's going to be the winner, you don't know if they're going to prioritize your vertical. So that's really challenging. And then finally, and I'm sure you see a lot of this too is mergers and acquisitions and shadow IT. So talk to me a little bit about the pain that you see it from your engagements and your discussions with customers about how they're seeing this manifest in their business.
>> Yeah in some cases, they're just new, consistent I'll say, language. They almost all require their own decoder ring in some cases, so now if I'm running IT ops, I have to have an expert in the cloud A. Then I have to have another expert in cloud B. They may be doing the same thing and offering similar services, but the language is a little bit different, the service may be more advanced or less advanced than the other. And now that gets me very difficult from an organizational perspective from me.
Then I've got developers that have a lot of demand what they may be doing on one cloud, but there may be an advantage of another cloud that they should be using. So now I've got kind of the IT with the developers trying to work together, and I end up with organizational barriers inside of there. Then I back their information security as well. So it gets very difficult, just from an organizational perspective, nevermind some of the technology or architecture. >> You're seeing a lot of complexity because of management tools being different, infrastructure being different so you can't migrate workloads around very easily.
And even the developer tooling is a challenge too. Finally, the last thing here is on certain security postures. One of the things I always see is the shared responsibility model. I keep hammering that because every customer I talk to, only about 30% of them know like what they're actually on the hook for. Talk to me a little bit about how the different ways in which all of these clouds operate and the responsibility that still sits with the business. How does that leave them open and exposed to risk?
>> Yeah so it does, you're absolutely right. And just because you move a workload or (mumbles) service from one of the cloud service providers, doesn't mean you're automatically free. Doesn't mean you're automatically free and then lost all responsibility. So in some cases it's a matter of I'm going to hand over that responsibility of maybe pieces of my infrastructure over to that cloud provider. I may hand over pieces of that network together. But it's a matter of, to your point, you got to understand what that shared responsibility model is.
So everybody on the team within the IT organization truly understands what they're responsible for. So they still need to apply security practices, they need to still have training maybe even around these different cloud solutions so they can really understand have I exposed myself to anything when I moved something in the cloud or have I actually tightened up and improved security in some cases? >> Sure makes perfect sense. We talked a little bit about data, and the challenges of data early on, and I kind of just wanted to do a double click on this and just really emphasize this.
There's these things about data in the cloud that aren't very intuitive, that you wouldn't necessarily think about the first time you're putting an application up in the cloud, but over time become the gotchas, if you will. The first one is data gravity and latency issues. So I always like to joke with people, we still haven't found a way to travel faster than the speed of light. The distance between where the data is and where it's being consumed creates latency. And especially with rather sizable workloads that use massive files, we're also seeing that simply the size and the heaviness of the data, it can be really difficult to transport that to and from the cloud.
That's really problematic. And then we get into data management across the clouds. As we talked about, different tools, different methods, it's really difficult and confusing sometimes for people to figure out where the data is, how much data they have, what the bill for that data is. It's all hidden away and it's almost as cloudy as the cloud itself. And then the next one I really want you to maybe touch on a little bit, costs of availability and access. This is one of those things that I find that it's really interesting here.
It runs so much differently than say the spin up and spin down mentality of an application. So you talk to me about, with data, how access and egress fees can really impact your cloud deployment? >> Yeah I mean it's, as you can imagine, it's a pretty easy street to get from where your data is, likely in your data center, into the cloud. That's pretty simple highway to truck down. And what the challenge happens, and what I see people not prepared to do, is they kind of think about it as migration.
That's the migration kind of going into the data center and moving things up to one of these clouds. And they consider those costs, so they say okay even if there's a cost to transport to do that, they're thinking the benefit once I get there is big enough to outweigh those transportation costs. But then what I do see is them quickly realizing are we producing more data than we thought? Is our data growing faster than we thought? Is that an issue? Can we access that data whether it be geographic regions, whether or not the way the application's accessed?
Then it becomes very tricky and I've seen scenarios where days, even I've seen it measured in weeks, to be able to say okay now I need to in some cases, truck that data, from that cloud provider back into a data center or back into even a different data center. And it's not an ideal situation. I mean nobody wants to go through that. But we see it happening in cases. >> And you're kind of hitting on the next point, which is vendor lock-in.
Getting in is easy, getting out is hard, so be mindful about where you put your data 'cause it's such a huge competitive advantage for your company, you want to make sure that it's accessible and usable. Shifting gears to data protection. This is another one of those things, like I said, people mistake availability zones for backup. What we found, we did some recent research here, and multi-cloud has been great for giving people access to all these different technologies that help drive their business forward, but it's created massive data protection challenges.
And so, what we saw here from the 1200 people or 1,000 people that we interviewed, was that 82% of them said they had an event in the last year, they had an outage. Or they lost data. And it almost seems like it's inevitable, you're going to have a problem. And so then when you see that it takes them 66% said that it takes them three hours to recover from that downtime, especially right now in these challenging business climate where the business is IT, that's just a non-starter.
So how should organizations be thinking about data protection, business continuity, disaster recovery, all those things in one little ball here as they plan their cloud strategy? >> Yeah it has to kind of ship that those type of priorities to top of the list. And in the case of, I kind of view data protection and recovery and all these pieces, and business continuity in general, very similar to the way I look at security. There, in many cases, we don't want to go reinvent the wheel of something we're already doing well in our IT architecture.
So if we don't have to reinvent the wheel, how can we involve our data protection experts and our team early on in the cloud process? Because likely we're seeing now where those tools do have the capabilities to work with different type of cloud architectures. Maybe not all of them, but there's some ability that that can help understand okay what can we move, what shouldn't we move when going (mumbles)? And then it's just also an understanding of okay what's the, and here you got it great, what's the cost of that downtime?
What's the cost I associate with what the downtime can be? And how can I quickly get the business back up? Because you're absolutely right, it's going to happen. It's just a matter of what the impact is and what's it going to cost the business? >> That's great. So I really want to shift up gears and go to the research that you guys did for us. It was really very intriguing. So here we think that having a centrally managed cloud strategy can address some of these issues.
I'll click through it and we can talk a little bit about this what you're seeing. >> Absolutely so the first thing we kind of looked at is so we're looking at very early points, right? Are you using more than one provider? And guess what we knew, 86% saying they're using more, utilizing multiple CSPs today, and expect to kind of even do more (mumbles). So I think this is the norm that we have to operate in. It's not unreasonable to think that people are going to consume from every major cloud service provider out there for different reasons.
And by the way, for the right reasons in a lot of cases. But in order to do that, kind of went to your point that you just said, is okay how can I do that without treating each of them as an island? I don't want to treat that as an island, this is a new island, I want to be able to say okay same piece of land, how can I kind of operate across all of those together? >> Yeah makes sense. And so what did you find around this idea of consistency that we had talked about?
>> Yeah so one of the things we asked, you can kind of see here in the question is in respect to other things you're doing in IT, how important is it for your organization to enable this choice of workload placement? 'Cause that's really what it comes down to. In the past, the choice was I want my workload on this server that has all the CPU and all this memory in it so it runs faster. Well that was kind of easy back then because you just arm wrestled the other IT guy and you say all right you can use it on my server or their server, so that was pretty easy back then.
Now that there's so many different places, so it's now a matter of 83% of the organizations here want to have that flexibility, and have that as a top priority when they're making these technology decisions. >> And this is why, right? We're seeing that people continue to have struggles when it comes to figuring out which workloads to put in the cloud, right? >> Yeah, I mean we aren't perfect. I mean I know if I'm in IT, I want to move something to the cloud.
But I may run a little bit faster than I should or I may make it without kind of having more understanding the full impact of that decision. So we're seeing here, 77% have repatriated or moved cloud workloads back because whoops you know what, I may not involve the security team, that doesn't have that security posture that we're looking for. The cost model of that wasn't something I thought through, and I can actually run it through our entire IT or our (mumbles) IT, better than I could in the cloud.
And even limited access, we see this all the time where, some of these clouds are have technology or services that are new. They're still developing, they don't have the full features that new technologies (mumbles) technology has had for many many years. >> And so how can organizations address this? >> Yeah so here we wanted to know how can you, what are ways you can kind of reduce the number of security breaches, half outages, all stuff we just talked about, kind of other events here?
>> Some of the reasons why people are moving back yeah. >> Yeah exactly, so when we asked that, here respondents wouldn't believe that they can reduce that downtime, and they can reduce those security events by having a strategy that has that management consistency across clouds. I don't need to have my IT ops person speak a different language than my information security person than my developer people. I want them all to speak the same language. All under some consistent management, as they've got more clouds I'm able to avoid these events, avoid these security outages and (mumbles).
>> But the benefits of consistency go well beyond just security, right? >> Oh there's no question. Even to think about it from IT operations perspective. So can I provision resources? Can I protect resources? Can I have visibility into data that's moving? The list goes on. Where there's a lot of things that even here you're kind of seeing is a typical organization that's actually moving workloads, if they have that consistent workload, they're able to move things faster. 7 days.
Well what if I don't? That's taking 16 days, Nick that's crazy, right? >> And that's per app. >> It's crazy. >> When I heard that I always ask my customers, they go yeah you know I'm going to re-platform everything on cloud first. And then the next question I ask them is how long is it going to take you to do that per app? And they go oh 16 days. And I go how many apps do you have? And they say 1,000.
It's a multi-year project. >> It is. And I think that that's the trouble. I mean if you don't have that level of, and some of it's just if I have to relearn and retrain my organization, and I have to go out and spend money on training, if I have to spend money on new certifications, that gets very hard, it takes time for all that to happen. So here it's just a matter of let's, with that consistency where I can have that viewpoint in that management that is the same across those clouds, just really helps me be able to make these decisions.
Obviously much faster. >> For sure. And speaking of speed, developers, what do they get from this? >> Yeah no there's a little bit more, so this comes into the software development pipeline has changed. You kind of see things about certainly integration with CICD pipelines kind of being very important. So I think that really what we see happening is developers operate at a pace that IT operations is kind of needs to catch up to. So what they're seeing here is hey if I have some of that consistency, 96% expect improvements in ability to push code and production, gives them that ability to do so.
And they also, look at they're moving from the back where this isn't something that's going to happen weekly or monthly, they're looking to push that to daily, multiple times a day, and even in some cases, hourly to be able to do so. >> Yeah that's crazy especially when you think about being across multiple environments and having all these developers pushing code all over the place. >> And having no, whatever you didn't have visibility into it. What if there was simple just access problems to it?
I mean it can be a nightmare. >> And considering on this fact, I thought this was a pretty cool data point here for the business. What's the benefit in investing then in this for the business? >> Yeah I mean and so we asked here is ultimately what's management consistency give you as far as being able to have and reduce that lifecycle time? And these are big numbers when it comes to IT. Where 37% saying one to two months, 38% saying between one and four weeks.
This is all due to the fact that there isn't, what it eliminates (mumbles) and what I see it eliminates is there isn't a okay now developers have to stop and wait for IT. And then IT has to go off and do their thing which involves working with information security, information security comes back, IT operations kind of figure out some more, then kind of eventually gets back to the developer. Since there is such that consistency across those teams and across those environments, it allows them to respond.
So then maybe some approval process in there but it happens much quicker. >> Yeah and I had to think, especially this year with digital transformation becoming the thing that organizations need to do just to survive, taking on new customer touchpoints, improving productivity for workers as they're sitting at home, it seems like getting this kind of agility is very critical on top of money. >> There's no question. >> But at the end of the day, IT budgets, they're not limitless. So I think this was also a good point that we saw that there was actually a cost reduction as part of this too.
>> Yeah well think about it, if I have multiple management tools that I have to manage, operate, maintain, train on, there's a cost to doing that. Nevermind okay now what if I have to make a change or what if something different happens? So yeah we saw that on average, the cost improvement with that consistent management really looks like it's almost 20%. And I really see things move inside IT organizations when things, without single digits, nothing happens. Double digits and the teens, people start to consider it.
When it gets upwards of 20%, people start to realize okay. Even if there's an initial cost to that investment, I know my savings over time is just going to save me from hey I can operate better throughout the day, I can sleep better at night, and I don't have to worry myself about the weekends as much either. >> Good yeah. Our final slide here, just summarizing what we found from all these data points across everyone we interviewed. Why don't you just take us home here and explain?
>> What's kind of cool here is some people have lept out in front of this, and do have established some consistency across those platforms with management solutions. And here are the benefits. I bet nearly, the one we just talked about, nearly 20% expect cost reduction. There's a big cost too, at some point, I'm going to move something so if I've got to move something that's going to take me time, well I can save 35% expected that kind of reduction in time. We talk a lot about how many person hours is this going to save a week?
Exponentially based on the number of applications of workloads it can have, can make a huge difference. Certainly from the project cost savings, so a new application, maybe modernizing a new application, developers are off using these cloud services, we want to encourage that, we want to encourage the builders to do so, being able to have that consistent management drives down those costs even better. Kind of the timeline, every project, especially those bigger application development projects, always have overages, they're able to contain it more because there's consistency sure, but there's options to get to tools that they wouldn't been able to get to in the past, or new cloud services, and they can do so there.
And then the other one, obviously, people that are kind of built on IT and built on products, this where I can get changes out and I can get product out faster than I ever have. >> So I mean it seems like there's something for everyone in the business, developers, business units, IT. But in order to get there, I think that it's IT's role to really become that business enabler, and if you want to call it a cloud broker, maybe that's the term, I don't know.
They need to be the people who empower the business to get to the cloud and tap into the benefits of wherever they're going to deploy. >> I totally agree. And it has to there's no, the business benefits alone, forget about the technology benefits, are really can be phenomenal. And you're absolutely right, IT has an excellent opportunity to step in and say all right let us really be that point where we can help decide and make the best decisions, maybe where a workload may go.
But, you mentioned all these things earlier, we can still protect the data, we can still do good things with that data and apply it's elegance to it. And all along where we're kind of providing that more predictable operations across all those environments, taking advantage of all those pieces. >> Well Mark, you've been great. Thank you for joining me today. And for everyone who's listening at home, thank you for taking the time to learn a little bit more about our approach to cloud, and some of the things that we found in the cloud.
com/cloud and there's an infographic and a lot of great research there that you can tap into, and you can also learn more about the Dell Technologies cloud. Thank you for joining me. (swoosh)
