Striking a Balance: Optimizing Enterprise Data Centers and Public Cloud
[00:00:00] John Kundtz: Hi, everybody. My name is John Kundtz, and welcome to this edition of the Cloud Collective podcast. Today's guest is Robert Erickson. He's a serial entrepreneur, cloud computing pioneer, and vice president for strategy and growth for Kyndryl's cloud practice.
I am super excited to discuss a hot industry topic today: balancing cloud and traditional data centers. Robert will share valuable insights and advice, especially regarding common pitfalls organizations face as they transform into digital businesses.
Welcome to the show, Robert.
[00:00:44] Robert Erickson: Hi, john. Thank you. Happy to be here.
[00:00:46] John Kundtz: It's great to be with you. Robert, you and I have worked together since 2015. That's going on eight years now, and we've been through a lot in those eight years. But I'd like you to take a moment to tell our audience a bit more about you. And you're welcome to start anywhere you feel fit.
[00:01:07] Robert Erickson: Thanks, John. It's always difficult to answer the question of where to begin. As you said, I've been through many startups. The last one was Gravitant, where we built out what was the first cloud brokering platform and sold it to IBM in 2016. Then, we went through the separation spinoff of Kyndryl from IBM Global Technology Services.
And then I see the future; why am I still at Kyndryl? Kyndryl has a unique opportunity to take advantage of a major inflection point in the industry. I've been helping them do that and laying the groundwork for the next round of new innovations in the IT industry.
[00:01:48] John Kundtz: That's great. Yeah, why I let you introduce yourself because, like I said, we could go on and on about your background.
We got onto this podcast partially because of an article that was published on December 12, 2023, in the Wall Street Journal entitled, The Cloud Isn't the Answer to All IT Problems, at least for now.
From your perspective, what's your point of view on the idea of balancing enterprise data centers and the public cloud, and what are some of the significant benefits organizations to successfully integrate this balancing act into the organizations?
[00:02:25] Robert Erickson: We started seeing this trend about a year ago in Ernst, and something that was demonstrated in data that we track our customers, if you roll the clock back a bit, circa prior to the pandemic, in 2019, There was a lot of interest in saving money, taking advantage of the hyperscaler and cloud providers services.
They were working well together, and we saw a lot of people exploring, “Can I move my workloads?”
Get out of the business of running data centers, the costs associated, the management headaches of having to keep it up to date, licensing agreements, and all of that. Can I place that on to a third party, use their services, and come out ahead?
It started with a drip, like any of the technology trends, and this became a major movement; then, in 2019, when the pandemic hit, we went from just cost savings to mere survival. There was an existential threat. Customers and the companies serving those customers need to be reached.
So quickly, we moved to online digital experiences and accelerated that. It was a massive acceleration. Overnight companies were suddenly available online. Even small mom-and-pop restaurants were seeking to place orders online and pick them up in the driveway or the parking lot without touching them.
So that really moved the industry forward. And so, there was even more of an emphasis on the cloud and the potential of the cloud. But as we reach 2023 and the end of the year and look forward, Our customers have had enough experience. Now, there's enough maturity and understanding what the cloud can and can't do.
And so, talking to the analyst groups like Gartner and Forrest or IDC, they're coming to the same conclusion. Why are we today in 2023 with all that momentum behind it? Less than about 30 percent of mission-critical workloads are in the cloud. And that's because the easy stuff, the low-hanging fruit, has already been done.
The remaining 70% will continue to be retired and rebuilt in the cloud. Some of it may be moved to the cloud, although that's decreasing. The reason that we're at that high percentage is because it's difficult to move. These applications may have dependencies on licensing agreements and contractual agreements.
They may have technical considerations like integrations, data, and system interoperability that aren't easily replicated in the cloud. More importantly, they may have regulatory compliance and security concerns, and more customers are beginning to understand that the cloud was initially positioned to save a bunch of money; they didn't really realize that.
And so with those learnings, there's almost a pause in the market while people reconsider, and in that reconsideration, they're balancing, too.
- What do I already have a sunk cost?
- What do I need to keep?
- What is core to me?
- How do I secure it and protect it?
Versus, how do I make things faster?
What can I use the cloud capabilities and the hyperscalar capabilities now become a mission or a force multiplier.
But it's part of the solution, not the end solution, so they're balancing those two, and that's where the complexity comes in.
[00:06:07] John Kundtz: It makes a lot of sense. We certainly saw this notion of moving from a CAPEX model, which is essentially your traditional data center where you go out and buy a whole bunch of equipment and depreciate over time, to more of an OPEX model.
Certainly, certain workloads and points in the development cycle are probably more suited for that OPEX type of approach. As you alluded to, certain things make sense to do as more of a traditional and or own that infrastructure.
A lot of organizations are trying to push everything into a public cloud model what are some of the common pitfalls or challenges organizations face as they try to transform themselves into digital businesses?
[00:06:50] Robert Erickson: Yeah, the first thing is that you could start with technology as the basis and build your way into value and outcomes without understanding those outcomes upfront.
So, I think it's reversed. And that's that maturity I alluded to and talked about earlier. The customers we're seeing have now become fairly adept and knowledgeable about what the cloud can and can't do.
- So, if we're talking about, what is the final outcome?
- What are we trying to achieve?
- Is it a cost-saving in general?
- Is it reaching new markets?
- Is it creating innovation and instantiating that and making it, repeatable and part of the core culture?
Even reaching new markets globally. So you start with that, and the pitfall is that if I think that I can avoid that conversation and just build something in the cloud and they will come, you risk what we do, who we internally call doing cloud poorly.
It's usually more costly, and you don't achieve the results that you are likely to seek. So, it starts with a firm understanding, usually in collaboration with the customer, which is the best idea.
If you have a joint meeting of “the minds,” customers often don't even know the art of the possible. They don't know what the destination is. I don't know what's possible with these new technologies. So having that conversation up front, working with that client, working through it, is critical.
We have playbooks that have emerged, but trying to do it without that and just building in the cloud to start with, you end up with, what is the old adage? I'm a solution, seeking a problem.
[00:08:34] John Kundtz: I'm a hammer looking for a nail.
You saw this over and over again, at least in my career, where technology emerges, and people just try to rush into it for the technology's sake.
Quickly, they learned that it's not necessarily about the technology, the raw technology, even though it's super innovative and there's great potential for benefits. Still, sometimes you just have to take a step back, like you said, and not just do it because it's there and it's better.
But what are you really trying to accomplish?
And what's the best method and process to move forward?
I've seen that in my career over and over again.
[00:09:09] Robert Erickson: Yeah, absolutely. And, part of it is this assumption, especially prior to the pandemic, that somehow, we're going to take all the headache of I.T. and transfer it to somebody else, like the hyperscalers.
In the end, it's your business, and customers understand there are things that properly belong in the cloud, things like the tooling that they provide, the speed, the new digital experiences, and the fast, innovative access to services, all perfect for a cloud situation.
But we also have tremendous investment in things like data centers and data. And I would say that even the data center itself and the legacy or traditional stack IT are significant investments. But it's not static. It's not staying still, either. It also is progressing. Having an investment in a data center and updating it to handle things like large language models, generative AI, being able to process and protect data on-prem, even the, AI models themselves are core IP.
Customers are beginning to transfer what is tribal knowledge into AI models and then train them with that data. Do you really need or want that data transmitting back and forth at a high cost across the internet? Maybe, but it depends on the situation. So that balance is coming through.
[00:10:38] John Kundtz: If you were to give one piece of advice to someone who is considering embarking on a digital business transformation, what would that be?
How do you get started?
Where do you go first?
[00:10:48] Robert Erickson: Like any knowledge or technical aspect, you start with somebody that's been there and done it. You find patterns or examples of who's actually done this at scale or has done it with the outcomes that you're looking for. Find that group and join some of the guilds out there.
LinkedIn is a good place. Your Kyndryl representative, if you want to go the Kyndryl route or your system integrator may be available, start there by asking questions and start to develop knowledge and education on that aspect. Maybe even try a Hello World on some of these things and get started with that. Low cost, low risk.
I wouldn't try to plan some new technology paradigm on a mission critical that has a short timeline. Start with something a bit more innocuous, something if you have changes along the way that doesn't disrupt your business.
We've done that with a few customers.
[00:11:47] John Kundtz: It sounds like the classic crawl, walk, run analogy where we pick something. That is fairly small and isn't, going to run your whole business, but important to team up with somebody who has knowledge of both the traditional and the cloud world. And then in the agile process, you learn, and then you pivot where it makes sense, or you continue to do what's working.
[00:12:12] Robert Erickson: Yeah, that's a great analogy.
[00:12:16] John Kundtz: Well, Robert, as always, it's great to be with you. I always learn something new when I talk to you, so thanks for joining our show today.
I have one last question, and that is, how can people learn more about what we just discussed today?
[00:12:30] Robert Erickson: Yeah, there will be a series of blogs that we're publishing over the next couple of weeks. Around this topic to really dig in. So this is the first in the series. You can come to, and we'll have the links and all of that available to you.
I'll share those, John, with your audience and look over the next couple of weeks. We're going to have some interesting customers. You can hear it from the customer, and we're going to have some analysts from the industry.
And it's always good to hear it from where the rubber hits the road; the development managers and the dev managers who actually implement it will have a series of those included. So you can hear for yourselves how to do this.
[00:13:10] John Kundtz: Excellent. I also encourage the audience to connect with Robert directly on LinkedIn.
We will post the links to Robert's profile, as well as other items we talked about during the show.
Remember, you can succeed in managing and optimizing your cloud and traditional data centers by bringing financial accountability and predictability to the highly variable expenses of the cloud by creating a modern operating model between IT, finance, and your business teams.
Well, that's it for today. I'm John Kundtz and thanks for joining us on the Cloud Collective Podcast.
Have a great day