Q&A with ColoSpace’s Aaron Sawchuk: Buckle In, Hybrid Architecture is Going to Be Around Awhile
BOSTON, MA — Aaron D. Sawchuk is the chairman and co-founder of ColoSpace. Aaron sets and helps implement the overall technology and product development strategy at ColoSpace, focusing on building highly-available IT services that add significant value to customer operations. Aaron brings extensive experience in the IT-services and Internet Networking space, having spent more than twenty years founding and leading technology firms in New England. Sawchuk will speak on a panel at CapRE’s upcoming Second Annual Great Boston and New England Data Center Summit on June 22. In anticipation of the summit, we connected with Sawchuk to learn about his latest activity and observations on his region.
CapRE: Thank you for connecting with us, Aaron. Please share with us some thoughts on what you’re seeing out in the industry.
Sawchuk: From our perspective, to look at the landscape here, the biggest changes or things that I think are interesting is that we’re seeing larger and larger enterprise customers really embracing hybrid as a real, long-term strategy to their overall data center and IT outsourcing initiatives. And we’ve been talking about hybrid for the last several years.
CapRE: Clarify for our readers. What is hybrid?
Sawchuk: When I talk about hybrid it’s typically a mix of data center colocation, some level of managed services and some level of cloud services that organizations may be procuring directly from other public cloud providers. But what has changed in the last year is that hybrid strategy was seen as a temporary phenom. It used to be that they would want to colocate some existing infrastructure, have them age out, and then move to the cloud. The organizations that embraced the hybrid strategy were small to medium organizations and now what we’ve seen, especially in our marketplace, is much larger organizations adopting that strategy. Multiple cloud vendors, multiple data center vendors, a mixture of traditional colocation, software-as-a-service and cloud.
When I talk to CIOs and CFOs and senior managers at those firms, they realize that hybrid will be their IT strategy for the long-term. So rather than getting to some end-game where everything will move to a single platform, they expect to work with multiple vendors, technology, and delivery models for the next five years — not just the next year or two. That’s a big shit.
CapRE: What other developments or dynamic characterize your observations of 2018 so far?
Sawchuk: The other is sort of an interesting point. As we look for our planning and budgeting for 2018, we wonder what the changes in the tax legislation will do for our customers’ decisions in purchasing. Some may have been more reluctant in the past to make capital investments. They can now realize the immediate benefit of that capital investment instead of writing it off over a number of years with an OpEx strategy. Colo growth is typically driven by OpEx and Cloud growth by CapEx.
CapRE: So has that expectation been borne out?
Sawchuk: The interesting thing is that we haven’t seen a seismic shift. We continue to see cloud growth accelerate but what we did see is a much more of a willingness for firms to make those CapEx investments and deploy infrastructure in a data center. While we haven’t seen the Cloud growth come down, we are seeing pockets of demand develop. Everything else being equal they now have much more of a willingness to write checks. Also, this is likely reflection of their an appetite for investment and their outlook on the economy in general. It may be one of the drivers for hybrid approach as well.
CapRE: Very interesting.
Sawchuk: Another underlying trend is customer infrastructure deployments on a net-new basis. They’re not forklifting infrastructure out of an office data center and into colocation, but making net-new tech investments. Those investments are consuming dramatically more electricity and cooling than what we have seen.
CapRE: Not surprising.
Sawchuk: We’ve heard this story for the last ten years – deployments getting denser. But we’re seeing it rapidly accelerate in the last 10 months. The average customer 18 months ago was deploying about 5 kilowatts per rack in a deployment. The net-new ones today are closer to 15. That’s almost a 3x increase in just the last couple of years. Some of that has to do with the types of workloads being deployed – there’s a lot more investment in AI, big data and analytics. IT solutions that require GPUs to do complex math operations as well. Those are all absolutely increasing the overall electricity footprint.
CapRE: So how has this impacted your strategy for the future?
Sawchuk: As we look to our medium-term strategy over the next three years, we fully expect that data centers will consume significantly more power than they are today. That’s going into our engineering and planning as we contemplate opening new facilities. This “density conundrum” has certainly re-emerged, after a period of several years where it flatlined.
CapRE: How about sales?
Sawchuk: Our customers demand for us to make it very easy for them to execute on their hybrid plan. We’ve deployed AWS Direct Connect, Azure ExpressRoute, Salesforce Direct Connectivity, and we have access today to about 2 dozen cloud on-ramps and our peers in the industry are doing the same thing. As we think about our strategy in other ways, it’s incumbent on us to continue to invest in our product capabilities to continue to deliver a depth and breadth of services to our customers to help them meet their medium and long-term goals.
We like to differentiate ourselves from the large REITs through those services and it’s been very effective, but we can’t take our eye off of the ball. Our goal is to do four major product launches a year and some might be an enhancement, but some might be a net-new capability. Last year we launched a net-new disaster recovery as a service offering for example. And building upon some unique features and benefits that when we survived the landscape weren’t readily available.
It’s about being able to accommodate denser workloads and on the services side, it’s about making sure we can enable customers to leverage a lot of these technologies and differentiate ourselves from our peers.
CapRE: What’s the last word on all of this?
Sawchuk: What I’ll close with is that the data center industry in general has really over the last year or two years continued to mature. And continued to become much more of a diversified industry. What I think is interesting is that in addition to the four data centers that we operate in New England, we also partner with 3rd party providers in Ashburn and Dallas — and soon will be in Germany later this year — to be able to deploy workloads out of a host of different regions.
I’ve met with a number of providers in various geographies, and it used to just be a rack is a rack is a rack. A rack in Dallas was the same as Boston. That is still true, but what we are seeing as part of the industry maturation is that the types of infrastructure and companies deploying, and the regions they are deploying in, has become much more strategic.
While we have, for example, the most expensive electricity in the U.S. in New England, we continue to see demand from firms who want to be here. And part of that is the Edge network concept. Part is about lower latency. But there really are unique demand characteristics in each of these markets. You’re not going to see massive hyperscale data centers in Boston because of power costs and labor costs. But you are seeing healthy, steady growth in that economy anyway. The industry has matured and it’s been a development of various demand characteristics in unique markets. And I believe that it will drive even more success for the industry as whole, as we find our competitive advantage based on our geography. It allows us to become much more focused. We can still service a customer more concerned about latency from a low-cost location such as Dallas.
CapRE: Thank you for your time Aaron. We’ll see you in Boston.