It’s Time to Take a Chapter from the Hyperscale Location Strategy & Site Development Playbook
Mike Lebow from Stream Data Centers explains how industry trends have made location strategy and site development more complex for everyone, and what colos can learn from the hyperscale playbook.
With perspective as both a hyperscaler and a colocation provider, in this edition of Voices of the Industry, Mike Lebow from Stream Data Centers explains how industry trends have made location strategy and site development more complex for everyone, and what colos can learn from the hyperscale playbook.
An important trend is taking shape in the data center industry. Traditional colocation providers are being driven to find and develop larger sites to serve a variety of customer needs. Hyperscalers are increasingly looking to colocation providers for massive build-to-suit data center buildings, and even campus-scale deployments. At the same time, enterprise customers are also taking down larger and larger tranches of data center capacity.
The evolution of hyperscale demand is driven by the world’s largest tech platforms, including the exponential growth of the Cloud; Cloud providers are entering many new markets simultaneously; supply chain and labor constraints are making self-performing at massive scale across a huge footprint even more difficult; and hyperscalers are struggling to meet their demand signals with internal staff, resources, and existing site portfolios. On the other hand, the evolution of enterprise demand is driven by multi-tenant colocation and enterprise requirements scaling to levels previously limited to the largest tech companies on the planet. As my colleague Chris Bair has written, “In enterprises’ drive for efficiency, resiliency, and value, they’ve effectively started to deploy how and where hyperscalers do.”
Historically, the site selection process for colocation providers included finding and buying sites that were already fully developed, often in existing business parks. These ‘off-the-shelf’ sites were shovel-ready – with utilities, graded land, roads, etc. They didn’t require bringing in high voltage power lines and building new substations, moving or vacating existing easements or utilities, or mitigating complex environmental risks on greenfield land. However, in today’s environment, those ‘off-the-shelf’ sites are increasingly rare, and it’s even harder to find straightforward sites that meet the massive size and infrastructure capabilities necessary to serve hyperscale and enterprise customer needs.
So old site selection methods don’t work anymore. In order to mitigate risk and enable scalable capacity in this new environment, colocation providers need a robust and methodical location strategy and site development processes – the kind of processes hyperscalers have long relied on.
A robust and methodical location strategy and development process
To help illustrate what this new location strategy and site development process could look like in today’s environment, I’ll outline the steps we follow as a colocation provider pursuing a build-to-suit project or planning for our own turnkey colocation facility.
Define critical location factors: Our first step is to gather data points from our customers and develop a list of Critical Location Factors (CLFs). Some of these factors are consistent across customers – for example, access to reliable high-voltage power infrastructure, acreage to enable scalable growth, etc. Others are more dependent on the particular customer’s requirements, such as access to water (some customers use water for cooling; others do not) or proximity to existing sites or networking nodes.
Look broadly at a range of potential sites: With our customer’s CLFs in mind, we work with various stakeholders in a region or utility territory to identify a long list of on- and off-market properties. Some of those stakeholders (large economic development agencies and utilities, for example) may be very well versed in the needs of data center operators and come to the table with vetted and qualified sites. Others, such as independent landowners or family farms, need guidance on defining site characteristics, conducting due diligence, and positioning their site to be sold to a data center operator.
Due diligence and development: Once we identify sites with the strongest fundamentals, we get to work on the due diligence and site development process. We conduct a rigorous, multi-phase due diligence process to identify and characterize any possible site risks. We develop mitigation plans when possible and walk away from sites early in the process if a risk will be too costly or time consuming to overcome. Once we settle on a site, we dive headfirst into development activities. Long before a single shovel hits the ground, we work with power, water, sewer, gas, and other utility providers to assess current system capacities and determine what upgrades or interconnection projects are required to secure as much capacity as possible to enable a range of densities and master plans for potential customers. We complete rezonings or special use permits, striving to make the allowable uses as flexible as possible. We mitigate wetlands (when appropriate) to maximize land use and minimize permitting and construction delays. We work with government agencies to outline development and investment parameters and negotiate responsible and mutually beneficial economic development incentives. Bottom line, we strive to identify all possible risks and opportunities, then work to address them all to ensure smooth delivery of capacity.
Maintain a creative development mindset: In today’s environment, there is no such thing as a perfect site. Every site will require risk mitigation; the key is to identify each project’s unique challenges, and to be creative in assessing how risks can be most quickly and easily mitigated. Creativity enables us, for example, to find optimal solutions for utility upgrades and interconnections, expand the site through consolidation with adjacent landowners, or work with local communities and planning/zoning departments to move roads, re-route infrastructure, secure easements, etc. to optimize the site’s growth potential. At the same time, mitigation of any given risk has multiple potential paths, and oftentimes the various project attributes are interrelated. For example, expanding a substation means less room for stormwater retention, providing flexibility for future easements restricts building locations, and taller buildings require additional setbacks or landscaping. All factors must be managed when optimizing a site.
Be prepared to work for it
Complexity is a defining characteristic of location strategy and site development in today’s environment. Even the most sophisticated hyperscalers and colocation providers face the complexity challenge and have to engage a variety of stakeholders and work hard to mitigate risks associated with any site.
An example from my time in site selection and site development for a major hyperscaler is particularly illustrative. We were looking for a site in the eastern U.S. that would provide an enormous amount of power capacity to enable long-term portfolio growth. We engaged with a large utility company on a multi-hundred acre site that was being used to extract dirt for remediation of a retiring coal-fired power plant. Development required complex legal contracts, a massive site grading plan, construction of a new substation and high voltage transmission lines extensions, and a multi-party development agreement for multiple water and sewer infrastructure upgrades. A lot of hard work, but the result was a site that enabled low-cost growth in one location for years to come.
Another example is from five years ago in the most well-known of data center markets: Northern Virginia. As one of the most in-demand data center markets in the world, ‘off-the-shelf’ sites had long since been absorbed. We were able to find a relatively straightforward property, but it was not large enough to ensure future scalability. With our creative development mindset, we identified and purchased adjacent parcels of land that were earmarked for an alternative use, which enabled us to more than double the footprint of the site. However, in order to make the site a single parcel, we had to vacate a county road, dedicate a portion of our land to create a new public right-of-way, move underground utilities, and create a new emergency egress for an adjacent water treatment plant. With five parties in the mix, it took months to get all stakeholders onboard, and even longer to complete the upgrades, but the efforts resulted in a large contiguous campus in a critical market.
These are just two examples of the kind of work hyperscalers have had to do in order to develop suitable, scalable sites and the approach colocation providers will have to embrace in this new environment. (In fact, that is one reason Stream Data Centers is growing our Location Strategy and Development team – to support Stream’s mission to find high-quality sites and deliver capacity to our customers when and where it’s required.)
For hyperscalers and colocation providers alike, location strategy and site development are critical components of the data center development project lifecycle. It’s more complex and challenging than ever before, but getting it right – with a robust and methodical diligence and development process – is what enables us to deliver sustainable, scalable, and cost-optimized capacity on time and without any surprises.
Mike Lebow is Senior Vice President of Location Strategy & Development at Stream Data Centers, which builds and operates data centers for large and sophisticated enterprises and hyperscale cloud providers. Prior to Stream, Mike spent nearly eight years at Google on the Energy and Location Strategy team.