Last updated: August 24, 2026
How Hyperscalers Evaluate Data Center Sites
Landowners often assume fiber, location, or price per acre is what makes a parcel attractive to a data center buyer. In 2026, it's power deliverability, by a wide margin — the discipline of separating sites that can actually be energized from sites that merely look available. Here's how site-selection teams actually weigh the criteria, and what that means for how you should describe your own land.
⚡ TL;DR — What Buyers Actually Screen For
- • Power deliverability is the top criterion — substation/transmission headroom and a realistic interconnection timeline, not just proximity to a line on a map
- • Water is both engineering and political — wet-bulb profile and free-cooling hours drive the cooling design; local opposition to water use can stall a project independent of the engineering
- • Fiber and cooling are "site-makers," not the primary driver — necessary, but they don't compensate for weak power access
- • Land cost matters least — typically a low-single-digit share of total hyperscale project cost; land quality (acreage, topography, zoning) matters far more than price
- • Buyers plan for expansion — campuses increasingly scale from tens of MW to several hundred MW in phases, so room to grow matters beyond the first building
Criteria, Ranked by How Much They Actually Move a Decision
| Criterion | Weight | What buyers look at |
|---|---|---|
| Power deliverability | Primary driver | Substation/transmission headroom and realistic interconnection timeline — now weighted above fiber and land cost combined |
| Water access & cooling fit | High, and rising | Wet-bulb profile, free-cooling hours, municipal/well/surface source, and local political sensitivity to a large new water user |
| Fiber connectivity | Site-maker, not primary driver | Proximity to existing routes; a site can have great power and still fail without adequate connectivity |
| Land quality & acreage | Moderate | Contiguous acreage, topography, environmental constraints, and zoning status matter far more than raw price per acre |
| Land cost | Low | Typically a low-single-digit share of total project cost for hyperscale-size deals — rarely a deciding factor on its own |
| Expansion capacity | Strategic | Buyers increasingly plan phased campuses scaling from tens of MW to several hundred MW — room to grow beyond phase one matters |
Site Intake
Not sure how your land stacks up against these criteria?
Submit it for a confidential review — we'll evaluate it against current buyer standards.
Your information is reviewed privately. We only use submissions to evaluate potential fit and relevant opportunities.
Why Power Now Beats Fiber as the Top Criterion
For years, fiber proximity was treated as close to a gating requirement — no fiber, no deal. That's shifted. Grid operators in some of the most active data center territories have reported large-load interconnection requests that vastly exceed what can realistically be energized in the near term; in a few grid territories, requested capacity has reportedly run into the hundreds of gigawatts against only a small fraction of that actually drawing power today. Fiber, by contrast, can usually be extended to a promising site given enough lead time and budget. New transmission and substation capacity generally cannot be built nearly as fast, which is why power deliverability — not fiber — is now the criterion that makes or breaks a site.
That doesn't mean fiber is unimportant. It means fiber functions more as a "site-maker" than a primary screen: a site with strong power access and weak fiber can often still work with investment, but a site with weak power access rarely gets a serious look regardless of how good its connectivity is. See our fiber connectivity requirements guide for what buyers specifically check.
Water: An Engineering Input and a Political One
Site-selection teams model water two different ways. On the engineering side, they look at wet-bulb temperature profiles and annual free-cooling hours to decide whether a site supports lower-cost evaporative cooling or needs a more expensive air-cooled or closed-loop design — a decision that materially changes both capex and how much water the facility will actually draw. On the community side, water availability has become one of the more common flashpoints for local opposition to new data center proposals, independent of whether the engineering numbers actually pencil. A site can be technically fine on water and still face real delay or rejection if local stakeholders are unconvinced. See our water and cooling requirements guide for how this plays out by region.
Land Quality Over Land Price
Landowners frequently overweight price per acre when thinking about what makes their parcel attractive. Buyers generally don't: land cost is typically a low-single-digit share of total project cost on hyperscale-size deals, so a below-market price rarely tips a marginal site into a serious offer. What matters far more is land quality — contiguous acreage without title complications, workable topography, minimal environmental constraints, and zoning that's either already industrial or realistically rezonable. A parcel that's clean on those dimensions but priced at or near market will usually outcompete a cheaper parcel with assemblage, wetlands, or zoning problems. See our colocation vs. hyperscale requirements guide for how acreage expectations differ by buyer type, and our land due diligence checklist for the specific issues that come up most often.
The practical takeaway for landowners
If you're deciding what to lead with when describing your land, lead with power: how close you are to a substation or transmission line, and anything you know about available capacity there, even approximate. Follow with water source and acreage. Fiber and price are worth mentioning but shouldn't be the headline — buyers now assume those are solvable if the power and land fundamentals are right.
Frequently Asked Questions
What's the single most important factor in data center site selection today?
Power deliverability — not fiber, not land cost, and not even raw acreage. Industry site-selection teams increasingly describe their job as separating sites that can actually be energized within an acceptable timeframe from sites that merely look available on paper. A parcel with excellent fiber and cheap land is still a non-starter if the nearest substation has no headroom and the interconnection queue runs five-plus years.
Why has power availability overtaken fiber proximity as the top criterion?
Because interconnection queues have become the actual bottleneck on project timelines, not connectivity. Several major grid operators have reported large-load interconnection requests that vastly exceed what can realistically be energized in the near term — in some territories, requested capacity has run into the hundreds of gigawatts against only a small fraction actually drawing power. Fiber can often be built to a site; new transmission and substation capacity generally cannot be built nearly as fast.
How much does water availability actually matter in site selection?
It's become both an engineering input and a political one. Buyers evaluate wet-bulb temperature profiles and annual free-cooling hours to model whether a site supports evaporative cooling or requires a more expensive air-cooled or closed-loop design — and separately, water availability has become one of the more common triggers for local community opposition, which can stall or kill a project regardless of the engineering numbers. A site with technically adequate water access but strong local opposition to a large new water user is still a real risk.
Does land cost matter much to a hyperscale buyer?
Less than most landowners assume. Land cost is typically a low-single-digit share of total hyperscale project cost, so buyers rarely walk away from an otherwise strong site over price alone. Land quality — contiguous acreage, proximity to substation capacity, topography, environmental constraints, zoning status — matters far more than the per-acre price, which is why a well-positioned parcel can outcompete a cheaper one that's harder to develop.
What should I document before submitting my land for review?
At minimum: approximate contiguous acreage, proximity to the nearest substation or transmission line (even an estimate), any known water source (municipal, well, or surface), whether fiber routes run nearby, and current zoning. You don't need a formal power study or engineering report to submit — describing what you know accurately is more useful than guessing at numbers you're not sure of.
Site Intake
Ready to see how your land stacks up?
Submit it for a confidential review against current hyperscale and colocation criteria. No obligation.
Your information is reviewed privately. We only use submissions to evaluate potential fit and relevant opportunities.