TL;DR
Most data center sustainability marketing is theater. Real infrastructure buyers ask three pointed questions instead: Is the PUE measured or marketed? Where does the power actually come from? And does any of this lower my cost, or just my guilt? In a well-run facility, sustainability and cost efficiency are the same conversation. Wasted power is wasted money. This post gives you the questions to ask and the answers that separate a real sustainability program from a green logo.
The Two Sustainability Conversations
There is a version of the data center sustainability conversation that is mostly marketing. Net-zero pledges with a target date far enough out that nobody signing the press release will still hold the role. Tree logos. Vague commitments to “green” operations.
Then there is the version actual infrastructure buyers have. It is much shorter and much more pointed. When a buyer raises sustainability in a colocation evaluation, they are usually asking three things: what is the real efficiency number, where does the power come from, and does this cost me more or less. This post is about that second conversation.
The framing matters because sustainability has quietly become a procurement-diligence problem rather than a feel-good story. Regulated buyers increasingly need this data for their own reporting, which means the answers have to survive an audit, not just a slide. Here is how to ask the right questions and read the answers.
The PUE Question, and How to Tell Measured From Marketed
Power usage effectiveness (PUE) is the headline sustainability metric in this industry, and it is also the most frequently misrepresented. Start by knowing exactly what it measures.
PUE = Total Facility Power ÷ IT Equipment Power
A PUE of 1.0 would mean every watt entering the building reaches your equipment, with nothing spent on cooling or overhead. That is physically unreachable. The useful question is not “is it close to 1.0” but “is this number real.”
What a good PUE actually looks like
| PUE | What it represents |
|---|---|
| 1.0 | Theoretical ideal, physically impossible |
| ~1.1 | Best-in-class hyperscale facilities |
| 1.7 | Industry average (Uptime Institute survey) |
A facility quoting a strong PUE should be able to share operating numbers, not just a design target. A provider that can only offer a design PUE, and cannot show what the facility actually runs at, is telling on itself.
Three ways a PUE number misleads
First, PUE can be gamed by where you draw the boundary. Move certain power loads outside the calculation and the number drops without anything actually improving.
Second, PUE ignores the energy source entirely. A facility running on coal and one running on wind can post identical PUE figures. The metric says nothing about carbon.
Third, climate skews it. A data center in a cool, dry climate will show a lower PUE than an identical facility in a hot, humid one, regardless of how well either is operated. Comparing PUE across facilities in different regions can mislead you.
Why the number lands on your bill
Here is the part buyers care about most. Efficiency is not only an environmental story. It shows up in cost. Industry analysis puts the value of a 0.1 PUE improvement at roughly $640,000 in annual energy savings per megawatt of non-IT load avoided. A high-PUE facility wastes resources, costs more to run, and those charges get passed on to the tenant. You pay for inefficiency whether or not it appears as a line item.
Where the Power Actually Comes From
Once PUE is settled, the next question is sourcing. This is where “green” claims get slippery, so ask for specifics.
There is a real difference between a power purchase agreement (PPA) that funds new renewable generation and an unbundled renewable energy certificate (REC) bought to offset existing consumption. Both can be legitimate. They are not equivalent. Ask which one backs the claim, and ask whether there is additionality, meaning the provider’s commitment actually brought new clean energy onto the grid.
Siting matters as much as procurement. Because grid mix varies enormously by region, a facility on a low-carbon grid can produce dramatically lower Scope 2 emissions than the same facility on a high-emissions grid. Where the building sits is part of the sustainability answer, not a detail.
For most buyers, the practical value of a provider’s sourcing program is that it removes work from your team. In a recent S&P Global Market Intelligence survey, 67% of colocation customers said working with a provider that prioritizes carbon-free energy sourcing reduces the administrative burden on their internal teams. The same research found a 10 MW customer could save roughly $1.3 million per year procuring renewable energy through a provider’s economies of scale rather than buying it alone. This is exactly where a provider’s wholesale scale, passed down to retail deals, becomes a sustainability advantage and a cost advantage at the same time.
Does Sustainability Lower Your Cost, or Just Your Guilt?
This is the question that separates a real program from green theater, so it deserves a direct answer.
In a well-run facility, sustainability and cost efficiency are not in tension. They are the same thing viewed from two angles. Wasted power is wasted money. A better PUE means less energy spent on overhead, which means a lower operating cost, which a disciplined provider can reflect in your pricing.
Billing structure is where this becomes concrete. Colocation increasingly bills on actual metered energy use plus a defined power and cooling overhead. That structure rewards tenant efficiency directly. The more efficient the facility and your deployment, the less you pay. Sub-metering turns sustainability from a marketing claim into a number on your invoice.
So the honest answer is that real sustainability usually lowers cost, and green theater usually raises it. A provider treating efficiency as an operating discipline tends to be better on both fronts. A provider treating it as a campaign is often quietly passing inefficiency through to you.
Beyond PUE: What Buyers Do Not Ask, but Should
PUE and sourcing cover most of the conversation. Two more factors are worth raising, because they separate a sophisticated buyer from a checklist buyer.
Water, in local context. Globally, data center water use is tiny relative to other industries. Locally, in water-stressed regions, it can be significant and politically sensitive. Counterintuitively, the move from evaporative cooling toward closed-loop liquid cooling tends to reduce water use rather than increase it. If your facility sits in a constrained region, ask about cooling design and water usage effectiveness (WUE), not just power.
Embodied emissions. On a clean grid, the carbon embedded in steel, concrete, and IT hardware can represent 30% to 50% of a data center’s total emissions, and it is very hard to reduce. You will rarely get a precise figure here, and that is fine. Raising the question tells you whether the provider thinks about sustainability as a full lifecycle problem or just a marketing surface.
One honest caveat worth keeping in mind through all of this: there are no official statistics for this sector. Most published figures are estimates, and water and emissions data in particular are weak. Treat any single number, including a provider’s, with appropriate skepticism. Ask how it was measured.
FAQ
What is a good PUE for a colocation facility?
The industry average is around 1.7, best-in-class hyperscale facilities approach 1.1, and 1.0 is a theoretical limit no one reaches. More important than the number is whether it is a measured operating figure rather than a design target.
Does choosing a sustainable data center cost more?
Usually the opposite. Efficiency lowers operating cost, and metered billing passes that savings to the tenant. Real sustainability tends to lower cost, while green theater raises it.
How do I verify a provider’s renewable energy claims?
Ask whether the claim is backed by a power purchase agreement that funds new generation or by unbundled certificates that offset existing use, and ask whether there is additionality. The two are not equivalent.
Is PUE a reliable way to compare facilities?
Only with caution. PUE can be skewed by where the measurement boundary is drawn and by local climate, and it says nothing about the energy source. Use it as one input, not the verdict.
Should I worry about data center water usage?
It depends on the region. Water use is negligible globally but can matter in water-stressed areas. In those regions, ask about closed-loop cooling and water usage effectiveness.
The Bottom Line
Sustainability is no longer a separate conversation from cost and reliability. It is a lens on how well a facility is actually run. The PUE answer, the sourcing answer, and the billing structure tell you more about an operator’s discipline than any pledge or logo will.
At Data Canopy, we treat efficiency as an operating discipline because it serves both our clients and the environment, and because our access to top-tier facilities lets us pass wholesale-scale efficiency down to right-sized retail deployments. If you are evaluating colocation this year, start with the PUE question, ask how it was measured, and let the quality of the answer tell you about everything else.
