How KPMG's Data Center Index Redefines Cost Structures

For years, data center investors have faced a fundamental problem: no standardized way to compare costs across markets, asset types, or operational models. Everyone had their own spreadsheets, their own assumptions, and their own definitions of what counts as capital expenditure versus operational overhead. KPMG's Data Centre Benchmarking Index is a direct response to that problem — and for anyone deploying serious capital into digital infrastructure, it changes how the numbers must be read.

What the KPMG Index Actually Measures

KPMG's benchmarking index sits at the intersection of M&A advisory, operations analysis, and regulatory risk — which tells you something important about who it's designed for. This isn't a white paper for engineers optimizing cooling systems. It's a strategic instrument for CFOs, institutional investors, and infrastructure developers trying to understand where their cost structures stand relative to the global market.

The index creates a common language for CapEx and OpEx across the data center sector — something the industry has lacked at scale.

At its core, the framework separates capital expenditure (the physical buildout: land, power infrastructure, mechanical and electrical systems, structural shell) from operational expenditure (power consumption, staffing, maintenance contracts, cooling management, software licensing). That distinction matters enormously in practice. A hyperscale campus in Northern Virginia and a colocation facility in Frankfurt may carry similar headline valuations, but their cost profiles are structurally different — conflating the two leads to mispriced assets and flawed underwriting.

By establishing benchmarks across global markets, KPMG gives institutional stakeholders a way to stress-test assumptions against real-world comparables rather than in-house projections.

The CapEx Picture: Where the Money Is Actually Going

Capital expenditure in data centers has never been simple, but the last three years have made it dramatically more complex. Power procurement — not land, not construction — has become the defining constraint and cost driver for large-scale builds.

In primary markets like Northern Virginia, Singapore, and London, power availability directly determines where development can happen at all. Utilities in some of these markets are quoting interconnection timelines stretching three to five years. That delay carries a capital cost that traditional CapEx models didn't adequately price in: carrying costs on land, pre-development expenditures, and the opportunity cost of delayed revenue generation.

CapEx benchmarking that doesn't account for power procurement timelines is already obsolete.

The KPMG index addresses this by folding in the full development lifecycle, not just hard construction costs. For investors, this reframes the question from "how much does it cost to build?" to "how much does it cost to build, and when will it start generating returns?" Those are very different calculations when a grid interconnection delay can push your commercial operation date back by 18 months.

The index also highlights divergence between hyperscale builds and edge deployments. A hyperscale campus might achieve significant economies of scale on mechanical and electrical systems — but edge facilities, by their nature, can't replicate that efficiency. Treating them with the same CapEx assumptions produces systematically inaccurate underwriting.

OpEx: The Cost Nobody Watches Closely Enough

Capital expenditure gets the headlines. Operational expenditure quietly determines whether an asset performs over a 20-year hold period.

Power Usage Effectiveness (PUE) — the ratio of total facility power to IT equipment power — remains the dominant operational variable, and the benchmarking data makes that visible in a way that anecdotal reporting never could. A facility running a PUE of 1.8 in a market where best-in-class operators are achieving 1.3 isn't just less efficient: it's carrying a structural cost disadvantage that compounds annually. At scale, the difference between a 1.8 and 1.3 PUE across a 100MW facility, assuming an average power cost of $0.07/kWh, translates to millions of dollars in annual operational drag.

Staffing models add another layer of complexity. Hyperscale operators increasingly rely on remote operations and automation to thin on-site headcount. Smaller colocation providers often can't make that same investment, which shows up directly in OpEx per kilowatt. The KPMG index captures this variance, giving investors a way to identify operational inefficiency that asset-level financials might obscure.

The gap between top-quartile and bottom-quartile operators on operational cost per kilowatt is wider than most investors assume — and it's the single biggest driver of long-term return divergence.

What This Means for Investment Strategy

The practical implication of a credible benchmarking index is that it raises the floor for due diligence. An investor who previously relied on a seller's pro forma can now compare those projections against independent market benchmarks. That's a structural shift in how negotiations happen.

For M&A specifically, the index becomes a tool for identifying mispriced assets. A facility with above-market OpEx but strong location attributes might be undervalued if an acquirer can see a clear path to operational improvement. Conversely, an asset with artificially compressed OpEx — achieved through deferred maintenance or underinvestment in redundancy — can look attractive until it doesn't. Benchmarking data makes that second scenario harder to hide.

Strategic implications extend beyond individual deals. Portfolio managers operating across multiple markets can use the index to identify where their assets are underperforming relative to regional benchmarks, prioritize capital improvement investments, and build more defensible return projections for LPs. That kind of systematic portfolio intelligence hasn't been easily available before.

Developers, too, have a new reference point for underwriting new builds. If a greenfield project in a secondary market is penciling at CapEx levels significantly above the regional benchmark, the index forces a sharper question: is that premium justified by demand fundamentals, or is it a cost structure problem that needs to be solved before the shovel hits the ground?

The Technologies Reshaping the Numbers

Benchmarking is always a snapshot of the present, but the KPMG index exists against a backdrop of technologies actively shifting what CapEx and OpEx look like in practice.

Liquid cooling is the most immediate disruptor. As GPU-dense AI workloads push rack densities beyond what air cooling can handle — deployments at 50kW, 100kW, or higher per rack are increasingly common — the CapEx associated with cooling infrastructure changes fundamentally. Liquid cooling systems carry higher upfront installation costs but substantially better PUE outcomes. Any benchmarking framework will need to account for this bifurcation between traditional air-cooled facilities and next-generation liquid-cooled deployments.

Battery energy storage is also reshaping the OpEx model. Operators integrating BESS with renewable procurement can flatten energy costs and reduce exposure to peak pricing — but BESS carries its own capital cost and lifecycle management burden. Facilities optimizing around storage-backed renewables will look different on both sides of the CapEx/OpEx ledger.

AI itself is creating a new category of workload intensity that didn't exist at scale five years ago. Training clusters running continuously at high utilization stress infrastructure in ways that traditional enterprise colocation never did. The benchmarks that hold today may need material adjustment within three to five years as AI-optimized facilities become a distinct asset subclass with their own cost profile.

The Shift Already Happening

Data center benchmarking used to be a largely internal exercise — operators comparing themselves against their own historical performance or against informal peer conversations. KPMG's index moves that exercise into the open, and the implications compound as more market participants adopt it as a reference standard.

Investors who incorporate this framework early gain a durable analytical edge: better deal sourcing, more precise underwriting, and the ability to spot operational improvement opportunities that competitors miss. Those who ignore it are essentially choosing to operate with less information in a market where the competition is getting smarter.

The cost structures of digital infrastructure are not standing still. Neither can the frameworks used to evaluate them.

[INTERNAL LINK: KPMG Index Overview]

[INTERNAL LINK: CapEx vs. OpEx in Data Centers]

[INTERNAL LINK: Investment Strategies in Digital Infrastructure]

Ready to redefine your investment strategy? Explore the KPMG Data Centre Benchmarking Index and gain insights that can transform your approach to digital infrastructure. Visit [InfraSale Marketplace](https://infrasale.com/marketplace) today!