Most organisations discover their IT asset ESG metrics the same way: an auditor, a customer questionnaire, or a regulator asks for a number that nobody has ever calculated, and the answer has to be assembled in three weeks from purchase orders, a spreadsheet of serial numbers, and a PDF from whoever collected the last pallet of retired laptops. The number gets produced. It does not survive contact with anyone who checks it.
That improvisation is running out of road. Under the EU’s Corporate Sustainability Reporting Directive, California’s SB 253, and the Scope 3 expectations already embedded in customer procurement questionnaires, the sustainability performance of your hardware estate has become a disclosed, assured, and in some jurisdictions legally consequential figure. The IT function is now a data supplier to the finance function, and the ESG metrics it hands over will be read by people whose job is to find the seam where the numbers stop being real.
The good news is that the ESG metrics that matter are finite and well defined. The bad news is that the ones most commonly reported — landfill diversion percentage, tonnes recycled, and the “carbon avoided” figure on the back page of a disposition certificate — are the weakest of the set, and at least one of them is something the GHG Protocol explicitly tells you not to put in your inventory. Teams that build their reporting around those three numbers end up with a dashboard that looks excellent and an audit position that is indefensible.
This article sets out the ESG metrics worth tracking across the full IT asset lifecycle — procurement, deployment, operation, utilisation, and disposition — what each one actually measures, where the data comes from, which regulatory disclosure it feeds, and which widely used numbers you should quietly stop reporting. It covers the environmental, social, and governance columns rather than treating ESG as a synonym for carbon, and it is written for the people who have to produce the figures rather than the people who present them.
ESG Metrics for the IT Asset Lifecycle: The Quick Answer
The short version is that IT asset ESG metrics fall into five lifecycle stages, and the leverage is distributed very unevenly across them. Procurement decisions lock in roughly four-fifths of a device’s total emissions before it is ever switched on. Operation is the stage everyone measures because the data is easy. Disposition is the stage everyone reports because the certificates are glossy. Utilisation and lifespan are where the actual carbon lives and where almost nobody has a defensible number.
If you are starting from nothing, the highest-value ESG metrics to stand up first are embodied carbon per device class, average asset lifespan by class, reuse rate at disposition, and a chain-of-custody completeness percentage. Those four are auditable, they map cleanly to disclosure frameworks, and they are the ones that change behaviour rather than describe it.
| Lifecycle stage | Primary ESG metrics | Where the data comes from | Disclosure it feeds |
|---|---|---|---|
| Procurement | Embodied carbon per unit, ecolabel coverage, recycled content share, supplier assurance rate | Vendor PCF datasheets, EPEAT/TCO registries, purchase orders | Scope 3 Category 1, ESRS E5 inflows |
| Deployment | Devices per employee, provisioning waste, packaging mass, freight emissions | Asset register, logistics records | Scope 3 Categories 4 and 5 |
| Operation | Endpoint energy per device, data centre PUE, WUE, ERF, REF, renewable share | Power meters, DCIM, EED reporting, utility bills | Scope 2, EU Energy Efficiency Directive |
| Utilisation | Average asset lifespan, idle asset ratio, utilisation rate, refresh deferral rate | CMDB, telemetry, HR joiner-leaver data | ESRS E5 resource use, internal targets |
| Disposition | Reuse rate, recycling rate, landfill diversion, chain-of-custody completeness, certification coverage | ITAD settlement reports, R2v3/e-Stewards audits, mass balance | Scope 3 Category 5, GRI 306, ESRS E5 outflows |
| Cross-cutting | Data sanitisation verification rate, supplier labour audit coverage, data ownership and assurance level | Sanitisation logs, supplier programmes, internal controls | ESRS S2 and G1, SB 261, assurance readiness |
The table is deliberately organised around where the number originates rather than where it is presented. That is the single most useful reframe available to an IT team being asked for ESG metrics for the first time: you are not being asked to write a sustainability report, you are being asked to become a reliable upstream data source for one.
Why IT Asset ESG Metrics Became a Reporting Obligation
For most of the last decade, IT sustainability reporting was voluntary, qualitative, and read by nobody. The change is not cultural. It is statutory, and it arrived on three fronts more or less simultaneously.
The first is the EU’s Corporate Sustainability Reporting Directive and its European Sustainability Reporting Standards, which put resource use, waste, and circularity on a mandatory footing through ESRS E5. Even after the Omnibus simplification package materially reduced the number of mandatory data points, the underlying obligation to trace material inflows, outflows, and waste treatment for material topics remains, and for an organisation whose largest physical purchase category is electronics, hardware is difficult to argue out of materiality.
The second is California. SB 253 requires large companies doing business in the state to disclose Scope 1, Scope 2, and eventually Scope 3 emissions, with the California Air Resources Board administering the regime and Scope 3 reporting arriving in 2027 for the prior fiscal year. Purchased IT hardware and its end-of-life treatment both sit squarely inside Scope 3, which means the ESG metrics behind them stop being marketing and start being a filing.
The third front is commercial rather than legal. Enterprise procurement questionnaires now routinely ask for supplier-level emissions data, and a supplier that cannot answer is increasingly scored down rather than merely noted. The practical effect is that companies far below any statutory reporting threshold are still being asked for ESG metrics by customers who are above it.
None of this makes the underlying measurement problem easier. It just removes the option of not doing it. The organisations handling the transition well are the ones that treated it as a data engineering problem in 2024 and 2025 rather than a communications problem, which is the same lesson enterprises learned the hard way when Scope 3 emissions reporting first landed on IT departments as a set of nine unwelcome steps.
The Five Lifecycle Stages Your ESG Metrics Have to Cover
A useful lifecycle model for IT assets has five stages, and the reason to be explicit about them is that the ESG metrics for each stage have different owners, different data sources, and different failure modes.
Procurement covers specification, supplier selection, and purchase. Its ESG metrics are dominated by embodied carbon and supplier assurance, and its data comes from outside your organisation entirely, which makes verification the central difficulty.
Deployment covers imaging, shipping, and issuing the asset to a user or a rack. Its ESG metrics are small in absolute terms — packaging, freight, provisioning waste — but they are the easiest to collect accurately, which makes them a reasonable place to prove your pipeline works before tackling harder numbers.
Operation covers the years the asset spends drawing power. This is where operational energy ESG metrics live, and it is the only stage where most organisations already have instrumentation. It is also where the greatest divergence exists between endpoint fleets, where operational energy is a minority of lifetime carbon, and data centre estates, where it is the overwhelming majority.
Utilisation is the stage that the classic five-stage model omits and that matters most. It asks not how much energy the asset used but whether the asset needed to exist. Its ESG metrics are lifespan, idle ratio, and utilisation, and they have more carbon leverage than anything else on the list.
Disposition covers retirement, data sanitisation, resale, reuse, recycling, and final disposal. Its ESG metrics are the most heavily certified and the most frequently misreported, for reasons the later sections of this article go into in detail.
Stage One: Procurement ESG Metrics
Procurement is where the largest share of your IT carbon is decided and where your ESG metrics are most dependent on other people’s disclosures.
The first of the procurement ESG metrics is embodied carbon per unit, expressed in kilograms of CO2 equivalent and sourced from the manufacturer’s product carbon footprint datasheet. The second is ecolabel coverage: the proportion of units purchased in a period that carry a recognised registration such as EPEAT or TCO Certified. The third is recycled content share, which the ESRS resource-inflow disclosures ask for directly and which most vendors now publish per product family. The fourth is supplier assurance rate, meaning the proportion of your hardware spend covered by suppliers who publish audited emissions data rather than estimates.
Those four are ordered by how hard they are to game. Recycled content and ecolabel coverage are verifiable from public registries. Embodied carbon depends on the vendor’s methodology and is comparable within a brand but only loosely comparable across brands. Supplier assurance rate is the one that tells you how much of the rest you should trust.
A procurement ESG metrics programme that stops at “percentage of spend with EPEAT-registered products” is a reasonable start and a poor destination. The registries were designed to steer purchasing toward better products, not to quantify a footprint, and treating registration coverage as a carbon number conflates a policy signal with a measurement.
The most common structural mistake at this stage is measuring only what was bought. The ESG metrics that matter also include what was not bought — the refresh cycles deferred, the redeployments that displaced a purchase, and the specification downgrades that avoided over-provisioning. Those show up as an absence in the procurement data, which means they have to be captured deliberately or they vanish.
Embodied Carbon: The Number Most Estates Never Calculate
Embodied carbon is the emissions associated with extracting materials, manufacturing components, assembling the device, and transporting it to you. For endpoint hardware it dominates everything else, and the ratio is not close.
Published product carbon footprints put manufacturing at roughly 80 percent of a laptop’s cradle-to-grave emissions, with the use phase accounting for most of the remainder. Vendor datasheets for individual models vary widely — a lightweight business laptop may sit near 200 kg CO2e while a high-specification mobile workstation runs several times higher — and independent benchmarking of published datasheets puts the median around the 215 to 230 kg CO2e range with a wide distribution around it.
The implication for your ESG metrics is uncomfortable. If four-fifths of the footprint is fixed at the moment of purchase, then every energy-efficiency initiative aimed at the endpoint fleet is operating on a fifth of the problem, and the sustainability programme that reports a 12 percent reduction in device power draw has moved roughly two percent of lifetime device carbon. Meanwhile the decision to refresh a year earlier than necessary quietly added an entire manufacturing footprint per device.
Servers invert the ratio. A rack-mounted server running continuously for five years in a facility with mediocre power efficiency will burn through its embodied carbon several times over, which is why data centre ESG metrics are legitimately energy-led while endpoint ESG metrics are legitimately procurement-led. Applying one framing to the other estate is the most common analytical error in this field, and the reason so many corporate digital carbon footprint programmes optimise the wrong half of their hardware.
To make embodied carbon usable rather than decorative, hold it in your ESG metrics as a per-device-class figure with a documented source, refresh it annually as vendors publish new datasheets, and multiply it by units purchased rather than units owned. That last distinction matters: embodied carbon is recognised in the year of acquisition under Scope 3 Category 1, not amortised across the service life, however much more intuitive amortisation feels.
Supplier and Ecolabel ESG Metrics
Ecolabels are the cheapest verification mechanism available to a procurement team, and their value in an ESG metrics programme is exactly that: verification, not quantification.
EPEAT, run by the Global Electronics Council, applies criteria across climate, circularity, chemicals of concern, and corporate ESG performance, with a climate criteria tier that all registered products were required to meet from the end of 2025 and a Climate+ designation for products meeting a substantial share of the optional mitigation criteria. TCO Certified takes a comparable four-pillar approach across climate, substances, circularity, and supply chain, with all criteria mandatory and independent verification required rather than self-declared.
The useful ESG metrics here are coverage ratios rather than counts: the proportion of units, and separately the proportion of spend, covered by a registration in each period. Reporting units and spend separately matters because a fleet can be 90 percent registered by unit and 55 percent by spend if the expensive specialist hardware is exactly the category with no registered options, which is a real finding rather than a rounding error.
Supplier-level ESG metrics sit alongside these. The ones worth tracking are the share of hardware spend with suppliers who have a validated science-based target, the share covered by third-party assured emissions disclosure, and the share subject to a supply chain labour audit within the last two years. The third belongs to the social column and is discussed later, but it is collected through the same procurement process and should be built into the same data flow rather than bolted on when someone asks.
There is a limit worth naming. Ecolabel criteria are product-level and threshold-based. A device either qualifies or it does not, which means these ESG metrics saturate: once you are at 95 percent coverage the figure stops carrying information, and continuing to headline it is a sign the programme has stopped developing.
Stage Two: Deployment and Provisioning ESG Metrics
Deployment ESG metrics are individually small and collectively worth tracking, mostly because they are the easiest place to prove that your data pipeline works end to end before you attempt harder numbers.
The deployment ESG metrics that earn their place are devices issued per employee, which detects the quiet accumulation of second and third machines; packaging mass per shipment, which vendors will report if asked and which feeds Scope 3 Category 5; and upstream transport emissions, which sit in Category 4 and are calculable from freight mode and distance without any new instrumentation.
Devices per employee is the sleeper metric in this group. An estate at 1.4 devices per employee is carrying roughly 40 percent more embodied carbon than one at 1.0 for the same headcount, and the excess is invisible in every one of the energy ESG metrics because the surplus machines are typically the ones sitting in drawers. It is also a cost metric, which makes it unusually easy to get funded — the same dynamic that makes IT audits that uncover hidden software costs an easier sell than a pure sustainability initiative.
Provisioning waste is worth a mention among the deployment ESG metrics and rarely worth a dashboard. Imaging, accessories, and consumables are real but marginal, and a programme that spends a quarter chasing them while embodied carbon and lifespan go unmeasured has confused effort with impact.
Stage Three: Operational Energy ESG Metrics for Endpoints
Endpoint energy consumption is the most measured and least consequential of the environmental ESG metrics, and it is worth being clear about why it still belongs on the list.
The direct case is weak. A business laptop drawing an average of 30 to 50 watts under load and far less at idle contributes perhaps 20 percent of its lifetime emissions through the socket, and that fraction shrinks further in grids with high renewable penetration. Halving it is difficult and moves the total by a tenth.
The indirect case is stronger. Endpoint energy telemetry is the same telemetry that tells you which devices are being used, which are idle, and which have not woken up in four months. The ESG metrics that come out of the utilisation stage — idle ratio, effective utilisation, redeployment candidates — are derived from the same collection infrastructure, so building it for energy reporting is a reasonable way to fund the data source that actually matters.
If you do report endpoint energy, report it as measured kilowatt-hours per device class per year rather than as a modelled figure derived from thermal design power. Modelled endpoint energy is close to meaningless, because real duty cycles bear almost no relationship to nameplate ratings, and an auditor who understands the difference will discount the number entirely.
The location-based versus market-based distinction applies here exactly as it does elsewhere in Scope 2. Report both. An estate that looks decarbonised on a market-based basis and unchanged on a location-based one has purchased attributes rather than reduced consumption, and the gap between the two figures is itself one of the more informative ESG metrics available.
Data Centre ESG Metrics: PUE, WUE, ERF and REF
Owned and colocated data centre space is the one part of the IT estate where standardised operational ESG metrics already exist, are widely instrumented, and in the EU are now mandatory to report.
Commission Delegated Regulation (EU) 2024/1364 established a reporting framework requiring facilities at or above 500 kW of installed IT power demand to report annually across roughly two dozen data points, including four headline sustainability indicators: Power Usage Effectiveness, Water Usage Effectiveness, Energy Reuse Factor, and Renewable Energy Factor. The European Commission’s rating scheme work builds a labelling regime on top of that data, with the first label cycle expected to draw on 2026 calendar-year reporting.
PUE is total facility energy divided by IT equipment energy, so 1.0 is a theoretical floor and anything below about 1.2 represents a genuinely efficient modern facility. WUE expresses litres of water consumed per kilowatt-hour of IT energy and has become the more contested metric as data centre siting collides with water stress. ERF captures the share of energy recovered and reused elsewhere, typically as district heat, and is close to zero in most estates outside northern Europe. REF is the renewable share of consumed energy.
For an enterprise IT team, the practical question is not how to compute these but how to obtain them from a colocation provider under contract. The answer is to write them into the agreement, because a provider that is not contractually obliged to report facility-level indicators will supply them annually, in a slide, at a granularity that cannot be audited. Organisations that discovered this late are the same ones now discovering that power availability constraints have turned facility efficiency from a sustainability line item into a capacity-planning dependency.
Why PUE Alone Is a Misleading ESG Metric
PUE deserves specific criticism because it is the most quoted number in data centre sustainability and one of the easiest to improve without reducing any emissions at all.
The mechanism is arithmetic. PUE is a ratio with IT load in the denominator, so raising IT load improves PUE even if total energy rises. A facility that consolidates workloads onto fewer, harder-worked servers will typically see PUE worsen slightly while absolute consumption falls, and a facility that runs a fleet of idle machines at high utilisation of the electrical plant will post an excellent PUE while wasting most of the energy it draws. Optimising the ratio and optimising the outcome are different projects.
PUE also says nothing about what the IT equipment achieved with the energy. Two facilities with identical PUE can differ by an order of magnitude in useful work delivered per kilowatt-hour, which is why the more informative operational ESG metrics pair infrastructure efficiency with a workload-normalised denominator — transactions, inferences, or delivered compute per unit of energy. That is the argument behind treating performance per watt as the decisive infrastructure metric rather than facility overhead, and it applies with particular force to AI estates where accelerator utilisation swings enormously.
The defensible position is to report PUE because it is required and comparable, and to report at least one workload-normalised efficiency metric alongside it because PUE alone will mislead anyone who reads it without that context. The Green Software Foundation’s Software Carbon Intensity specification is one available formalisation of that second metric, and its main virtue in an ESG metrics programme is that it forces an explicit functional unit rather than allowing a floating one.
Cloud and Colocation: The ESG Metrics You Cannot Collect Yourself
A large and growing share of the IT estate runs on infrastructure you do not own, cannot meter, and are not permitted to inspect. This is the structural weak point in every IT asset ESG metrics programme, and pretending otherwise is the fastest route to a qualified assurance opinion.
Hyperscale providers publish customer-facing carbon tools, and those tools are genuinely useful for trend analysis within a single provider. They are not comparable across providers, they rely on allocation methodologies that differ in material ways, they lag by one to three months, and they overwhelmingly report market-based figures that reflect the provider’s renewable procurement rather than the grid your workload actually ran on. Treating a provider’s dashboard export as audited ESG metrics is a category error.
The workable approach is to record three things: the provider-reported figure, the methodology version it was produced under, and your own independently modelled estimate based on consumption and regional grid intensity. Where the two diverge, the divergence is the disclosure. Assurance providers respond far better to a documented gap than to a single confident number with no provenance.
There is a governance dimension too. Cloud consumption is procured by engineering teams under a spend model that decouples the buying decision from any environmental review, which means the fastest-growing part of your footprint is the part with the weakest controls. That mismatch has been visible in the sector for a while, most publicly in the way the AI buildout has strained the climate commitments of the largest cloud providers themselves.
Stage Four: Utilisation ESG Metrics and the Cost of Idle Assets
Utilisation is where an IT asset ESG metrics programme stops describing and starts changing things, because it is the only stage where the intervention is not to run the asset better but to need fewer assets.
The core utilisation ESG metrics start with the idle asset ratio: the proportion of registered devices that have not been used within a defined window. Thirty days is aggressive, ninety days is defensible, and both are worth tracking because the gap between them tells you whether you have a redeployment problem or an accounting one. Most estates that measure this for the first time find between five and fifteen percent of endpoints have not checked in for a quarter, and the embodied carbon attached to those units is already spent.
The second of the utilisation ESG metrics is effective utilisation for shared infrastructure — average and peak server CPU and memory utilisation, storage allocated versus consumed, and virtual machine density. A virtualised estate running at 15 percent average utilisation is carrying roughly six times the hardware it needs, and every one of those chassis has an embodied footprint and a replacement cycle attached. Consolidation programmes such as VMware estate modernisation onto native cloud services usually produce a large one-off improvement here, which is worth capturing as an ESG metric rather than only as a licensing saving.
The third is asset register accuracy, expressed as the percentage of devices in the CMDB that can be confirmed present, assigned, and in a known state. This is not conventionally treated as a sustainability metric and it should be, because every one of the downstream ESG metrics in the disposition stage is computed against a denominator that this number determines. A programme reporting a 94 percent landfill diversion rate on an asset register that is 80 percent accurate is reporting a 94 percent figure about 80 percent of its estate, and the honest headline is neither number.
Asset Lifespan: Where IT Asset ESG Metrics Have the Most Carbon Leverage
If you track exactly one ESG metric across the IT asset lifecycle, track average asset lifespan by device class. Nothing else on the list moves as much carbon per unit of management effort.
The arithmetic follows directly from the embodied carbon share. If manufacturing is 80 percent of lifetime emissions and you extend service life from three years to four, you spread that fixed cost across a third more service, and published analysis puts the resulting reduction in annualised emissions at roughly 25 percent. Extending from four years to six, as TCO Certified has modelled, takes annualised emissions from about 74.7 kg CO2e to about 53.1 kg CO2e per notebook, a reduction near 29 percent, and longer extensions compound further.
The objection is always the same: users will not accept older hardware, and support costs rise. The first half is largely an untested assumption. Research from Atos on sustainable workplace practice found device lifespan could be roughly doubled without a measurable degradation in user experience, provided refresh was condition-based rather than calendar-based. The second half is real but quantifiable, and in most estates the incremental support cost of years four and five is well below the capital and carbon cost of replacement.
Track lifespan three ways, because each answers a different question. Average age of the deployed fleet tells you where you are now. Average age at retirement tells you what your policy actually produces, which is frequently shorter than the policy states. Refresh deferral rate — the proportion of devices eligible for replacement that were retained — tells you whether the programme is working this quarter rather than three years from now.
Calendar-based refresh is the single most carbon-expensive habit in corporate IT, and it survives mostly because it is administratively simple. Condition-based refresh requires telemetry, a defined health threshold, and someone empowered to say no to a replacement request. All three are cheaper than the hardware.
Repair, Reuse and Redeployment ESG Metrics
Between operation and disposition sits a set of interventions that most asset registers cannot see, which is why the ESG metrics covering them are usually absent.
Repair rate is the proportion of hardware faults resolved by component replacement rather than whole-unit swap. It is a genuine circularity indicator and one of the few reuse ESG metrics collectable from service desk data with modest effort, provided the ticket taxonomy distinguishes the two outcomes — which, in most service management configurations, it does not by default.
Internal redeployment rate is the proportion of returned devices reissued inside the organisation rather than sold or recycled. This is the highest-value circular outcome your ESG metrics can record, because it displaces a purchase one-for-one and the displaced embodied carbon is directly attributable. It is also the metric most likely to be undercounted, because redeployment is often handled informally by local IT teams and never touches the disposition process at all.
Spare parts availability and mean time to repair matter as leading indicators. An estate with a thin parts inventory will replace units it could have repaired, and the ESG metrics will show it as a disposition outcome rather than a procurement failure. Tracking the reason code for every retirement — end of support, physical damage, performance, user request, unrepairable — is what makes that distinction visible.
The broader point is that repair and reuse ESG metrics measure organisational capability, not environmental outcome. They tell you whether the option existed, which is why they belong in the management commentary alongside the outcome figures rather than being reported as a footprint reduction on their own.
Stage Five: Disposition ESG Metrics
Disposition is the most heavily instrumented and most systematically misreported stage of the IT asset lifecycle, and the reason is that the party generating the data is usually the party being paid for the outcome.
The ESG metrics that matter at disposition are, in descending order of usefulness: reuse rate by unit and by mass, recycling rate by mass with the processing route named, landfill and incineration rate, chain-of-custody completeness, and certification coverage of every downstream processor. Weight recovered and tonnes processed are volume statistics rather than performance ESG metrics and belong in an appendix.
The load-bearing distinction is between reuse and recycling, and it is not a technicality. A device that is wiped, refurbished, and resold displaces the manufacture of a replacement device and therefore avoids an entire embodied footprint somewhere in the economy. A device that is shredded into commodity streams recovers a fraction of its material value, consumes energy to do so, and displaces primary material extraction at a far lower rate. Both are better than landfill. They are not remotely equivalent, and an ESG metrics dashboard that adds them together into a single “diverted” percentage has erased the only distinction that matters.
Global context is worth stating plainly here. The Global E-waste Monitor 2024 recorded 62 million tonnes of electronic waste generated in 2022, up 82 percent from 2010, with only 22.3 percent documented as formally collected and recycled — and it projects that rate falling toward 20 percent by 2030 as generation outpaces collection. Against that backdrop, an organisation reporting a 98 percent diversion rate is either genuinely exceptional or measuring something narrower than it sounds. In practice it is almost always the latter, because the denominator is what reached the ITAD vendor rather than what left the estate.
Landfill Diversion Rate and Why It Flatters You
Landfill diversion is the most reported disposition ESG metric and the least informative, and its weakness is entirely in the denominator.
The figure is calculated as the mass of material that did not go to landfill divided by the mass that entered the disposition process. Everything that never entered the process — devices lost, stolen, kept by departing employees, sold informally, or sitting in a cupboard in a regional office — is invisible. So is everything that entered the process, was exported, and was disposed of somewhere your reporting does not reach.
The result is one of the few ESG metrics that improves when your asset tracking degrades. A company that processes only the hardware it can easily account for will post a better diversion rate than one that hunts down every stray unit, which is precisely backwards as an incentive.
There is a second problem. Diversion counts mass, and mass in IT hardware is dominated by steel, aluminium, glass, and plastics. The environmentally significant fraction — the precious metals and critical raw materials on the boards — is a small share of weight and a large share of impact. A process that recovers 95 percent of mass while sending boards to a low-grade smelter is scoring well on diversion and poorly on everything that made the recovery worth doing.
Keep it among your ESG metrics, because it is expected and comparable. Report it with the denominator stated explicitly — mass received by the disposition process, not mass retired from the estate — and report the reconciliation between the two alongside it. That reconciliation gap tells you more than any of the diversion ESG metrics on their own, and it is the number an assurance provider will ask for first.
Reuse Rate Versus Recycling Rate: The Distinction Auditors Look For
Reuse rate is the entry among disposition ESG metrics that best predicts whether a disposition programme is doing environmental work or waste management, and it should be reported two ways.
By unit, reuse rate is the proportion of retired devices that entered a second service life, whether internally redeployed, sold to a refurbisher, or donated. By mass, it is the same calculation weighted by device weight. The two diverge sharply when the fleet contains a small number of heavy items — servers, storage arrays, network chassis — that are more likely to be recycled than resold, and reporting both is what prevents a good laptop resale programme from masking a poor infrastructure outcome.
Mature disposition programmes routinely resell 60 to 80 percent of retired endpoints by unit when the fleet is refreshed on a three to four year cycle and the hardware was mid-range or better at purchase. That number collapses as refresh cycles lengthen, which produces one of the genuinely difficult trade-offs in this field: extending asset lifespan reduces total emissions substantially while simultaneously reducing your reuse rate and your resale recovery. An ESG metrics programme that rewards reuse rate without accounting for lifespan will push the organisation toward shorter refresh cycles, which is the opposite of the intended outcome.
The resolution is to report reuse rate as a quality measure of the disposition process and lifespan as the primary carbon measure, and to state the interaction explicitly in the commentary rather than letting two ESG metrics quietly fight each other on the same dashboard.
Recycling rate should always name the route. Material sent to a certified processor for materials recovery, material sent for energy recovery, and material sent for disposal are three different outcomes, and the GRI 306 waste disclosures ask you to separate them. Aggregating them into “recycled” is a reporting choice that will not survive review.
The Avoided Emissions Trap in ITAD Reporting
This section covers the single most common serious error in IT asset ESG metrics, and it is worth being blunt about it: the carbon avoidance figure on your disposition certificate almost certainly cannot go in your emissions inventory.
Disposition vendors routinely supply a report quantifying emissions avoided through reuse and materials recovery. The number is usually large, always favourable, and computed on a methodology that varies between vendors. It is presented in a format that looks exactly like an inventory figure, and it is frequently folded into corporate sustainability reporting as a reduction.
The GHG Protocol’s technical guidance is explicit on this point. In the guidance covering Scope 3 Category 5, waste generated in operations, and correspondingly in Category 12, end-of-life treatment of sold products, companies are instructed not to report negative or avoided emissions associated with recycling. The reasoning is straightforward: your inventory measures emissions attributable to your activities, and a credit for displacing someone else’s future production is a different accounting system.
The Science Based Targets initiative takes the same position from the target-setting side. Its guidance treats avoided emissions as sitting outside the corporate inventory entirely, so they do not count toward near-term or net-zero target achievement. A company that has been reporting ITAD avoided emissions as progress against a validated target has been reporting something the standard does not recognise.
None of this means the avoidance figure is worthless. It is a legitimate way to describe the value of a reuse programme, and it belongs in the narrative section of a sustainability report clearly labelled as avoided emissions outside the inventory boundary. What it cannot do is reduce your reported Scope 3, appear in the same table as inventory figures, or count against a science-based target. The ESG metrics that do belong in the inventory for this stage are the actual emissions from transport, processing, and final treatment of your retired assets — which are positive numbers, are considerably smaller, and are far less pleasant to put on a slide.
Teams that make this correction early take a one-time hit to their reported performance and gain an audit position that holds. Teams that leave it are carrying a restatement risk that grows with every year of published ESG metrics.
Certification and Chain-of-Custody ESG Metrics
Certification coverage is a governance metric wearing environmental clothing, and it is one of the few disposition ESG metrics that a third party will verify on your behalf.
The two dominant standards are R2v3, maintained by Sustainable Electronics Recycling International, and e-Stewards, maintained by the Basel Action Network. R2v3 uses a risk-based framework permitting controlled export to qualified downstream facilities under documented accountability, and has by far the wider adoption, with several hundred more certified facilities than its counterpart. e-Stewards is stricter on export, prohibiting shipment of electronics to developing countries regardless of working condition, and additionally requires NAID AAA data destruction certification and an ISO 14001 or equivalent environmental management system.
Which you require depends on your risk posture rather than on which is objectively better. The ESG metrics to track are the proportion of disposition volume handled by certified facilities, the number of downstream tiers you have documented, and the proportion of volume for which you hold a settlement report reconciling every serial number received to a final outcome.
That last one is chain-of-custody completeness, and it is one of the disposition ESG metrics that distinguishes a real programme from a paper one. A certificate of destruction covering a pallet is not chain of custody. Serial-level reconciliation from the point of collection through to resale, materials recovery, or destruction, with each transfer documented, is. Most organisations discover on first measurement that they have serial-level tracking for 60 to 75 percent of retired assets and a mass-level estimate for the rest.
The downstream question is the one that causes actual incidents. Your vendor’s certification covers your vendor. The processors your vendor sells to, and the processors they sell to, are where uncontrolled export and informal disposal occur, and the only defence is a documented downstream chain with named facilities. Ask for tier-two and tier-three visibility explicitly, because it will not be volunteered.
Data Sanitisation ESG Metrics Belong in the Report Too
Data sanitisation is usually filed under security, and it belongs in your IT asset ESG metrics because it is the governance control that makes the environmental outcome possible.
The connection is mechanical. Reuse requires that a device can leave your control with confidence, which requires verified sanitisation. Where that confidence is absent, the organisation defaults to physical destruction, which forecloses reuse entirely and converts a resaleable asset into shredded commodity. Every percentage point of drives destroyed rather than sanitised is a percentage point of reuse rate that was never available.
The ESG metrics to track are the proportion of storage devices sanitised to a recognised standard such as NIST SP 800-88 with a verification record, the proportion physically destroyed, and the reason code where destruction was chosen. A well-run programme destroys drives when policy, encryption state, or drive failure genuinely requires it and sanitises the rest. A poorly run one destroys everything because nobody wants to own the decision.
There is a compliance dimension that raises the stakes on the governance side. Improper decommissioning of hardware carrying regulated data has produced substantial enforcement action, and the compliance exposure from mishandled hardware decommissioning is typically larger than the entire commercial value of the assets in question. That asymmetry is why the destruction default persists, and why the fix is a verifiable sanitisation process rather than an exhortation to destroy less.
Report sanitisation verification rate alongside reuse rate. The two move together, and showing them together is what explains a reuse figure to anyone who asks why it is not higher.
Social ESG Metrics Across the IT Asset Lifecycle
The S in ESG is the column IT teams skip, and the IT asset lifecycle happens to be one of the places where social risk is most concentrated and most documented.
Upstream, the risks are mineral sourcing and manufacturing labour. The ESG metrics that map to this are the share of hardware spend covered by a supplier code of conduct with audit rights, the share covered by a conflict minerals declaration, and the number and severity of findings in supplier audits within the reporting period. Under the ESRS, workers in the value chain are a defined disclosure area, which means this stops being voluntary for in-scope organisations.
Downstream, the risk is informal recycling. Electronic waste that leaves the formal chain frequently ends up processed by hand in conditions involving open burning and acid leaching, with well-documented harm to workers and nearby communities. The social ESG metrics that speak to this are the same downstream chain-of-custody completeness described earlier, read as a social control rather than an environmental one. That reframing is often what gets it funded.
There is a positive social metric worth reporting where it applies: the volume of refurbished hardware donated or sold into digital inclusion programmes, tracked by unit and by recipient organisation. Unlike most social ESG metrics it is easy to verify, and it is one of the few places where the environmental and social columns produce the same number for the same activity.
The mistake to avoid is reporting a supplier code of conduct as if it were an outcome. Coverage of a policy is an input metric. Audit findings, remediation closure rates, and repeat-finding rates are outcome ESG metrics, and the gap between the two is usually where the real story is.
Governance ESG Metrics: Ownership, Assurance and Board Oversight
Governance ESG metrics are about whether the numbers can be trusted, which makes them the ESG metrics that determine the value of every other number in this article.
The first is data ownership coverage: the proportion of your reported ESG metrics that have a named owner, a documented calculation method, a defined source system, and a stated refresh frequency. Most programmes score badly here on first assessment, and the ones that score well are almost always the ones that ran a formal data lineage exercise rather than assembling figures on request.
The second is assurance level by metric. Limited assurance and reasonable assurance are different products at different prices, and disclosure regimes are phasing in requirements over time. Knowing which of your ESG metrics could survive limited assurance today, and which could not, is more useful than an aggregate readiness score.
The third is restatement rate — the number of previously published ESG metrics revised in the current period, and the magnitude of the revisions. A restatement rate of zero across several years of a maturing programme is a sign that errors are not being found rather than that they do not exist.
The fourth is board and executive oversight frequency, which sounds like the most box-ticking of the governance ESG metrics and is not, because the disclosure regimes ask for it directly. SB 261 and the TCFD-aligned frameworks it points to expect a description of governance over climate-related risk, and a description that cannot cite a cadence, a forum, and a decision record is a weak one.
The underlying reality is that IT asset ESG metrics fail on governance far more often than on measurement. The calculations are not hard. Sustaining them across a reorganisation, a vendor change, and a finance system migration is hard, and that is a governance problem with a governance solution.
Mapping Your ESG Metrics to CSRD and ESRS E5
For organisations in scope of the CSRD, the mapping from IT asset ESG metrics to the reporting standard is more direct than most IT teams expect.
ESRS E5 covers resource use and circular economy, and it is structured around three flows: resource inflows, resource outflows relating to products and services, and waste. Hardware procurement is an inflow, with the standard interested in the mass or volume of resources used and the share of recycled or secondary content within it. Retired hardware is an outflow and a waste stream, with the standard asking for total waste generated, the split between diverted and directed to disposal, the treatment routes applied, and the quantity of hazardous waste.
The Omnibus simplification adopted in early 2026 reduced mandatory data points substantially across the ESRS, with the revised delegated act cutting the count by a large margin and tightening the materiality filter, targeting application from FY2027. The effect on IT asset ESG metrics is that fewer data points are unconditionally required, not that the topic has become immaterial for an electronics-intensive organisation. Planning on the basis that hardware has been simplified out of scope is a bet against the materiality assessment your own auditors will run.
Practically, four of the ESG metrics discussed above carry most of the ESRS E5 load: recycled content share of purchased hardware, total mass of retired hardware, the treatment split across reuse, recycling, recovery, and disposal, and the hazardous fraction. If those four are auditable, the disclosure is largely writable. If they are not, no amount of narrative will fill the gap.
The climate standard ESRS E1 takes the emissions figures separately, which is where embodied carbon in Category 1 and disposition emissions in Category 5 land. Keeping the E1 and E5 data flows aligned to the same asset register is the single most useful structural decision available, and the most common cause of internally inconsistent reports when it is not done.
Mapping Your ESG Metrics to California SB 253 and SB 261
The Californian regime is narrower in subject matter than the CSRD and, for many US-headquartered organisations, closer in time.
SB 253 requires disclosure of Scope 1 and Scope 2 emissions on an earlier cycle, with Scope 3 following in 2027 covering the prior fiscal year, for entities above a billion dollars of revenue doing business in the state. The Scope 3 requirement is the one that pulls IT asset ESG metrics into a mandatory filing, because purchased hardware sits in Category 1 and retired hardware sits in Category 5, and both are calculable only from asset-level data that IT owns.
SB 261 addresses climate-related financial risk on a biennial basis for a lower revenue threshold, following TCFD-aligned reporting. Its relevance to IT assets is less about emissions and more about resilience and transition risk — supply chain concentration for critical hardware, exposure to regulated substances, and the cost trajectory of compliance itself. Enforcement timing has been affected by litigation, with the Air Resources Board indicating it would set an alternate date for the initial SB 261 cycle, so the operative planning assumption should be that the obligation persists and the date moves rather than the reverse.
The practical consequence for an IT organisation is a deadline structure rather than a new measurement problem. If your embodied carbon per device class, units purchased, retired mass, and treatment split are already produced monthly with documented lineage, both regimes are largely a formatting exercise. If they are produced annually by a person with a spreadsheet, they will not survive the assurance step, and the assurance step is where these regimes bite.
Assurance is the detail most teams underestimate. Limited assurance on Scope 1 and 2 arrives before Scope 3 assurance, but the effect of any assurance requirement is to make your calculation methodology, source data, and internal controls reviewable by someone with a professional obligation to be sceptical. Building ESG metrics that assume that reviewer exists is considerably cheaper than retrofitting them once she does.
Circularity ESG Metrics and ISO 59020
Circularity has its own measurement standard now, and it is worth knowing about even if you do not intend to certify against it.
ISO 59020:2024 sets requirements for measuring and assessing circularity performance, covering system boundaries, indicator selection, data collection, and interpretation, at levels ranging from product to organisation to inter-organisational system. Its practical value in an IT asset ESG metrics programme is that it forces the boundary question to be answered explicitly, which is the question most internal circularity dashboards silently dodge.
The headline indicator most often cited is circular material use rate, which measures the contribution of recovered material to total material use. Eurostat’s economy-wide version of this figure sat at 12.2 percent for the EU in 2024, which is a useful sense of scale: an organisation claiming a circularity rate several times that figure is either measuring a much narrower boundary or measuring something else.
For hardware specifically, the more tractable organisational ESG metrics are recycled content share on the inflow side, reuse rate and material recovery rate on the outflow side, and product lifetime extension expressed as the ratio of actual to design service life. Those three plus a stated boundary produce a defensible circularity picture without requiring a full ISO 59020 assessment.
The broader shift worth watching is the Ecodesign for Sustainable Products Regulation and its Digital Product Passport, which will eventually attach standardised material, repair, and recycling data to products themselves. Electronics and ICT are scheduled in later waves of the rollout, with obligations phasing in toward the end of the decade rather than immediately, but the direction is that a substantial share of the procurement ESG metrics currently assembled by hand will arrive as structured data attached to the product. Programmes designed to ingest that will age better than programmes designed around vendor PDFs.
Where IT Asset ESG Metrics Actually Come From
Every one of the ESG metrics in this article resolves to one of six source systems, and the reason most programmes stall is that nobody owns the joins between them.
The asset register or CMDB is the spine. It supplies unit counts, device classes, deployment dates, and retirement dates, and its accuracy caps the quality of everything downstream. Procurement and finance systems supply purchase dates, spend, supplier identity, and quantity, which is what converts a device class into an embodied carbon figure. Endpoint management and telemetry platforms supply last-seen dates, utilisation, and energy where instrumented. Facility and DCIM systems supply the data centre indicators. The disposition vendor supplies the settlement report, serial reconciliation, and treatment split. Vendor sustainability publications supply the emission factors.
The joins that break are consistently the same three. Purchase orders do not carry serial numbers, so linking spend to devices requires a reconciliation that is usually manual. Retirement dates in the asset register do not match collection dates in the ITAD report, producing period-boundary mismatches that make year-on-year comparison unreliable. And devices retired outside the formal process never appear in either system, which is the reconciliation gap discussed earlier.
The engineering answer is unglamorous: a single asset-level fact table keyed on serial number, with one row per device and columns for acquisition, cost, class, embodied carbon factor, deployment, last-seen, retirement, disposition route, and final outcome. Every one of the ESG metrics in this article is a query against that table. Organisations that build it stop having a reporting problem and start having a data quality problem, which is a considerably better problem.
The alternative — assembling each of the ESG metrics independently from whichever system is nearest when the request arrives — produces figures that cannot be reconciled with each other, which is exactly the condition that assurance is designed to detect.
Deployment Roadmap for IT Asset ESG Metrics
The following sequence assumes you are starting with an asset register of uncertain accuracy and no existing sustainability reporting, which describes most organisations honestly assessed.
Step 1: Establish the asset-level fact table
Before any metric, build the serial-keyed table described above and populate it from the asset register. Accept that it will be incomplete. Record the completeness percentage as the first of your ESG metrics, because every subsequent figure will be qualified by it and you will want the baseline.
Step 2: Attach embodied carbon factors by device class
Collect product carbon footprint datasheets for your top ten device models by volume, which will typically cover 70 to 85 percent of units. Assign a documented default to the remainder. Record the source and publication date of every factor, because the first assurance question will be where the number came from.
Step 3: Measure lifespan three ways
Compute average fleet age, average age at retirement, and refresh deferral rate from the dates already in the table. This requires no new data collection and produces the entry in your ESG metrics with the most carbon leverage, which makes it the fastest available demonstration that the programme is worth funding.
Step 4: Instrument the utilisation signal
Configure endpoint management to report last-seen dates and, where available, energy. Derive the idle asset ratio at thirty and ninety days. Expect the first result to be worse than anyone predicted, and plan the conversation before you run the query rather than after.
Step 5: Rebuild the disposition data contract
Renegotiate the ITAD reporting requirement to demand serial-level settlement reports, named downstream processors to at least tier two, treatment route per unit, and certification evidence. Most vendors can supply this and do not by default. Do this before you compute any of the disposition ESG metrics, because recomputing history is expensive.
Step 6: Separate reuse from recycling and correct the avoided emissions treatment
Split every disposition outcome into reuse, materials recovery, energy recovery, and disposal. Move any avoided emissions figure out of the inventory and into clearly labelled narrative. This is the step that costs you a headline number and buys you an audit position.
Step 7: Contract for facility and cloud indicators
Add PUE, WUE, ERF, and REF reporting obligations to colocation agreements at renewal, and establish a documented methodology for cloud emissions that records the provider figure, the methodology version, and your independent estimate. Treat the divergence as a disclosure rather than a problem to hide.
Step 8: Assign ownership and prepare for assurance
Give every one of your reported ESG metrics a named owner, a written calculation method, a source system, and a refresh cadence. Run an internal review as though it were a limited assurance engagement. The findings from that dry run are worth more than another quarter of metric development, and they arrive before the external reviewer does.
ESG Metrics That Matter
The following set is what a defensible IT asset ESG metrics programme reports, with the qualification each one requires. Anything not on this list is supporting detail rather than a headline.
| Metric | Definition | Target direction | Principal caveat |
|---|---|---|---|
| Embodied carbon per unit | kg CO2e from vendor PCF, by device class | Lower | Not comparable across vendors; recognised at purchase, not amortised |
| Average asset lifespan | Mean service life at retirement, by class | Higher | Fleet age and age at retirement differ; report both |
| Refresh deferral rate | Eligible devices retained rather than replaced | Higher | Leading indicator only; confirm against age at retirement |
| Idle asset ratio | Devices not seen in 30 and 90 days | Lower | Depends entirely on asset register accuracy |
| Devices per employee | Deployed units divided by headcount | Lower | Role mix distorts comparison across business units |
| Reuse rate | Retired units entering a second service life | Higher | Conflicts with lifespan; report the interaction |
| Recycling rate by route | Mass by materials recovery, energy recovery, disposal | Route-dependent | Never aggregate the three routes into one figure |
| Landfill diversion rate | Mass diverted over mass received | Higher | Denominator is mass received, not mass retired |
| Chain-of-custody completeness | Units reconciled serial-to-outcome | Higher | Distinguish serial-level from mass-level evidence |
| Sanitisation verification rate | Drives sanitised to standard with a record | Higher | Report destruction reason codes alongside |
| Certified processor coverage | Volume handled by R2v3 or e-Stewards facilities | Higher | Certification covers the tier you contracted, not tier three |
| PUE | Facility energy over IT energy | Lower | Improves when IT load rises; pair with a workload metric |
| WUE and REF | Water per kWh of IT energy; renewable share | Lower and higher | Site-dependent; disclose water stress context |
| Recycled content share | Secondary material in purchased hardware | Higher | Vendor-declared; verification is limited |
| Restatement rate | Published metrics revised this period | Non-zero, small | Zero suggests errors are not being found |
The column that carries the most weight is the caveat column. A programme that publishes these ESG metrics without their qualifications will be corrected by someone external. A programme that publishes them with the qualifications attached is doing the thing the disclosure regimes were designed to produce.
Common Mistakes When Tracking IT Asset ESG Metrics
The failure patterns in this field are consistent enough to enumerate, and most of them are structural rather than technical.
The first and largest is booking avoided emissions as reductions. It has been covered above and it deserves the top position because it is common, material, and explicitly contrary to the guidance.
The second is measuring only what is easy. Energy is metered, so energy gets measured, and endpoint energy is a fifth of endpoint lifetime carbon. The ESG metrics with real leverage — lifespan, utilisation, embodied carbon — require joining systems that were never designed to be joined, so they get deferred indefinitely.
The third is a denominator that is never stated. Diversion rates, reuse rates, and coverage percentages are all ratios, and in most published reporting the denominator is unspecified. Where it is unspecified it is almost always the convenient one.
The fourth is treating certification as an outcome. Holding a certified vendor is a control, not a result. The result is what happened to the assets, and only serial-level reconciliation demonstrates it.
The fifth is optimising a single one of your ESG metrics into a perverse outcome. Rewarding reuse rate shortens refresh cycles. Rewarding PUE discourages consolidation. Rewarding diversion rate discourages finding lost assets. Every ratio in this article can be improved by making the organisation worse, which is why they should be reported in sets rather than individually.
The sixth is annual collection. ESG metrics produced once a year by a person with a spreadsheet cannot be verified, cannot detect a trend, and cannot support a decision. Monthly production from a source system is a different product with the same name.
The seventh is ignoring the social and governance columns until asked. Both are cheaper to build alongside the environmental data than to retrofit, and both are explicitly required by the frameworks that are arriving.
The eighth is scope creep into precision that the data cannot support. Reporting embodied carbon to four significant figures when the underlying factor is a vendor estimate with an undisclosed uncertainty range signals a misunderstanding of the input, and reviewers notice.
Where IT Asset ESG Metrics Still Fall Short
An honest assessment has to include what this measurement framework cannot currently do, because the gaps are substantial and pretending otherwise is how programmes lose credibility.
Vendor product carbon footprints, the input behind your procurement ESG metrics, are not comparable across manufacturers. Methodologies, boundary choices, use-phase assumptions, and grid factors differ enough that a 15 percent difference between two vendors’ published figures for equivalent devices tells you very little. Within a vendor’s range the figures are useful for relative comparison. Across vendors they should be treated as indicative.
Cloud and SaaS remain largely unmeasurable from the customer side with any rigour. Provider tools are improving and are still not comparable, auditable, or timely. For an organisation whose compute has largely migrated, this means the fastest-growing part of the footprint carries the weakest data, and no amount of internal process fixes a disclosure the provider does not make.
Second-life tracking effectively stops at the point of sale. Once a refurbished device is sold, its subsequent service life, its eventual disposal, and whether it genuinely displaced a new purchase are all unknown. Reuse rate measures an intention rather than a confirmed outcome, and the avoided-emissions claims built on top of it inherit that uncertainty.
Critical raw material recovery is poorly measured everywhere. Mass-based recycling ESG metrics systematically undervalue the recovery of the materials that matter most, and the reporting infrastructure to track specific material streams from a corporate estate does not exist at usable cost.
Small and mid-sized organisations face a genuine proportionality problem. The measurement infrastructure described here is affordable at ten thousand devices and disproportionate at two hundred, yet the customer questionnaires increasingly do not distinguish. The realistic answer at small scale is a documented, simplified methodology with stated limitations rather than an attempt to replicate enterprise machinery.
Finally, the regulatory picture is unstable. The Omnibus simplification changed the ESRS data point count materially, Californian enforcement dates have shifted under litigation, and the ESPR product-group timetable continues to move. Building ESG metrics tightly coupled to any single regime’s current schema is a design error. Build them around the asset-level fact table and generate the schema on demand.
Frequently Asked Questions
Which ESG metrics should a small IT team start with?
Four: average asset lifespan by device class, embodied carbon per unit from vendor datasheets, reuse rate at disposition, and asset register completeness. All four are computable from data you already hold, they cover the highest-leverage decisions, and the fourth qualifies the other three honestly. Everything else can wait until those are produced reliably every month.
Can we count the carbon avoided by reusing our old laptops?
Not as a reduction in your emissions inventory. The GHG Protocol technical guidance instructs companies not to report negative or avoided emissions from recycling in Scope 3 Categories 5 or 12, and the SBTi treats avoided emissions as sitting outside the corporate inventory, so they do not count toward target achievement. You can report the figure as clearly labelled avoided emissions in narrative commentary. You cannot subtract it from your footprint.
Is PUE still useful among data centre ESG metrics?
Yes, with a caveat. It is required under the EU reporting framework, it is comparable across facilities, and it has driven real efficiency improvement. But it is a ratio with IT load in the denominator, so it can improve while total consumption rises, and it says nothing about useful work delivered. Report it because it is expected, and pair it with a workload-normalised efficiency measure so nobody reads it in isolation.
How do we get ESG metrics for cloud workloads?
Imperfectly. Record the provider’s reported figure, the methodology version it was produced under, and your own estimate derived from consumption and regional grid intensity. Disclose the divergence rather than choosing one number. Where cloud is a material share of the footprint, write reporting requirements into the commercial agreement at renewal, because that is the only mechanism with leverage.
What is the difference between reuse rate and landfill diversion rate?
Reuse rate measures devices that entered a second service life and therefore displaced the manufacture of a replacement. Landfill diversion measures mass that did not go to landfill, which includes shredding and materials recovery. Reuse avoids an entire embodied footprint; recycling recovers a fraction of material value. Combining them into one diversion figure erases the distinction that carries the environmental value.
Should we require R2v3 or e-Stewards certification?
It depends on your export risk posture. R2v3 permits controlled export to qualified downstream facilities under documented accountability and has substantially wider adoption, which matters for vendor availability. e-Stewards prohibits export of electronics to developing countries regardless of working condition and additionally requires NAID AAA data destruction and an ISO 14001 environmental management system. Many major providers hold both. Either way, certification of your direct vendor is not chain of custody — require named downstream processors as well.
How long should we keep laptops?
Longer than you currently do, and on a condition-based rather than calendar-based trigger. Moving from a three-year to a four-year cycle reduces annualised emissions by roughly a quarter, and four to six years takes annualised notebook emissions down by close to 29 percent on published modelling. The constraint is rarely hardware capability and usually policy inertia, warranty structure, and the absence of a health signal to justify individual retention decisions.
Does the CSRD Omnibus simplification mean we can stop tracking this?
No. The simplification reduced the number of mandatory data points substantially and tightened the materiality filter, with application targeted from FY2027. It did not remove resource use and waste from the standards, and for an organisation whose largest physical purchase category is electronics, hardware is difficult to exclude from a defensible materiality assessment. Plan for fewer required data points, not for the topic disappearing.
Who should own IT asset ESG metrics — IT, sustainability, or finance?
IT owns the source data and the calculation, sustainability owns the framework mapping and the narrative, and finance owns the disclosure and the assurance relationship. The failure mode is sustainability owning the calculation, because the team furthest from the source systems ends up reconciling them by hand. Give every one of the ESG metrics a single named owner inside IT with a documented method, and let the other two functions consume it.
Final Verdict
The ESG metrics worth tracking across the IT asset lifecycle are fewer and less flattering than the ones most organisations currently report. Embodied carbon, asset lifespan, idle ratio, reuse rate, chain-of-custody completeness, and a handful of facility and governance indicators cover the ground. Tonnes processed, diversion percentages without a stated denominator, and avoided-emissions credits do not.
The structural insight is that leverage and measurability are inversely related in this domain. The easiest numbers to collect — endpoint energy, tonnes recycled, ecolabel coverage — sit closest to the edges of the carbon picture. The numbers that move the outcome, principally how long assets live and how many of them exist at all, require joining systems that were never designed to be joined. That is an engineering problem with a known solution, and it is the work.
The regulatory timeline has removed the option of postponing it. CSRD reporting is running, Californian Scope 3 disclosure arrives in 2027 for the prior year, and customer questionnaires already ask questions that most IT organisations cannot answer with evidence. Assurance is the mechanism that will separate real programmes from presentational ones, and it does not respond well to a confident number with no lineage.
The recommendation is narrow. Build the serial-keyed asset fact table first, attach embodied carbon factors and lifecycle dates to it, and generate every metric as a query against it. Correct the avoided-emissions treatment now rather than after publishing three more years of figures. Renegotiate the disposition data contract before computing another disposition metric. And report every ratio with its denominator stated, because the denominator is where the credibility of an IT asset ESG metrics programme is won or lost — the same discipline that separates a genuine sustainable technology practice from a well-designed slide.
References
The Global E-waste Monitor 2024, published by the International Telecommunication Union and UNITAR, provides the generation and formal collection figures cited above, including 62 million tonnes generated in 2022 and a documented collection and recycling rate of 22.3 percent, with a projected decline toward 20 percent by 2030. The UNITAR summary states the headline finding that e-waste generation is rising five times faster than documented recycling.
The GHG Protocol’s Scope 3 calculation guidance and the technical chapters for Category 5 and Category 12 are the authoritative source for the treatment of waste and end-of-life emissions, including the instruction not to report negative or avoided emissions from recycling. The Science Based Targets initiative’s position on avoided emissions is set out in its published criteria and FAQ material.
The EU data centre reporting obligations derive from Commission Delegated Regulation (EU) 2024/1364 and the associated rating scheme work described by the European Commission. Circularity measurement requirements are set out in ISO 59020:2024. Ecolabel criteria are published by the Global Electronics Council for EPEAT and by TCO Development for TCO Certified. Data sanitisation guidance is NIST Special Publication 800-88 Revision 1. Disposition certification standards are R2v3 from Sustainable Electronics Recycling International and e-Stewards from the Basel Action Network.
Device lifetime figures are drawn from TCO Certified’s published modelling of four-year versus six-year notebook service lives and from Atos research on sustainable workplace practice. Product carbon footprint proportions are drawn from manufacturer-published datasheets and independent benchmarking of those datasheets; readers should treat cross-vendor comparisons as indicative rather than precise for the reasons set out above.
