TL;DR
- Equipment ESG reporting fails on evidence rather than ambition. Most companies can state a policy and cannot prove an outcome.
- Four things are provable from an asset register: how many devices you bought, how long you kept them, how many came back, and where they went afterwards.
- Everything beyond those four needs a vendor certificate, and a certificate you cannot produce on request is the same as no certificate.
- Refresh cycle length is the single highest-impact number you control and the easiest to evidence. Start there.
- Devices that are never recovered are the gap that invalidates everything else. An unreturned device has an unknown end of life by definition.
- Do not report an emissions figure derived from a generic per-device factor and present it as measured. Say what it is.
The Claim Nobody Could Support
The sustainability section of the tender asked a specific question. What proportion of retired IT equipment is reused or recycled through certified channels, with evidence.
The company's policy said all of it. The policy was real, it had been signed off, and it was on the website. What nobody could produce was the evidence, because the asset register recorded devices as "disposed" without recording where, the recycler's certificates were in an email folder belonging to somebody who had left, and roughly a fifth of devices issued in the previous three years had no disposal record at all because they had never come back.
The honest answer was that the company could evidence certified disposal for about 60 per cent of retired devices, had no record for a fifth, and had a policy that claimed 100 per cent. That is a worse position than having no policy, because the gap between the claim and the evidence is itself a finding.
Best tools for HR Operations
The real issue is not that the recycling was not happening. It mostly was. It is that equipment ESG reporting is an evidence exercise, and an evidence exercise fails on record-keeping rather than on behaviour.
The Four Things You Can Actually Prove
Start from what an asset register can support, because a report built on anything else will not survive a question.
How many devices you acquired, and when. Purchase records plus the register. This is the easiest number and the foundation of everything else, because every other metric is a proportion of it.
How long you kept them. Issue date to retirement date, averaged across a cohort. This is the number that matters most environmentally and it is entirely within your records.
How many came back. Devices recovered against devices issued, per cohort. This is where most registers start to struggle, and the struggle is the finding.
Where they went. Reused internally, sold, donated, recycled, or unknown. The honest category list includes "unknown" and most published reports do not have it.
| Metric | Source | How hard |
|---|---|---|
| Devices acquired | Purchase records | Easy |
| Average service life | Issue and retirement dates | Easy if dates are recorded |
| Recovery rate | Issued against recovered, per cohort | Moderate; exposes register gaps |
| Disposition split | Retirement reason plus vendor records | Hard; needs the certificate chain |
| Emissions | A factor applied to the above | Not measured, and should not be presented as such |
What to do: publish the four you can prove and state plainly that the fifth is an estimate with its method named. A report with four solid numbers and one labelled estimate is stronger than one with five numbers of unstated provenance.
Refresh Cycle Is the Number That Matters
If you report one thing, report this, because it is both the highest-impact lever and the easiest to evidence.
Most of a laptop's lifetime footprint is in its manufacture. That is the widely accepted shape of the problem and it means the environmental question is overwhelmingly about how long the device stays in service rather than how it is powered. Extending a fleet's average service life is the largest single action available to most companies.
And you already have the data. Issue date and retirement date are in the register, or recoverable from it. No vendor, no factor, no estimate. An average service life by cohort is a measured number, which is rare in this area.
Report it as a distribution, not just a mean. A fleet with a three-year average made up of some devices retired at 18 months and some at five years tells a different story from one where everything is retired at three. The short tail is where the avoidable replacement sits.
And report the reasons for early retirement. Failure, damage, loss, role change, or policy. Each has a different fix, and the split is far more actionable than the average. Damage and loss in particular connect to a different conversation, one we treat separately in device damage liability, because the handling of it is a people question first.
What to do: publish average service life, the distribution, and the early-retirement reason split. Three measured numbers, no estimates, and they support an improvement plan rather than only a claim. The split is also the one output of the whole exercise that an operations team can act on next quarter rather than next year, because each reason has a different and fairly obvious owner: failures point at the specification, damage and loss point at handling and at the recovery process, and role changes point at how devices are reallocated internally rather than replaced.
The Recovery Rate Is the Gap That Invalidates Everything
This is the finding that undermines most equipment ESG reporting, and it is rarely stated.
A device you never recovered has an unknown end of life. It cannot be counted as recycled, it cannot be counted as reused, and it cannot honestly be left out of the denominator. If a fifth of your fleet is unaccounted for, no disposition claim about the other four-fifths can be presented as a claim about your fleet.
The rate is usually worse than people assume and worse for remote staff. Devices in an office come back. Devices with somebody who left a company they joined remotely, in a market with no local logistics partner, frequently do not, and the gap is structural rather than a matter of diligence.
The fix is upstream of the reporting. Recovery is an operational problem solved by shortening the lead time between a departure being known and a retrieval being started, which is a workflow question rather than a sustainability one. We looked at the mechanism in asset register HRIS integration, and the recovery side specifically in IT asset recovery for small teams.
And report the unrecovered devices as their own line. Not folded into recycling, not excluded from the denominator. A report that says 78 per cent of retired devices went through certified channels, 14 per cent were reused internally and 8 per cent were never recovered is credible. One that says 100 per cent were recycled is not, and an auditor will find the gap faster than you expect.
What to do: calculate the recovery rate per cohort before writing any disposition claim, because it sets the ceiling on what you can claim. A 90 per cent recovery rate caps every disposition statement at 90 per cent, and saying so is the difference between a report and a marketing page.
What Needs a Certificate, and What a Certificate Needs to Say
Anything that happens after a device leaves your control is a vendor claim, and a vendor claim you cannot produce is not evidence.
Data destruction. A certificate naming the method, the date and the specific devices by serial. A certificate that says a batch of 40 devices was wiped, without serials, cannot be matched to your register and therefore cannot support a device-level claim.
Recycling and recovery. The downstream chain matters here. A recycler that passes material on needs to be able to say to whom, because a claim about certified recycling that stops at the first vendor is a claim about your contract rather than about an outcome.
Resale and reuse. If a device was sold, it was reused, which is the best environmental outcome available and is frequently reported worse than recycling because the paperwork is thinner. Worth recording properly: a sale record with serials is as good as a recycling certificate for reporting purposes and represents a better outcome.
Donation. The same, with the extra requirement that the data destruction happened first and is separately evidenced.
| Disposition | Evidence required | Common failure |
|---|---|---|
| Data destruction | Certificate with serials, method, date | Batch certificate with no serials |
| Certified recycling | Certificate plus downstream chain | Chain stops at the first vendor |
| Resale or reuse | Sale record with serials, plus wipe evidence | Not recorded at all; reported as disposal |
| Donation | Recipient record plus wipe evidence | Wipe evidence missing |
| Unrecovered | Nothing to evidence; report as unknown | Folded into the recycled figure |
What to do: store certificates against the asset record rather than in an email folder belonging to one person, which is how the evidence in the opening story went missing when that person changed jobs, and check that they carry serials when they arrive rather than two years later during a tender. A certificate without serials can sometimes be reissued with them if you ask within a reasonable period, and essentially never if you ask three years on.
Emissions Figures, and How to Be Honest About Them
This is where equipment ESG reporting most often crosses from evidence into something else, usually without intending to.
A per-device emissions figure is a factor, not a measurement. You take a published figure for the embodied emissions of a device type, multiply by your device count, and get a number. That is a legitimate estimate and it is not a measurement, and the difference matters because the two get presented identically.
Vendor-published figures vary and are not comparable across manufacturers. Different boundary assumptions, different methodologies, different reporting years. Adding figures from three manufacturers produces a total whose error bars nobody can state.
So say what you did. Name the source of each factor, name the device counts it was applied to, and state that the result is an estimate derived from published manufacturer figures. That sentence costs you nothing in credibility and protects the whole report, because the alternative is a number that cannot be defended under questioning.
And never present a saving you did not measure. An extended refresh cycle avoids manufacturing emissions, which is real, and the quantity is an estimate built on the same uncertain factors. "We extended average service life from 36 to 48 months across 1,200 devices" is a measured claim. The tonnage that implies is an estimate, and the measured claim is the stronger one to lead with.
What to do: lead with the operational numbers you measured and put the derived emissions estimate second, labelled. It is a better report and it is the one that survives a due diligence question.
The Cohort Method, and Why Annual Snapshots Mislead
How you slice the data decides whether the report is informative or flattering, and the default slice is the flattering one.
An annual snapshot compares this year's retirements against this year's purchases. In a growing company that always looks good, because you bought more than you retired. The ratio is a statement about growth rather than about equipment management, and it moves with hiring rather than with anything you did.
A cohort follows the devices you issued in one period. Of the 400 laptops issued in a given year, how many are still in service, how many were retired and for what reason, how many came back, and what happened to them. That is a closed population with a denominator that does not move, and every number derived from it means something.
The cohort view is where the unrecovered devices become visible. An annual snapshot never surfaces them, because a device that was never returned is simply absent from the retirement count, which makes the disposition split look complete. A cohort view shows them as a residual that has to be explained, which is exactly why it is the harder report to produce and the only one worth producing.
Cohorts take time to mature, and that is fine. A cohort issued two years ago is still partly in service, so its numbers are provisional. Report it as provisional. A cohort from five years ago is essentially closed and tells you the truth about your process as it was then, which is a useful comparison against a more recent one.
What to do: report by issue cohort, publish two or three cohorts side by side, and mark the recent ones provisional. It is more work than a snapshot and it is the difference between a report that measures your process and a report that measures your headcount growth.
What Auditors and Tenders Actually Ask For
The questions are more predictable than the frameworks suggest, and preparing for the four below covers most of what arrives.
Can you trace a specific device from purchase to disposal. Pick one serial, show the purchase record, the issue record, the retirement record and the disposal certificate. This is the single most common test and it fails on the certificate link rather than on any of the first three.
What proportion of retired equipment went through certified channels, with evidence. This is the opening story. The answer has to include the unrecovered residual, and a figure under 100 per cent with a stated gap is received far better than a round number that cannot be substantiated.
What is the average service life of your fleet. Measured, from your own records, and the easiest question on the list to answer well. Have the distribution ready too, because the follow-up is usually about the short tail.
Who is responsible, and how is this reproduced next year. A process question rather than a data one, and the one that catches reports produced by hand for a single deadline. A named owner and a written field list is the whole answer.
What to do: rehearse the first question before you need it. Trace one device end to end, in your own systems, and see where it breaks. In most companies it breaks at the certificate, which tells you what to fix with months rather than days of notice.
When You Don't Need a Formal Report
Under about 100 devices. The data gathering costs more than the finding is worth, and nothing you report will change a decision. Keep the records so you can report later.
No external requirement. If no tender, customer, investor or regulation is asking, a formal report is work with no reader. Fixing the recovery rate is still worth doing, because that is an asset recovery argument rather than a reporting one.
Your records cannot support it yet. This is the common case and the honest response is sequencing, not scope reduction. Fix the retirement reason field and the certificate storage this year, report next year, and say internally that this is what you are doing.
Nobody owns it. An equipment sustainability report with no owner gets produced once, for a specific tender, by whoever was asked, and is never reproducible. Worse than useless, because the next person has to start over and will produce different numbers.
What to Fix First, in Order
The sequence matters more than the ambition, and this order is chosen so each step makes the next one possible.
One: record a retirement reason. A single field, a short fixed list: end of life, failure, damage, loss, role change, policy refresh, unrecovered. Without it, every other analysis is guesswork. This is the cheapest high-value change in the whole list.
Two: store certificates against the asset record. Not in email. The link between a serial and its disposal evidence is the thing a tender asks for, and reconstructing it later is expensive or impossible.
Three: measure the recovery rate per cohort. It sets the ceiling on every claim you can make, so knowing it tells you what is reportable before you draft anything.
Four: publish average service life and its distribution. Measured, defensible, and the number with the most environmental significance.
Five: only then estimate emissions, with the factor sources named and the result labelled as an estimate.
What to do: do one and two this quarter. They are both small, they are both prerequisites for everything else, and they are the two that cannot be done retroactively.
Procurement Decisions That Change the Numbers
Reporting measures what buying decisions produced, and four of those decisions do most of the work. Each is made before any reporting exists, which is why they are usually made without reference to it.
The specification you buy at. A machine specified at the edge of what the role needs gets replaced when the role grows. One with headroom stays in service longer, which moves the biggest number in the report. That is a cost-per-hire conversation with an environmental consequence, and the two are usually held separately.
Whether the device is repairable. Replaceable memory, storage and batteries extend service life materially, and increasingly manufacturers publish repairability information. A fleet of devices that cannot have a battery replaced has a hard ceiling on service life regardless of policy, and it is a ceiling you chose at purchase.
Whether you buy refurbished. A refurbished device issued to a new joiner is a reuse outcome you can evidence, and it is the only line in the report that improves at the point of purchase rather than at the point of retirement. It also has a shorter remaining life, so it shortens your average service life figure while improving the underlying outcome, which is a reporting artefact worth being ready to explain.
Who disposes of it, and what the contract requires. This is the cheapest of the four. Requiring serials on every certificate, and a statement of the downstream chain, is a contract clause rather than a project. Adding it at renewal costs nothing and fixes the evidence failure that the opening story turned on.
What to do: put the four above into the procurement checklist rather than into a sustainability plan. A decision made at purchase is cheap and a decision made at retirement is mostly already determined.
What a Second Year Should Look Like
The first report is an archaeology exercise. The second should not be, and the difference is designed rather than hoped for.
Same fields, same cohort definition, same factor sources. Changing any of the three makes the year-on-year comparison meaningless, and somebody will ask for that comparison. Write the three down with the first report so the second can reproduce them.
Derived, not assembled. If the first report was built by hand, the second should come from a query. The work between them is making sure the four fields are populated as part of ordinary record-keeping, not a data-gathering push in the final month.
Narrower, not wider. The temptation in year two is to add metrics. The better move is to improve the evidence behind the ones you have, particularly the recovery rate, because the credibility of the whole report sits on its weakest number rather than its count of numbers.
With the gap closing visibly. A report that says last year 8 per cent of the cohort was unaccounted for and this year 3 per cent is a report about a company that fixed something. That is a stronger artefact than a flawless-looking first report, and it is the one that stands up to a follow-up question.
What to do: write the method note at the same time as the first report, not afterwards, and keep it with the report rather than in somebody's inbox. It is three paragraphs and it is what makes the second year cost a fraction of the first.
Who Should Own Equipment ESG Reporting
Ownership is the reason these reports are either reproducible or not, and the common arrangements both fail in predictable ways.
Sustainability teams own the framework and not the data. They know what needs reporting and under which standard, and they do not have access to an asset register or a relationship with a recycler. A report owned entirely by sustainability becomes a request for numbers that somebody else produces differently each year.
IT owns the data and not the requirement. They can produce device counts and dates, and they have no visibility of what the report has to satisfy, so the numbers arrive in a shape that needs rework. They also have no reason to prioritise it against operational work.
The workable split is that IT owns the record and sustainability owns the report. With one explicit agreement: the fields the report needs are specified once, in writing, and IT maintains them as part of ordinary record-keeping rather than as an annual exercise. That is what makes the second year cheaper than the first.
Procurement owns the certificate chain, because they own the vendor relationship and the contract that should require serials on certificates in the first place. Putting that requirement in the contract is a five-minute change that solves the single most common evidence failure.
What to do: agree the field list in writing between the three, once, and put the serial requirement into the disposal vendor contract at the next renewal. Those two steps do more for reproducibility than any amount of reporting effort.
The Tools, and What They Hold
No product produces this report, and the honest framing is that your register either holds the fields or it does not.
What a register must hold. Issue date, retirement date, retirement reason, and a link to disposal evidence. Those four, against a serial. Everything in this article is derived from them.
AssetTiger publishes a plain ladder by asset count, with tiers at $15, $20, $40, $75, $140 and $275 per month, and holds custom fields, so the retirement reason can be added. Snipe-IT is free and open source self-hosted, with managed hosting published at $399.99 per year for the small tier and $2,499.99 at the top, and its documented API makes a certificate link and a cohort query straightforward.
Lansweeper, Device42 and Asset Panda publish no readable price and route to a quote. Each holds lifecycle fields to a degree a spreadsheet does not, and the question to ask each is specific: can a disposal certificate be attached to a device record, and can you report recovery rate by issue cohort.
RemoAsset is an equipment lifecycle platform aimed at distributed teams, and it is owned by the same people who publish this site, which you should weigh against everything said about it here. Its relevance here is the recovery rate specifically: because it handles retrieval and disposition rather than only recording them, the unrecovered gap that invalidates most reporting is the part of the problem it is built around. Its limitations are real. Pricing is quote-based with no published figure, so it cannot be compared on price from a web page, and it is a newer platform with a smaller footprint than the long-established ITAM names.
| Capability | Why the report needs it |
|---|---|
| Retirement reason as a field | Separates end of life from avoidable replacement |
| Certificate attached to the serial | The thing a tender actually asks for |
| Recovery rate by issue cohort | Sets the ceiling on every disposition claim |
| Service life distribution | The highest-impact measured number |
| Resale and donation recorded as reuse | Otherwise the best outcome reports as disposal |
What to do: ask any shortlisted vendor to produce a recovery rate by issue cohort from demo data. It is a narrow question and it separates products that record a lifecycle from products that merely record devices.
Three Claims to Avoid Making
Equipment ESG reporting goes wrong in three specific sentences, and each is common enough to be worth naming.
"All our retired equipment is recycled responsibly." Unsupportable the moment any device is unaccounted for, which at a distributed company is always. The defensible version names a proportion and states the residual, and it reads as more credible rather than less because a round number in this area invites exactly the question it cannot survive.
"We have reduced our device emissions by a stated tonnage." Derived from factors with unstated error bars, presented as an achievement. The measured claim underneath it, that average service life rose from one number of months to another across a stated fleet size, is both true and more impressive to anybody who understands the area. Lead with that.
"Our devices are reused rather than disposed of." Frequently true and almost never evidenced, because resale and donation records are kept worse than recycling certificates. If you are going to make a reuse claim, keep the sale or donation record with serials and the separate wipe evidence, because without those two it is the hardest claim on the list to substantiate.
And one framing to avoid. Presenting equipment sustainability as a target rather than as a measurement produces pressure on the number rather than on the outcome, and the number is the part that is easy to move. A report that gets narrower and better evidenced each year is doing the job. One that gets rounder is not.
What to do: read your current published policy against the four numbers you can actually produce, before anybody external does it. If the policy claims more than the evidence supports, narrowing the policy is faster, cheaper and safer than closing the evidence gap, and you can do both in that order.
The Decision Table
| Situation | Scale | Setup | Primary Pain | Recommended Starting Point |
|---|---|---|---|---|
| Under 100 devices | Under 100 | Register exists | No reader for a report | Keep records, do not report yet |
| Tender asks for evidence | Any | Policy but no records | Claim exceeds evidence | State what you can prove, with gaps |
| High unrecovered rate | 200 plus | Distributed team | Caps every claim | Fix retrieval lead time first |
| No retirement reason field | Any | Register without the field | Every analysis is guesswork | Add the field this quarter |
| Certificates in email | Any | Vendor relationship exists | Cannot match serial to evidence | Attach to the asset record |
| Reporting under a standard | 500 plus | Sustainability team involved | IT and sustainability disconnected | Agree the field list in writing |
| Asked for an emissions figure | Any | Device counts available | Estimate presented as measured | Name the factor source, label it |
What Getting This Wrong Costs
The direct cost is a failed tender question, and that is recoverable.
The expensive version is a published claim that the evidence does not support. A policy saying all retired equipment is recycled through certified channels, on a website, where a fifth of devices were never recovered, is not an aspiration that fell short. It is a statement about an outcome that cannot be substantiated, and the exposure attached to that has grown considerably as reporting expectations have tightened. The company in the opening story was in a worse position for having the policy than it would have been with a narrower, truer one.
The second cost is that the reporting exercise hides the operational finding. A company that spends a quarter producing an emissions estimate from published factors has learned nothing it can act on. A company that spends the same quarter finding out that its recovery rate is 78 per cent has found a recoverable-asset problem worth real money and an environmental problem at the same time, and the fix for both is the same fix.
And the third is reproducibility. A report produced once for a specific requirement, by whoever was available, from numbers assembled by hand, costs the same again next year and produces different figures. The difference will be asked about, and the honest answer is that the method changed, which is the answer nobody wants to give.
So the question is not what your equipment sustainability policy says. It is which four numbers you can produce from your own records this afternoon, and what proportion of your fleet has an end of life you cannot account for.
When You're Ready to Report
Do the two cheap prerequisites first. Add a retirement reason field with a short fixed list, and start attaching disposal certificates to the asset record instead of filing them in email. Neither can be done retroactively, which is why they come first.
Then measure the recovery rate by cohort, because it tells you the ceiling on everything you can claim, and publish average service life with its distribution and its early-retirement reasons. Those are measured numbers from your own records and they support an improvement plan rather than only a statement.
And when you get to an emissions figure, name the factor source and call it an estimate. The report that leads with four measured operational numbers and labels the fifth honestly is the one that survives the follow-up question, which is the only test that matters.
One last check, and it takes an hour. Pick a single serial from three years ago and trace it end to end through your own systems: what you paid, who got it, when it was retired, why, where it went, and which certificate covers it. Whichever link breaks is the one to fix, and you will know considerably more about your reporting position after that hour than after a month spent drafting.
Frequently Asked Questions
What can equipment ESG reporting actually evidence?
Four things from your own records: how many devices you acquired and when, how long you kept them, how many came back, and where they went afterwards. Everything past those four depends on a vendor certificate or a published emissions factor, and both should be labelled as what they are. A report with four measured numbers and one clearly marked estimate is considerably stronger than one with five numbers whose provenance is unstated.
Why does the recovery rate matter so much?
Because a device you never recovered has an unknown end of life by definition, so it cannot be counted as recycled or reused and cannot honestly be dropped from the denominator. The recovery rate therefore sets the ceiling on every disposition claim you can make. If 8 per cent of a cohort never came back, no statement about certified disposal can describe more than 92 per cent of that cohort.
What should a disposal certificate contain?
The method, the date and the specific devices by serial number. A batch certificate stating that 40 devices were wiped, with no serials, cannot be matched to your asset register and so cannot support a device-level claim. Check this when certificates arrive rather than during a tender, because a vendor can often reissue with serials within a reasonable period and essentially never three years later.
Is resale better or worse than recycling?
Better, because reuse extends a device's service life rather than recovering materials from it. It is frequently reported worse because the paperwork is thinner: a recycling certificate looks like evidence and a sale record often is not kept at all. A sale record with serials plus separate wipe evidence is as good for reporting purposes and represents the better outcome, so it is worth recording properly.
Can we report an emissions figure?
Yes, as a labelled estimate. You take published manufacturer factors, apply them to your device counts, and name the sources. What you should not do is present that as a measurement, or add figures from several manufacturers without noting that their boundary assumptions and methodologies differ. Lead with the operational numbers you measured and put the derived estimate second.
What single change improves the reporting most?
Adding a retirement reason field to the asset register, with a short fixed list covering end of life, failure, damage, loss, role change, policy refresh and unrecovered. Without it every other analysis is guesswork, and it cannot be filled in retroactively for devices already retired. It takes an afternoon and it is a prerequisite for the service life and early-retirement analysis that carries the report.
Who should own this?
IT owns the record, sustainability owns the report, and procurement owns the certificate chain, with the field list agreed once in writing between them. The arrangement that fails is sustainability owning the whole thing, because it becomes an annual request for numbers that somebody else assembles differently each year, and the second year costs as much as the first.
What if our records cannot support a report yet?
Sequence it rather than narrow it. Add the retirement reason field and move certificate storage onto the asset record this quarter, measure the recovery rate next quarter, and report the following year. Say internally that this is the plan, because the alternative is a report assembled by hand for one deadline that nobody can reproduce or defend afterwards.