Why verification is the control that matters
Everything else in fixed asset management is a computation on top of an assumption. The depreciation working, the insurance cover, the impairment assessment and the disposal accounting all assume the asset exists, is where the register says it is, and is in the condition recorded. Physical verification is the only control that tests the assumption.
This is why auditors ask about it early and why the answer is often uncomfortable. Verification is expensive in effort, produces bad news, and generates work for people who did not create the problem. The result in many organisations is a verification exercise that is performed but not evidenced, or evidenced but not reconciled.
A verification that does not change the register is not a verification. It is a walk around the building.
Planning the exercise
Plan coverage before mobilising anyone. Four decisions determine whether the exercise produces a usable result.
Scope and coverage
Decide which locations, which asset classes and what proportion of each will be covered. Full coverage of a large estate in one round is rarely achievable, and pretending otherwise produces partial coverage that is reported as complete. A stated coverage percentage per class per location is more honest and more useful.
Cut off
Fix the date at which the register extract is taken, and record movements that happen during the verification window separately. Without a cut off, every discrepancy can be explained away as a movement that happened after the count.
Team and independence
Decide who verifies. Departmental staff know where things are; independent teams classify more consistently and are not verifying their own custody. A common arrangement is an independent team accompanied by a departmental representative who can open rooms and identify equipment.
Classification rules, agreed in advance
Agree what found, not found, moved, damaged, idle and untagged mean before the first site. If these are decided during the exercise, the first three sites are classified differently from the rest and the whole dataset becomes questionable.
Tagging before counting
Verification without tagging is an argument. Two people looking at four identical pumps cannot agree which is asset number 4471 unless the pump says so.
Tag under your own numbering policy rather than a vendor format, so that the tag remains meaningful if you change systems. Barcode and QR labels suit most environments. RFID is worth the additional cost where density or access makes line of sight scanning impractical, such as stores, server rooms and armouries, because it turns a count into a read.
Label durability matters more than it seems. A label that does not survive a wash down, a furnace area or a year of handling means the next verification round starts by re-tagging, which doubles the cost of the exercise for no benefit.
What counts as evidence
The purpose of evidence is to allow somebody who was not present to satisfy themselves later. That sets a clear standard.
- Geo coordinates and timestamp captured at capture. Not typed in afterwards. This is the difference between a photograph and evidence.
- Verifier identity recorded. Attribution is what allows a query to be resolved by asking the person who was there.
- The tag visible in the frame. A photograph of a machine proves a machine exists. A photograph of a machine with its tag proves which machine.
- Condition and remarks recorded structurally. Free text alone cannot be reported on. Structured condition codes with optional remarks can.
- Stored against the asset record. Evidence in a shared folder organised by date of visit is evidence nobody will find in eighteen months.
Offline capability is not a convenience feature in Indian conditions. Basements, plant floors, district offices and remote sites routinely have no usable connectivity, and a tool that requires a live connection produces either gaps in coverage or data entered from memory afterwards. Capture offline, sync later.
Classifying exceptions honestly
The output of verification is not a count. It is an exception list. How exceptions are classified determines whether the exercise leads to correction or to a spreadsheet nobody actions.
| Exception | Meaning | Action it triggers |
|---|---|---|
| Found as recorded | Asset located in the recorded location with the recorded custodian | Update last verified date and evidence |
| Found, location differs | Asset located, but not where the register says | Raise a transfer for approval, investigate why it was unrecorded |
| Found, custodian differs | Asset located, held by someone other than the recorded custodian | Reassign with acknowledgement, review the handover process |
| Found untagged | Asset present but carrying no tag or an unreadable tag | Tag during the visit, then match to the register or investigate as unrecorded |
| Not found | Asset in the register could not be located during the visit | Second search, then escalate for explanation before any write off |
| Found, not in register | Asset present that does not appear in the register at all | Investigate capitalisation history, add with proper approval |
| Damaged or idle | Asset located but not in usable service | Assess for impairment, repair or disposal |
Note the two categories that organisations most often suppress: not found and found but not in register. The first is uncomfortable because it implies loss. The second is uncomfortable because it implies purchases that were never capitalised. Both are more expensive to discover later, and both are exactly what an auditor is looking for.
Reconciliation and adjustment
Reconciliation compares three things: what verification found, what the register holds, and what the general ledger carries. Each difference needs a reason code and an owner, not a balancing entry.
Adjustments should run through the same approval workflow as any other register change, with the verification evidence attached. Where an asset is written off following verification, the file should show the search, the escalation, the explanation and the approval, in that order. Where an unrecorded asset is added, it should show the investigation into why it was never capitalised.
Any adjustment that cannot be explained should stay open as an exception rather than being cleared to make the totals agree. An open exception with an owner and an age is a manageable control weakness. A cleared exception with no explanation is a finding waiting to happen.
Turning it into a cycle
The organisations that stop having asset problems are the ones that stop treating verification as a project. A rolling cycle works better than an annual event for three reasons: the effort is spread, the exceptions arrive in volumes that can actually be actioned, and every asset carries a recent last verified date rather than a date from the previous financial year.
A practical pattern is risk based: high value, mobile and high theft risk assets verified more frequently; static, low value assets on a longer rotation; and any asset involved in a transfer, an incident or a custody change verified as part of that event rather than waiting for its turn.
Readiness checklist
Before the audit, these are the questions worth being able to answer from the system:
- What percentage of assets by value has been verified in the last twelve months.
- What is the last verified date for any asset an auditor selects.
- Can you produce the photograph, coordinates and verifier for that asset.
- How many exceptions are open, by category and by age.
- For each write off in the period, can you show the search, escalation and approval.
- Does the register reconcile to the general ledger, and are the differences explained.
MeltX FAM is built to answer those six questions without anyone preparing a report, and MeltX also puts trained verification teams on the ground through its asset verification and compliance service for organisations whose estate is larger than their own capacity to cover it.