The complete guide to physical asset audits

A physical asset audit is the moment your asset register meets reality. Done well, it confirms that recorded assets exist, finds the ones that do not, and reveals the process failures that let the register drift. Done badly, it is weeks of walking, a spreadsheet of unexplained differences and a write-off nobody can justify.

This guide covers the complete audit cycle: audit types, roles and controls, planning, register preparation, counting methods, sampling, variance reconciliation, write-offs, reporting and building a cycle-count program that keeps the register accurate all year.

Why physical audits matter

An asset audit supports several goals at once:

• Financial reporting. Capitalized fixed assets on the balance sheet must exist. Auditors often test this directly, and a reliable internal audit makes their work — and your year-end — much smoother.
• Loss control. Missing assets found quickly can often be recovered; missing assets found a year later usually cannot.
• Operational accuracy. People only use a register they trust. Every audit that corrects locations and custodians increases that trust.
• Compliance. Safety-critical and regulated equipment must be accounted for, inspected and in date.
• Insurance. Accurate schedules support correct cover and claims.

Types of audit

See cycle count and stock-take in the glossary.

Roles and controls

Good audits separate duties so that the results can be relied on:

• Audit owner — plans the audit, defines scope and signs off the report. Often from finance or internal audit.
• Counters — perform the count. Ideally not the custodians of the assets they count.
• Area guides — people who know the space and can open locked rooms, but do not record results.
• Reconcilers — investigate variances.
• Approvers — authorize adjustments and write-offs; independent of the custodian.

In smaller organizations one person may wear several hats. The principle to preserve is that nobody should be able to both lose an asset and quietly close the record of its loss.

Planning the audit

Define scope and objective

Write a short audit plan stating:

• The objective (for example, "verify existence and location of all capitalized assets at the Riverside site").
• Scope by location, category and value threshold.
• Count dates and cut-off time.
• Counting method(s).
• Teams and area assignments.
• How variances will be handled and who approves adjustments.

Set a cut-off

Movements during the count are the enemy of accuracy. Agree a cut-off: after it, movements in and out of the scope area are either frozen or recorded separately so they can be reconciled. For a count that spans several days, count areas where movement is lowest first, or freeze each area as it is counted.

Choose timing

Avoid peak operational periods. Allow time after the count for reconciliation before any reporting deadline — reconciliation usually takes longer than counting.

Preparing the register

Most time lost in audits is lost to data problems. Before the count:

1. Snapshot the register for the scope, so counts are compared with a fixed list.
2. Record known movements: transfers, open check-outs, items at repair vendors, pending disposals.
3. Remove duplicates and fix records with no location.
4. Confirm disposed assets are closed, not still marked active.
5. Check tags: are there records without tag IDs? Plan to tag them during the count.
6. Prepare area lists, sorted in walking order.
7. Pre-brief area owners and ask them to flag known discrepancies.

Our asset register guide covers the data structure that makes this easy.

Counting methods

Informed vs blind counts

• Informed count — counters see the expected list and confirm each item. Faster, but prone to confirmation bias: it is easy to tick an item that looks similar.
• Blind count — counters record what they find without seeing the expected list; comparison happens afterwards. Slower, but stronger evidence.

Scanning tags largely resolves the dilemma. A scan identifies the specific asset, not "something that looks like it", so an informed scan-based count keeps most of the rigour of a blind count.

Technology choices

• Paper — only for very small scopes. Slow and error-prone because of re-keying.
• Mobile QR or barcode scanning — the practical default. Results are timestamped and attributed as you go.
• NFC tap — useful where labels are covered or dirty.
• UHF RFID sweeps — many tags read per pass; best for dense storage. Watch for reading tags in neighbouring areas.

You can combine methods in one audit.

Existence and completeness testing

Auditors think about two directions of testing, and it is worth adopting their vocabulary:

• Register to floor (existence). Pick items from the register and find them physically. This tests whether recorded assets really exist — the key risk for overstated balance sheets.
• Floor to register (completeness). Pick items physically present and find them in the register. This tests whether everything you own is recorded.

A full count tests both directions at once. Spot checks should deliberately include both. If your external auditor tests a sample, they will usually specify how many items and how they are selected — ask early so your internal checks align.

Count day

1. Brief every team on scope, method, how to handle unlabelled or damaged items, and how to record items found in the wrong area.
2. Assign areas, not lists. Each team owns a zone completely.
3. Walk systematically — a consistent direction, shelf by shelf, desk by desk.
4. For each asset: scan, confirm location, note condition if in scope.
5. For unregistered items: record description, serial number, photo and location. Tag them if your process allows.
6. Before closing an area, review its "not yet found" list and search again.
7. Log everything — who counted where and when.

Handling difficult situations

Every audit meets situations the plan did not anticipate. Decide your approach to the common ones in advance.

• Locked or restricted areas. Arrange access with the area owner before count day, and schedule a follow-up visit rather than marking everything inside as missing.
• Assets in use off-site. Items at customer sites, on projects or with field staff cannot be walked to. Verify them through the custodian — for example, by asking them to scan the tag and submit a photo — and record the method used, since it is weaker evidence than a physical sighting.
• Remote workers' equipment. The same approach works for laptops and phones at home offices. Set a deadline and follow up on non-responses.
• Assets at repair vendors. Confirm with the vendor in writing, and check against open work orders.
• Items in transit. Check shipping records and confirm at the destination once they arrive.
• Assets too large or dangerous to approach. Verify visually from a safe distance and read the asset plate, or ask a qualified person to confirm.
• Unreadable tags. Identify the asset by serial number, re-label it with the same ID, and record the replacement.
• Multiple identical items. Where serial numbers are not recorded, identical items can only be counted, not individually verified. Record the count, and plan to capture serial numbers so the next audit can be precise.

Whatever the situation, record how each asset was verified. A physical scan, a custodian photo and a vendor confirmation are all legitimate, but an auditor will want to know which was used.

Audit readiness checklist

Use this list in the week before count day:

• Audit plan written, with objective, scope, dates, method and approvers.
• Register snapshot taken; known movements recorded.
• Duplicates and records without location cleaned up.
• Area lists prepared in walking order and loaded onto devices.
• Devices charged, app installed and tested, offline mode checked for areas without signal.
• Spare labels and cleaning materials packed for each team.
• Teams assigned to areas, with access arranged for locked rooms.
• Briefing held on unregistered items, wrong-area items and damaged tags.
• Cut-off communicated to everyone who moves assets.
• Reconciliation time booked in calendars for the days after the count.

Variances and reconciliation

Every audit produces the same categories:

• Found in the expected location.
• Found in a different location — update the register.
• Missing — in the register, not found.
• Unregistered — found, not in the register.
• Condition exceptions — damaged, idle, obsolete.

Work the variances in this order:

1. Match missing against unregistered by serial number, description and location. Replaced or damaged labels create many false pairs.
2. Check open check-outs, repair vendors and transfers for missing items.
3. Check disposal records — scrapped items often never reach the register.
4. Search likely locations — adjacent areas, overflow storage, vehicles.
5. Decide on unregistered items — add, flag as non-capitalized, or exclude (leased, third-party, personal).

The article on asset reconciliation walks through a full worked example.

Valuing variances

Report missing items both by count and by value. For capitalized assets, the value that matters for the accounts is net book value — that is what a write-off removes from the balance sheet. Showing original cost alongside is useful context. See net book value explained.

Approvals and write-offs

Write-offs should never be a spreadsheet edit. For each:

• Document the search performed and its results.
• Obtain approval from someone independent of the custodian, within defined authority limits (for example, higher values need more senior approval).
• Record the accounting entry and date.
• Keep the evidence with the asset record.

Where theft is suspected, follow your security and incident process before writing off.

Reporting

A good audit report is short and factual:

• Objective, scope, dates, method and teams.
• Counts by category: register records in scope, found (correct location), found (other location), missing, unregistered.
• Value of missing items at NBV and cost.
• Adjustments made and approvals.
• Root causes and corrective actions, with owners and dates.

Register accuracy

One headline metric summarizes the result:

Register accuracy = assets found in their recorded location ÷ register records in scope

Illustrative example: 1,200 records in scope; 1,090 found where the register said. Accuracy = 1,090 ÷ 1,200 = 90.8%. Track it audit over audit; it tells you whether your processes are improving.

Root-cause analysis

Variances are symptoms. Group them by cause and fix the process:

• Unrecorded transfers → make transfers a scan at the destination.
• Unrecorded disposals → make register update a mandatory step in the disposal process.
• Shared equipment going missing → introduce check-in/check-out.
• Assets never registered → tag and register at goods receipt.
• Label damage → better materials and placement.

Building a cycle-count program

Cycle counts spread the audit effort across the year. A common approach is risk-based (ABC) classification:

• A — high-value, portable or high-risk items. Count more often.
• B — medium value or risk.
• C — low-value, static items. Count least often.

Illustrative example: 3,000 assets classified as 300 A (counted quarterly), 900 B (counted twice a year) and 1,800 C (counted once a year).

• A: 300 × 4 = 1,200 asset-counts per year
• B: 900 × 2 = 1,800
• C: 1,800 × 1 = 1,800
• Total: 4,800 asset-counts per year, or 400 per month

Four hundred scans a month is a manageable, predictable workload — and every area is verified at least once a year. Schedule counts by location where possible, so each session covers a physical area completely.

Audit KPIs to track

• Register accuracy (as above).
• Missing asset value at NBV, per audit.
• Unregistered items found per audit.
• Time to complete count and reconciliation.
• Share of variances resolved within a target number of days.
• Cycle-count schedule adherence.

Audits in Asetavo

Asetavo's audit sessions are scoped by location, floor or category. Counters use the mobile app with QR, barcode, NFC or a UHF sled — online or offline — and counts build in real time. The variance report sorts every asset into found, missing and unregistered, and you resolve each item into the register with a full audit trail. Recurring audits support a cycle-count program, role-based permissions control who can run and reconcile sessions, and reports export to CSV and PDF.

Plan your count with the free audit checklist.