What a Good Asset Code Looks Like
A good asset code is short, unique and permanent, and it carries almost no meaning.
The number identifies the asset; the register describes it. That separation is the whole design principle, and most numbering schemes fail because they ignore it. Six to eight characters are enough to number every asset a mid-sized company is likely to own, and a code of that length can be read aloud over a phone without confusion.
Location, department, cost centre, purchase year and asset class all stay out of the code, because each of them can change while the asset stays exactly where it is. When one does, the code either starts telling a lie or has to be reissued, and a reissued code breaks the trail the number exists to protect. Meaning belongs in fields that can be edited.
Code, register or reference field?
Where each piece of information about an asset belongs.
In the code
- A running sequence
- At most one prefix that never changes
- A check digit, if the scheme uses one
In the register
- Location and custodian
- Asset class and cost centre
- Purchase date, cost and status
In a reference field
- Legacy codes from earlier schemes
- The maker's serial number
- Numbers an insurer or lender holds
Intelligent Codes and Why They Break
An intelligent code carries meaning in its structure, most often the location. It is the scheme almost every business tries first, because it reads well on a printout: plant, then department, then a sequence, and the reader knows where the asset is without opening the register.
First failure
The asset movesA machine transferred to another plant now carries a code saying it is somewhere it is not. Leave the code and every later reader is misled; reissue it and the asset's history sits under one number and its present under another.
Second failure
A segment runs outEach segment is given a fixed number of digits. When a site outgrows its block, the only remedy is to re-code: relabel every asset at that site and rebuild the mapping between old and new numbers.
Over time
Nobody trusts the codesA few years of ordinary reorganisation leave a register full of codes that disagree with the location field, and staff stop relying on either.
The appeal of an intelligent code is a printout that can be read without the register. A report that pulls the location from the register gives the same convenience without fixing the location into the number.
Sequence, Check Digits and Length
A plain sequence avoids every problem an intelligent code creates, but it needs three design decisions taken on purpose.
- Block allocation. Where several sites will tag at the same time, give each a reserved range within one continuous sequence, so two teams never issue the same number. The block is an administrative convenience, not a meaning: once the numbers are issued, nothing about the site is encoded in them.
- A check digit. A number copied by hand, read over a phone or typed from a worn label will sometimes be wrong. One extra digit calculated from the others by a standard scheme catches every single wrong character and most swapped pairs of neighbouring characters at the moment of entry, instead of at the next reconciliation. On a large estate it is worth the extra character.
- Length. The code has to fit legibly on the smallest tag the estate needs, survive printing at that size and stay readable after wear. In practice that keeps most schemes well short of what the sequence alone would allow.
Add the prefix and the check digit to the length before choosing tag sizes: a code that only fits the larger tag forces the larger tag onto every small asset.
Mapping an Existing Register
Most numbering projects start from a register that already has identifiers, and those are the hardest part of the work.
- Legacy codes that must survive: numbers stamped into castings, codes in maintenance contracts, identifiers an insurer or a lender already holds. Keep them in a separate field beside the new code, so the new number identifies the asset and the legacy field keeps the reference. Forcing both into one number is where schemes collapse.
Separate legacy field - Duplicates, usually from site-level sequences that were never reconciled. Trace each one to the assets involved and reassign one of them before anything is printed.
Resolved before printing - Gaps, which are investigated rather than closed: a gap may be a disposal nobody posted.
Explained, not filled - Two schemes in parallel during the changeover. Hold both codes against each line, state on every report which scheme it uses, and fix a date after which the old code is reference only.
End date fixed in advance
Projects that leave the changeover open-ended never finish it. The columns a fixed asset register needs, including the legacy code field, are set out in a separate guide.
What the Auditor Tests
An auditor applies three tests to a numbering scheme, and a scheme that passes all three is normally accepted without further work.
- Unique across the whole estate, not only within a site or department. Schemes that restart at each location create duplicates the moment two sites are combined in one register, and the duplication stays invisible until somebody sorts the file.
- The number on the asset matches the number on the register line, tested in both directions: from selected lines out to the assets, then from selected assets back to the lines. Only the second direction finds an asset that was never recorded. That two-way test is the core of a fixed asset verification.
- Gaps and reissues explained. A gap is unremarkable if a spoiled label or a disposal accounts for it, and a finding if nobody can say what happened to the number. A number reissued to a second asset after the first was disposed of breaks the trail for good, because two assets then share one identity in the history.
Setting the Convention Before You Buy Tags
Four decisions come before a single label is ordered, because each is expensive to reverse afterwards:
- How long the code is, which fixes how many assets the scheme can ever hold.
- Whether any prefix is carried and what it identifies; anything that can change does not belong there.
- Whether the sequence runs continuously across the estate or restarts anywhere.
- Who issues new numbers, so two people never allocate the same one.
Write the convention down in a short note covering those four points, the format with a worked example, and the rule for disposals: a disposed asset's number is retired, never reissued. That note lets someone joining in three years continue the scheme instead of starting a parallel one, which is how most registers end up with two incompatible conventions. The tagging process that follows the convention is covered in its own guide.
An external team is worth considering where the estate spans several sites with incompatible existing schemes, or where the register has to be rebuilt at the same time. Asset tagging can then settle the convention and the reconciliation together rather than one after the other.
