# Tag assets at intake so the register starts complete > Decide where identity begins, choose an identifier that carries no changing facts, and make the first record at intake the authoritative one. Canonical page: https://datacenterlabeling.com/guides/asset-tagging-at-intake Source review: 2026-09-12. This is a source-review date, not a claim of standards certification. Rackstamp provides editorial field guidance. IDs, cases, and worksheet entries are fictional examples unless identified otherwise. Distinguish local conventions from adopted standards. Source notes describe the evidence scope; a linked overview is not the full standard or a project approval. ## The practical answer Tag equipment at the point it becomes your responsibility, usually receipt, and do not let it pass that point untagged. Use a fixed-width identifier that encodes nothing changeable, excludes look-alike characters, and is never reused. Decide the check-character question before the first print run, define placement per equipment class so the tag stays readable in service, and make the intake record carry identifier, serial, owner and destination so later processes update it rather than diverging from it. ## Decide where identity begins Most inventory gaps are not created in the data hall. They are created at the point where equipment arrives and nobody has yet decided who owns it, what it is called, or which record it belongs to. Equipment that reaches a rack unrecorded is far harder to capture afterwards, because by then it is powered, cabled, in service, and inconvenient to disturb. Choose the moment identity begins and write it down. For most organizations the defensible answer is receipt: the point where the equipment becomes the organization's responsibility and a receiving record already exists. Tagging at receipt lets the physical identifier, the purchase record, the serial number, and the eventual location all attach to the same object from the start. The alternative positions are legitimate but should be chosen deliberately rather than by drift. Tagging at staging suits organizations that receive through a third party. Tagging at installation is the weakest option, because the window between arrival and service is exactly where equipment becomes invisible. Whichever point is chosen, the rule should be that equipment does not move past it untagged. ## Choose an identifier that does not try to be clever An asset identifier answers one question: which object is this. Resist encoding location, owner, function, purchase order, or hardware type into it. Every encoded attribute is a fact that can change, and when it changes you are left choosing between a wrong label and a renumbering exercise. Location and role belong in the record, where they can be updated; the identifier should be stable for the life of the object. Keep the format fixed width and machine friendly. Decide the character set deliberately and exclude glyphs that are read back incorrectly by people: the pairs of letter O with digit zero, letter I with digit one, and in handwriting letter S with digit five cause most of the transcription disputes. Decide case handling once. Decide whether the printed human-readable string matches the encoded symbol exactly, and make sure it does, because a barcode that disagrees with the text under it produces a discrepancy nobody can resolve in the field. Never reuse a retired identifier. Reissuing a number that previously belonged to a disposed asset corrupts the history of both objects and produces audit findings that are unusually hard to explain. ## Decide the check character question before printing anything A check character is a digit or letter computed from the rest of the identifier so that a single mistyped or misread character can be detected. It is worth adding when identifiers are frequently keyed by hand, read aloud, or transcribed between systems. It is unnecessary overhead when every read is a scan and the identifier never leaves the symbol. Make the decision before the first batch is printed, because adding a check character later means either two coexisting formats or a full reprint. If you adopt one, record the algorithm in the labeling standard, implement it in whatever generates the identifiers, and make sure the receiving system validates it rather than merely storing it. A check character that nothing verifies adds length without adding protection. ## Place the tag where it can be read years later Placement is a decision about the future, not about the moment of tagging. The question is whether someone can read or scan the tag during a routine audit, without removing the equipment, disturbing cabling, or bringing a mirror. A tag applied to a convenient flat surface at the receiving bench is often on the top or rear face, which becomes unreachable the moment the device is racked between neighbours. Decide placement per equipment class rather than per unit: servers, switches, storage chassis, power distribution units, and rack furniture each have a face that stays visible in service. Write the placement rule into the intake procedure with a photograph, because a written description of a location is interpreted differently by different people. Avoid vents, moving parts, hot surfaces, existing manufacturer markings such as serial and regulatory labels, and any surface that is removed during maintenance. Where equipment is deep, dark, or dense, a second tag on a reachable face is cheaper than a failed audit. Record that a duplicate placement exists so the two are not later mistaken for two assets. ## Make the intake record the authoritative one Tagging is only half the control. The record created at intake should carry the identifier, the manufacturer, model, serial number, purchase or receipt reference, the responsible owner, and the intended destination, and it should be the record that later processes update rather than a parallel list that quietly diverges. Systems that support a component inventory expect exactly this kind of accountability detail; the configuration management control family in [NIST SP 800-53](https://csrc.nist.gov/CSRC/media/Projects/risk-management/800-53%20Downloads/800-53r5/SP_800-53_v5_1-derived-OSCAL.pdf) is written around an inventory that reflects the system and is kept current, and that is much easier to achieve when the first record is complete. Capture the serial number at intake even when the asset identifier is the primary key. Serial numbers are how warranty, support contracts, recalls and disposal records reconcile, and collecting them later means handling the equipment again. Scan rather than type wherever a manufacturer symbol is present. Decide what happens when intake is bypassed. There should be a defined, unembarrassing route for registering equipment that arrived through an exception, because the alternative is that it never gets registered at all. Vendor audit workflows that scan a population and reconcile it against the record, such as the workflow described in [Sunbird's asset audit application note](https://www.sunbirddcim.com/sites/default/files/AN011_Sunbird_Application_Note_Asset_Audit_0.pdf), work far better when exceptions are rare and expected than when they are the normal path. ## Worked fictional case: the pallet that arrived during a change freeze In fictional DC01, twelve switches arrive during a period when the usual receiving clerk is unavailable. They are moved directly to a staging cage and, two weeks later, eight are installed. Four remain in staging. None were tagged at receipt. The recovery costs more than the original tagging would have. The eight installed units are now in service between neighbours; three have their most convenient face against a cable manager, so the placement rule cannot be followed and a second placement is improvised. Two serial numbers require a maintenance window to read. The four staged units are straightforward, but nobody can now say with confidence whether the pallet contained twelve or thirteen devices, because the delivery note was filed against a purchase order covering several shipments. The corrective action recorded is not twelve tags. It is a change to the intake procedure: a named backup receiver, a rule that equipment does not enter staging untagged, and a weekly check of the staging cage against open receipts. The quantities and dates here are fictional. ## Handle equipment that arrives already tagged Equipment often arrives carrying somebody else's identifiers: a manufacturer asset tag, a reseller's sticker, a leasing company's barcode, or a previous owner's number on refurbished hardware. Decide in advance which of these are recorded and which are removed, and never simply add your tag beside three others and hope the right one gets scanned. The workable rule is that exactly one identifier is the operational one, it is visibly the primary tag, and any retained foreign identifier is recorded as an attribute of the asset rather than as a second identity. Where a leased or loaned asset must keep the owner's tag, record the relationship and the removal obligation, because disposal and return processes depend on it. ## Acceptance checklist - [ ] The point where identity begins is defined, and equipment does not pass it untagged. - [ ] The identifier carries no location, owner, or function, and retired values are never reused. - [ ] Character set, width, and case are fixed, with look-alike characters excluded. - [ ] The check-character decision is made and recorded before the first batch is printed. - [ ] Human-readable text and encoded symbol are verified to match on a printed sample. - [ ] Placement is defined per equipment class, with a photograph, and avoids vents and service panels. - [ ] The intake record carries identifier, serial, model, owner, receipt reference, and destination. - [ ] A defined route exists for registering equipment that bypassed intake. - [ ] Foreign identifiers have a stated treatment and only one tag is operationally primary. ## Frequently asked questions ### Should the asset identifier encode the location? No. Location changes and identifiers should not. Encoding a rack or room into the number forces a choice between relabelling every move and tolerating labels that assert something untrue. Keep location in the record, where updating it is a normal transaction rather than a reprint. ### Do we need a check character? Only if identifiers are keyed or read aloud in practice. In scan-only workflows it adds length without benefit. If you adopt one, decide before printing, record the algorithm in the labeling standard, and make sure the receiving system actually validates it rather than storing an unverified character. ### What about equipment that never enters a rack, like spares and test units? Tag them on the same rule. Spares and test equipment are common sources of inventory error precisely because they move without a change record. A stored spare with an identifier and a location record is far easier to reconcile than an untagged box that everyone assumes belongs to someone else. ### Can we tag at installation instead of receipt? You can, but expect a higher loss rate. The interval between arrival and installation is where equipment is most easily misplaced, borrowed, or installed by someone outside the process. If installation is the only practical point, add a compensating check on received-but-not-yet-installed stock. ### Who owns the intake tagging step? Whoever owns the receiving process, with the inventory owner defining the format and placement rules. Splitting it so that receiving applies a label whose structure nobody owns produces consistent tags that carry inconsistent meaning. Name both roles in the procedure. ### How do we deal with refurbished hardware covered in old tags? Decide the treatment before the pallet arrives. Record any identifier you are contractually required to keep, remove or clearly overmark the rest, and make your own tag visibly primary. The failure mode to avoid is four barcodes on one chassis and no rule about which one a scanner should read. ## Sources and applicability - [NIST: SP 800-53 official control catalogue](https://csrc.nist.gov/CSRC/media/Projects/risk-management/800-53%20Downloads/800-53r5/SP_800-53_v5_1-derived-OSCAL.pdf): Official control catalogue containing the configuration management family, including the system component inventory control and its accountability detail. Cited for the expectation that an inventory reflects the system and is kept current, not as a compliance interpretation. - [Sunbird: Asset audit application note](https://www.sunbirddcim.com/sites/default/files/AN011_Sunbird_Application_Note_Asset_Audit_0.pdf): Vendor application note describing a scan-based audit and reconciliation workflow. Product-specific; used here to illustrate why a complete first record reduces later reconciliation effort. - [GS1: RFID read range](https://support.gs1.org/support/solutions/articles/43000734166-what-is-the-read-range-for-a-typical-rfid-tag-): Publisher guidance on factors affecting read range. Referenced for placement considerations where scanning technology is used; it does not establish a placement rule for any specific equipment. ## Related guidance - https://datacenterlabeling.com/problems/asset-versus-location - https://datacenterlabeling.com/problems/duplicate-identifiers - https://datacenterlabeling.com/problems/barcode-qr-scan-quality - https://datacenterlabeling.com/problems/label-record-reconciliation - https://datacenterlabeling.com/problems/moves-and-retirement-closeout