GenFrontier homeScope a project
Menu
Reader questionWhich reason codes and approval thresholds should we use for inventory adjustments so that the data explains losses and passes audit?

Inventory adjustment reason codes and approval rules

A minimal set of inventory adjustment reason codes tied to root causes, plus approval thresholds, segregation of duties, evidence rules and a weekly review.

field-notes / inventory-control / inventory-adjustment-reason-codes
A coiled garden hose with several small leaks, each leak marked by a red clothespin clipped beside it.

Use about a dozen inventory adjustment reason codes, each naming one cause someone can fix: transaction errors, physical losses with a known cause, unexplained losses and gains, and short-lived system corrections. Route approval by the value and the reason of each adjustment, and count the threshold per item over a rolling week so it cannot be split. Keep the person who counts, the person who approves and the person who reconciles the books separate. Then review the totals by code and location every week, so the adjustment history explains where stock went and an auditor can follow each line back to its evidence.

This field note is written for inventory control leads, controllers, internal auditors and the ERP administrators who configure the codes. It starts where a count variance investigation ends: the difference is confirmed, and someone has to post it.

What an adjustment is, and why “count correction” explains nothing

An inventory adjustment changes the record to match the shelf. The reason code on it should say why the two disagreed. A code that only says “the count changed” records the symptom and throws away the cause.

The ASCM Supply Chain Dictionary defines an inventory adjustment as “A change made to an inventory record to correct the balance in order to bring it in line with actual physical inventory balances.”

The same dictionary says the key purpose of cycle counting “is to identify items in error, thus triggering research, identification, and elimination of the cause of the errors.” The count finds the error. The reason code is where the cause gets written down.

Many systems go live with a code called something like count correction, cycle count adjustment or inventory variance. Every adjustment fits it, so within a few months it carries most of the volume. A report by reason code then tells you that inventory was adjusted because inventory was adjusted.

The three jobs a reason code does

A reason code earns its place only if it does at least one of these jobs:

  1. It explains the loss. Finance and operations can see which process lost stock: receiving, picking, handling damage, expiry or something no one has explained yet.
  2. It routes the money. In many ERPs the code selects the offset account, so damage lands in a damage expense account and an approved write-off lands against the obsolescence reserve.
  3. It triggers the right check. An unexplained loss deserves a different approver, and more evidence, than a receiving error with the receipt document attached.

A code that does none of the three is a label. Delete it.

Where this note starts

The investigation of a variance is its own discipline: recount, check open transactions, look for the stock in nearby locations, read the transaction history. That process is covered in how to investigate a cycle count variance before you adjust. This note starts at the end of that work, when the difference is confirmed and someone must decide which code to use, who approves it and what evidence goes with it.

A minimal reason-code set mapped to root causes

Use 10 to 15 codes in four families, and give each code one cause, one sign convention and one default offset account. The set below is a starting point. Rename the codes to fit your conventions, but keep the rule that each code names a cause someone could act on.

Code Family Use when Sign Typical offset (your controller decides)
RCV Transaction error A receipt posted the wrong quantity, item or lot and can no longer be reversed Plus or minus Inventory variance, or a vendor claim if the supplier shorted
PCK Transaction error A pick, pack or ship confirmation recorded a different quantity than left the shelf Plus or minus Inventory variance
UOM Transaction error Stock was received, picked or counted in the wrong unit, such as cases posted as eaches Plus or minus Inventory variance
PRD Transaction error Production consumption, output or scrap was reported wrong Plus or minus Production variance
DMG Known physical loss Stock was damaged in the building and destroyed Minus Damage expense
EXP Known physical loss Stock expired or spoiled and was disposed of Minus Spoilage expense
CLM Known physical loss A shortage or concealed damage is traced to a carrier or supplier after receipt Minus Claim receivable
SMP Known physical loss Stock was used for samples, testing or internal use without an issue transaction Minus Samples or internal-use expense
OBS Planned write-off Management decided to scrap, donate or sell off obsolete stock Minus Obsolescence reserve
UNL Unexplained A recount confirms a loss and the investigation found no cause Minus Shrink expense
UNG Unexplained A recount confirms a gain and no source document was found Plus Shrink expense (credit)
CNV System correction Opening balances are corrected during a cutover window Plus or minus Conversion clearing account

Twelve codes is a workable starting point. Add codes only when a different person would act on them.

A manufacturer may split PRD by line or cell. A food distributor may split EXP into expired and temperature abuse if the two have different owners.

Five rules that keep the list honest

  1. One cause per code. If a definition needs “or” to join two unrelated processes, split it.
  2. Name a cause someone can fix. “Inventory variance” is not a cause. UOM points straight at the item master or the receiving screen.
  3. No “other” code. An “other” code becomes the next count correction. If nothing fits, use UNL or UNG with a written note, and add a code at the quarterly review if the pattern repeats.
  4. Keep gains and losses apart. UNL and UNG are separate so a found pallet cannot quietly net against a missing one. Reviews should add up absolute values, not just the net.
  5. Give temporary codes an end date. CNV and any project code should stop being accepted after hypercare. If your system can deactivate a code, do it on the date you agreed.

Shrink is a result, not a reason

ASCM defines inventory shrinkage as “Reductions of actual quantities of items in stock, in process, or in transit,” and adds that “The loss may be caused by scrap, theft, deterioration, evaporation, and so forth.”

That definition is useful on a financial report. On an adjustment it hides the cause. Post the cause code, and let the shrink report add DMG, EXP, UNL and the rest together.

Do not give floor staff a “theft” code. An adjustment entered at a scanner is not the place to accuse anyone. Post UNL, escalate it through the approval matrix, and let loss prevention or management reclassify the line after an investigation.

Which differences should not be adjustments at all

Many differences that reach the adjustment screen are really a missing or wrong transaction, a location problem or a status problem. Fix those with the transaction that should have happened. A quantity adjustment is for stock that genuinely appeared or disappeared.

Choosing the code for a confirmed difference Recount confirms a difference Did a receipt, pick, issue orunit of measure post wrong? Is the missing stock sittingin another location? Is the stock here butunsellable or on hold? Is another cause proven: carrier,samples, approved write-off? Correct the transaction, or postRCV, PCK, UOM or PRD withthe document number. Post a bin-to-bin move. Thesite quantity does not change. Change its status. If destroyed,post DMG or EXP with photos. Post CLM, SMP or OBS with theclaim, request or approval. No cause found: post UNL or UNG with a required note, routed to the next approval tier up. NoNoNoNo YesYesYesYes
Decision path recommended in this field note, not a published standard. Code names follow the starting set in the table above.

Location problems are moves

If the missing 24 units are sitting in the next bay, the site quantity is right and only the location is wrong. Post a bin-to-bin move. An adjustment out of one bin and an adjustment into another produce the same balance, but they inflate both your loss and your gain totals and hide a location discipline problem behind a shrink number.

Status problems are status changes

Damaged stock that is still in the building, waiting for a vendor return or a disposal decision, should change status, not quantity.

Oracle’s store inventory product shows this design clearly. In its inventory adjustments documentation, a reason code such as Damaged - Hold carries a disposition of “+ Unavailable” that moves units from available to unavailable stock. The quantity on hand stays the same until the goods actually leave.

Missing transactions get posted, not adjusted

A receipt still sitting in a receiving queue, a transfer that left one site and has not arrived at the other, a production order not yet reported: each has a document.

Post the document. An adjustment would make today’s balance right and leave a duplicate waiting to post tomorrow. Negative balances are a special case of this, usually caused by the order in which transactions post, and are covered in negative inventory in your ERP.

How ERPs enforce reason codes: Dynamics 365 and Oracle as examples

Many ERP and WMS products can make a reason code mandatory, limit which codes a role may use, and post each code to its own account. The two products below are examples of the settings to look for, not a recommendation of either.

Dynamics 365 Supply Chain Management

Microsoft’s article on reason codes for inventory counting describes counting reason code policies that are either Optional or Mandatory. You assign a policy to a warehouse or to a product, and the product setting can override the warehouse. Where the policy is Mandatory, “the counting journal can’t be completed and closed until a reason code is provided.”

Each counting reason code can carry an offset account. When it does, “the value is posted against the assigned offset account instead of the default inventory posting profile account.” That is the mechanism behind the second job of a reason code: the code decides where the money goes.

The same article describes reason code groups, which limit the list a worker sees on the Adjustment in and Adjustment out mobile menu items. Each mobile menu item can also set a default reason code, show or hide the code, and allow or block editing.

Those mobile settings deserve a warning. A default code that is hidden from the worker produces exactly the count correction problem by configuration. Either show the code and make it mandatory, or send the variance to review and let the approver choose the code after investigation.

Variance review before posting

Microsoft’s cycle counting article describes a review step. For a worker who is not set up as a cycle count supervisor, a count outside the deviation limits on the work user record goes to Pending review. Its example: with 50 on hand and a limit of 10, a count outside 40 to 60 creates a difference.

A supervisor resolves it on the Cycle count work pending review page, and “The counted value is rejected if you don’t approve the difference.”

The same article notes that “The system never shows the expected quantity to count. This design prevents intentional miscounts.” Blind counting is a preventive control in its own right, and it matters more once approvals depend on the counted quantity.

Oracle Retail Store Inventory Operations

Oracle’s documentation states: “When creating a reason code, a disposition is assigned to the reason code and it defines how the stock will be moved.” Reason codes are therefore tied to stock buckets, such as available, unavailable and customer order reserved, rather than being free labels.

Access is also controlled per code. “The user will only see those reason codes in the list that the user has data permissions for.”

To confirm an adjustment, the user needs confirm permission and data permissions for every reason code on it. The documentation adds that “The Save button can be used to implement a Save and Approval process,” and the adjustment’s information screen records the create user, approve date and approval user.

Use the same idea in any system: floor roles get the transaction-error codes and the hold statuses, only inventory control can post UNL and UNG, and only finance can post OBS.

What to check in any ERP or WMS

Whatever you run, confirm these capabilities in a test environment before you write the policy around them:

  • A reason code can be made mandatory for each adjustment type: counting journals, mobile adjustments in and out, and quantity changes keyed on a screen.
  • Each code maps to its own offset account, and the mapping cannot be changed by the people who post adjustments.
  • The list of codes can be limited by role or by menu item.
  • An approval step can hold an adjustment before it changes the balance, route it by value and by code, and refuse self-approval.
  • Codes can be deactivated without deleting their history.
  • The audit log keeps the initiator, approver, before and after quantities and timestamps for every line.

If a capability is missing, write down the manual control that replaces it, who performs it and how often. That note is what an auditor will ask for.

When a 3PL holds the stock

If a third-party warehouse holds your inventory, its adjustments usually reach you electronically. X12 lists the 947 Warehouse Inventory Adjustment Advice, which “can be used to inform a warehouse/depositor or other business party of a quantity or status change to inventory records.”

Map each of the 3PL’s reason values to your own code list in the interface, and hold any unmapped value for review instead of posting it under a default. That mapping is typical work in an ERP and WMS integration project, and the question of which system posts the adjustment first is covered in ERP or WMS as the system of record.

Approval thresholds by value and by reason

Set approval tiers on the absolute extended value of each adjustment, then override them by reason: unexplained losses and gains escalate sooner, and planned write-offs always go to finance. Measure each threshold per item and location over a rolling period so a large difference cannot be posted in small pieces.

What published policies require

Corporate approval matrices are rarely public. University policies are, and they are written to satisfy auditors. Three examples, checked in November 2026:

Source Rule as published
Yale Policy 4210 Adjustments that “equal or exceed 5 percent or $10,000, whichever is greater, are considered significant and must be reported to the University Controller’s Office.” Departments “must investigate the reasons.” Write-offs “must be approved by the department Lead Administrator.”
University of Idaho APM 20.25 “Variances and inventory write-offs must be reviewed and approved by a next-level supervisor.”
University of Florida procedure Stock adjustments need “the signed approval of the accountable officer or designee who must not be the individual responsible for the inventory.” “Large adjustments shall be supported by an explanation.”

Three patterns repeat: a value or percentage triggers escalation, the approver sits above or outside the person who holds the stock, and larger adjustments need a written explanation.

Yale also asks that the journal entry say whether a gain or loss is due to “a physical count difference, obsolescence, or damage”, which is a three-code reason list in its own right.

The federal Green Book gives the principle behind tiers. Under its authorization category, “Transactions are authorized and executed only by persons acting within the scope of their authority,” and management “may require approval from multiple levels or units (multilevel authorization)” for transactions that present greater risk.

Example approval matrix by reason family and value
Reason familyUnder $500$500 to $5,000Over $5,000
Transaction error with a document (RCV, PCK, UOM, PRD)Inventory control leadInventory control supervisorOperations manager, controller notified
Known physical loss (DMG, EXP, CLM, SMP)Inventory control supervisorOperations managerController
Unexplained (UNL, UNG)Inventory control supervisorOperations manager and controllerController, loss prevention notified
Planned write-off (OBS)ControllerControllerController and CFO
Cutover correction (CNV)Project lead and controller sign the cutover reconciliation; no individual line approvals
Hypothetical thresholds for illustration, not a standard. Values are absolute extended cost, summed per item and location over a rolling seven days. The pattern (value escalation, an approver outside custody, a written explanation for large items) follows Yale Policy 4210, University of Idaho APM 20.25 and the University of Florida procedure, checked November 2026.

Setting the dollar bands

There is no standard dollar figure, and anyone who quotes one is quoting their own company. Start from your data instead:

  1. Export twelve months of adjustments with item, location, code, quantity, unit cost and user.
  2. Sort by absolute extended value and look at where the volume sits. If the first tier catches most lines but little of the value, the inventory control lead can clear routine items quickly while managers see what matters.
  3. Ask the controller what single unexplained loss would need to be explained to the CFO. That number, or something below it, is the top of the middle band.
  4. Check the approvers’ calendars. A tier that sends dozens of approvals a day to a controller invites rubber-stamping.

For items with high theft risk, serial-numbered items or controlled substances, use a zero threshold: every adjustment goes to the top tier for that family, whatever its value.

Value alone can miss a process failure on cheap items. For example, suppose a count finds 400 fewer fasteners than the record, at two cents each. That is $8 and clears the first tier, yet it means a count, a pick or a unit of measure went badly wrong.

Add a quantity or percentage trigger for low-cost items, such as any difference above a set share of the recorded balance, and route those lines to inventory control for a cause review even when the dollar tier is low.

Why found stock needs an approver too

Gains feel harmless, but they need the same approval as losses. A positive adjustment increases the inventory asset on the balance sheet. It can also hide an earlier loss: if stock left without a record and the shelf was refilled from an unrecorded receipt, a later gain makes the books look right while the supplier invoice and the missing goods tell a different story.

UNG in particular should route like UNL. A found pallet with no source document is as much a control question as a missing one.

Stop threshold splitting

An approval tier based on single lines invites splitting. Count the threshold on the sum of adjustments for one item at one location over a rolling seven days, or over the open accounting period, and route on that total.

Example (fictional). A count of item EX-20118 in bin 14-03-B finds 84 units against a record of 120. The unit cost is $42.50, so the difference is 36 units, or $1,530.

The investigation finds 24 units in bin 14-03-C, one level up. Those become a bin move, with no adjustment. The remaining 12 units have no explanation, so the analyst posts UNL for 12 units, or $510.

Under the example matrix above, $510 of unexplained loss goes to the operations manager and the controller. Two days earlier, the same analyst had posted UNL for 10 units of the same item at the same location ($425), which a supervisor approved on its own. The rolling total is $935, so the system routes the new line using the $935 total, and the controller sees both lines together.

Segregation of duties: initiate, approve, post

Three different people should count or initiate, approve, and reconcile the books for any adjustment above the lowest tier. No one who handles the stock should approve changes to its record. Where the site is too small for that, add a compensating control and write it down.

What the standards say

The 2025 revision of the Green Book, GAO’s Standards for Internal Control in the Federal Government (GAO-25-107721), lists segregation of duties among its common control activities. It describes it this way: “Management divides or segregates key duties and responsibilities among different people to reduce the risk of error, misuse, or fraud. This includes separating the responsibilities for authorizing transactions, processing and recording them, reviewing the transactions, and handling any related assets so that no one individual controls all key aspects of a transaction or event.”

Paragraph 10.22 frames it as separating activities “related to authority, custody, and accounting of operations.” The Green Book is written for federal entities, but GAO notes it may also be adopted by nonfederal entities as a framework, and it adapts the principles of COSO’s Internal Control - Integrated Framework.

The University of Colorado’s inventory internal controls guide applies the same idea to stock. It calls for “Adequate separation of duties between those responsible for the physical inventory (e.g., ordering, receiving, distributing/selling) and those responsible for the inventory accounting records.” The records side includes “maintaining the perpetual inventory balance in the Finance System, and reconciling the Finance System.”

Who does what

Role Does Must not
Counter (custody) Counts blind, records the quantity, notes what was seen Approve or post their own count
Inventory control analyst (initiates) Investigates, chooses the code, attaches evidence, submits Approve their own adjustment
Approver (authority) Approves or rejects within their tier Approve adjustments they created, or on stock they handle daily
Accounting (records) Reconciles posted adjustments to the general ledger by account Change quantities or codes
Reviewer (monitoring) Runs the weekly review and the access review Post adjustments

Enforce it in the system, not in the policy binder

A rule that depends on people remembering it is hardest to keep at 2 a.m. on a peak weekend. Configure the system so that:

  • A user cannot approve an adjustment they created. Find the setting that blocks it and test it in user acceptance testing.
  • Adjustment entry and adjustment approval are separate security roles, and a quarterly access review lists every user who holds both.
  • Changes to the reason code list and to offset accounts are restricted to finance or a system administrator, with a change log.
  • Counts are blind, so the counter never sees the expected quantity.

Small sites and night shifts

Paragraph 10.23 of the Green Book covers the common case where there are not enough people: “If segregation of duties is not practical within a business process because of limited personnel or other factors, management designs alternative control activities to mitigate the risk of fraud, waste, or abuse in the business process.”

Alternatives that work at a small site include a recount by someone from another shift, remote approval by a regional manager, and a weekly finance review of every adjustment at that site regardless of value. The University of Idaho’s procedure offers two more for full counts: “Assign two persons where possible to conduct inventory counts,” and have “a staff member not involved in the physical count spot-check the physical counts” for significant items.

Evidence each adjustment needs

Each adjustment must let someone who was not there understand what happened, who decided, and why. The minimum is a complete transaction record plus evidence that matches the code.

The Green Book’s documentation category sets the bar: “Management clearly documents the performance of control activities and all transactions and other significant events that occur in a manner that allows the documentation to be readily available for examination.” The University of Colorado guide says the same about counts: “All differences should be investigated and explained.”

Fields on every adjustment line

  • Item, site, location, lot or serial number, and status.
  • Quantity before, quantity counted and quantity after, with the unit of measure.
  • Unit cost and extended value at the time of posting.
  • Reason code, and a free-text note (required for UNL, UNG and anything above the first tier).
  • Initiator, approver, approval time and posting time, as separate fields.
  • The count task, recount task or source document it came from.

Evidence by code family

Family Evidence to attach or reference
Transaction error (RCV, PCK, UOM, PRD) The original document number (receipt, order, shipment, production order) and what was wrong on it
Damage and expiry (DMG, EXP) Photos, the disposal or destruction record, and the lot and expiry date for EXP
Claims (CLM) Carrier or supplier claim number, and the delivery receipt with exceptions noted
Samples (SMP) The internal request and who received the stock
Write-off (OBS) The approved write-off decision and the disposal, donation or sale record
Unexplained (UNL, UNG) Recount result, the investigation note, locations searched and transactions checked
Cutover (CNV) The signed cutover reconciliation that lists each corrected balance

Keep this evidence as long as your record-retention policy requires. As one published example, the University of Idaho retains count sheets and reconciliations “for a period of 3 years past the ending date of the fiscal year in which the physical inventory was performed.”

The weekly trend review

Once a week, someone who does not post adjustments should read the totals by code, location, item and user, and assign an owner to every pattern that repeats. That review is what turns a stack of approved lines into fewer adjustments next month.

The Green Book draws the line between the two activities in its Appendix II. Personnel “may perform a control activity to reconcile inventory, which identifies, investigates, and resolves individual discrepancies.”

Management may then perform “a corresponding monitoring activity” that checks the reconciliation was done properly “and notes any trends in the discrepancies identified to address a deficiency in the process.” Approving each line is the control. The weekly review is the monitoring.

A 30-minute agenda

  1. Totals. Gross loss, gross gain, net value and number of lines for the week and the last four weeks, by code.
  2. Top locations and items by absolute value. Look for a single aisle or item family that carries a large share of the total.
  3. Repeat items. As a starting rule, flag any item adjusted three or more times in four weeks for a root-cause review instead of another count.
  4. Pairs. A loss in one location and a gain of the same item elsewhere in the same week points to put-away or move discipline, not shrink.
  5. By user and shift. Who initiates and who approves. A single approver clearing everything is a control finding, whatever the totals say.
  6. Open items. Adjustments pending review for more than a few days. Stale reviews mean the stock and the record disagree while orders are being promised.
  7. Actions. Each pattern gets an owner, a fix and a date, recorded in the minutes.
Four weeks of gross loss value by reason code UNL unexplained UOM unit error PCK pick error DMG damage RCV receiving EXP expired SMP samples $6,840 $3,920 $2,710 $1,880 $1,150 $640 $310 Gross loss value at standard cost, four weeks, one site
Fictional example data for illustration. In this example, unexplained loss is the largest code and unit-of-measure errors are second, which points the review at investigation quality and at item master units rather than at counting more often.

Reading the pattern

In the fictional chart, unexplained loss is the largest code. That is a finding about the investigation as much as about the stock: either the analysts are not finding causes, or something is leaving without a record. Sample ten UNL lines and read their notes before deciding which.

Unit-of-measure errors in second place usually lead back to the item master: a case quantity that changed without anyone updating the system, or a receiving screen that defaults to eaches. Item master data readiness covers how to find and fix those fields.

Track accuracy next to the adjustment totals. ASCM’s definition of inventory accuracy measures the percent of items within tolerance, which tells you how many records are wrong, while the adjustment report tells you why. The inventory accuracy calculator computes the share of count lines within tolerance, the total unit variance and the net value of the differences from a list of counted items.

What auditors and controllers will ask for

Expect requests for the population of adjustments for a period, a sample traced back to evidence and approvals, proof that the approval rules work in the system, and proof that someone reviewed the trend. Build the reports for those requests before the first audit, not during it.

The usual requests

  • The population. Every adjustment in the period, with code, value, initiator, approver and timestamps, tying to the general ledger accounts it posted to.
  • A sample traced to evidence. For each sampled line: the count, the recount, the investigation note, the approval inside the right tier.
  • Configuration proof. Screenshots or exports showing that the code is mandatory, that self-approval is blocked and how tiers are set.
  • Access review. Who held adjustment entry and approval roles, and what was done about conflicts.
  • Change log. Who added, changed or deactivated reason codes and offset accounts.
  • Monitoring evidence. Minutes of the weekly review with actions and owners.

The Green Book lists physical control over vulnerable assets, including inventories, among its control activities, and adds that “Management periodically counts and compares such assets to control records.” If your counting program is meant to replace an annual wall-to-wall count, expect auditors to look at the adjustment controls as part of that judgment. Cycle counting versus the annual physical inventory covers what has to be true for that to work.

Prefer preventive controls

The 2025 Green Book revision emphasizes prioritizing preventive control activities. Its paragraph 10.12 says management designs a mix of preventive and detective controls, “prioritizing preventive control activities where appropriate.”

The same paragraph says preventive controls “generally offer the most cost-efficient use of resources.” For adjustments, preventive means a mandatory code, a blocked self-approval and a blind count. Detective means the weekly review, and you need both.

Rolling out reason codes and approvals

Run the change as a short project: measure what you have, agree the code list and tiers with finance, configure and test, train with real examples, and hold the first four weekly reviews with the project team present.

Step by step

  1. Export the history. Twelve months of adjustments with code, value, location, item and user.
  2. Classify a sample. Take 50 to 100 lines from the catch-all code and read the notes, documents and transaction history to see which causes are hiding in it.
  3. Draft the code list from the families above, with a one-sentence definition, a sign and an offset account for each.
  4. Set the tiers with the controller, using the history to check how many approvals each approver would receive.
  5. Configure roles and workflow. Separate entry and approval roles, block self-approval, restrict code maintenance, and set the rolling-total rule.
  6. Test it. Write test cases for each tier boundary, for self-approval, for a split adjustment and for an unmapped 3PL reason value.
  7. Train with examples. Walk counters and analysts through ten real differences from your own history and agree the code for each.
  8. Run the reviews. Hold the weekly review from the first week, and revisit the code list after one quarter.

In our engagements we start by writing these rules as numbered requirements with acceptance tests before any configuration, and we deliver the code definitions, the approval matrix and the review agenda as documents your team keeps.

Checklist

  • Every code names one cause and has a sign, an offset account and an owner.
  • The catch-all and “other” codes are deactivated.
  • Location and status problems are handled as moves and status changes.
  • Tiers use absolute value, reason overrides and a rolling total.
  • Gains route like losses.
  • Self-approval is blocked and tested.
  • Evidence requirements are listed per code and enforced for UNL and UNG.
  • A weekly review runs with minutes, actions and owners.

GenFrontier is a Kansas City firm that builds inventory and shipment-tracking software around your existing ERP, WMS and TMS. If your adjustment controls need an inventory layer with these rules built in, our inventory management software page describes that work, and you can size the counting workload that feeds it with the cycle count planner. To talk through your own code list, scope a project with us.

Working on this problem?

If this field note describes a process you need to fix or build software around, tell us about it. We start with your process, your systems and your deadline.