Create an Inventory Data Dictionary That Makes Reports Agree

We see reporting disputes start with labels that look clear but carry unstated assumptions. When we call a field "available," one team may mean stock physically present, another may subtract reservations, and another may include future supply.

We solve that problem by recording the formula, status inclusion, reporting grain (the level a number is counted at, such as SKU, bin, or warehouse), cutoff time, and accountable owner beside every label.

Indiana state guidance describes a data dictionary as a collection of names, definitions, and attributes about data elements that helps people interpret data consistently. We apply that same discipline to inventory reporting, where a field must be understandable to an analyst, a picker, a buyer, and a controller without a side conversation.

We recommend documenting these differences before we compare totals:

• Status scope: We state whether damaged, quarantined, reserved, or quality-held units belong in a field.
• Location scope: We specify whether a total represents a bin, warehouse, region, legal entity, or the full network.
• Timing rule: We record the event and timestamp that update the field, plus the report's refresh time.
• Unit of measure: We identify the base unit, conversion rule, and rounding method.
• Ownership: We name the business role that approves a definition and the system that supplies it.

We use custom reporting to make those rules visible in the same place as the totals. That way, we can answer "why does this report differ?" with a documented definition instead of a spreadsheet debate.

These three terms are often used interchangeably, but they answer different questions. A business glossary says what a term means to the business. A data dictionary records how a specific field is built. A data catalogue is the searchable index that tells people where a dataset or report lives.

In practice, fixing conflicting inventory reports does not require three separate documents. One register can carry the plain-language meaning and the calculation side by side, with a pointer to wherever the report lives. That is the register this guide helps us build.

We build the terminology map before we build formulas. Our goal is not to impose one universal vocabulary, because purchasing terms, customer promises, quality workflows, and ownership rules vary. Our goal is to make our chosen meaning explicit and stable.

Define the Core Quantity Fields

With inventory visibility, we can keep the location, status, and timing behind a number visible rather than treating a network total as self-explanatory.

Write Available Inventory as an Auditable Formula

We document available inventory as a formula, not a feeling: on-hand quantity minus committed quantity minus non-sellable hold quantity. We then define non-sellable holds as a non-overlapping set of statuses, so we do not subtract damaged and quarantined stock twice when a unit carries both labels.

We also keep incoming and in-transit quantities separate from immediate availability unless our documented promise policy says otherwise. That distinction lets us report what we can fulfil now alongside supply that may become usable later, such as open purchase orders handled in purchase planning.

Worked Example: Why Two Reports Show Different Available Stock

Take one SKU in one warehouse. None of the committed units are damaged or quarantined. The numbers are illustrative.

A report that subtracts damaged and quarantined stock separately would subtract 10 instead of 8 and show 60 available. The warehouse report that counts each unit once shows 62. Both reports follow a formula, but only one follows a documented, non-overlapping definition. Writing that rule into the dictionary is what makes the two numbers agree.

Link Commitments to Real Events

We define the exact event that creates a commitment, such as approved order reservation, allocation release, or pick confirmation. We also document the event that removes it, such as shipment, cancellation, expiration, or reallocation. Our order commitments workflow should use the same trigger logic as the report, or we will recreate the mismatch inside our system.

Treat Quarantine as a Controlled Hold

We use quarantined stock for items awaiting release, inspection, or disposition, not as a vague catch-all. The FDA's good manufacturing practice rules for finished pharmaceuticals (21 CFR 211.82) provide a useful model: components must be stored under quarantine until they have been tested or examined and released. We adapt that discipline by naming who can release the hold and what evidence they need.

Every quantity field should answer one precise question at one precise level of detail. We call that level the grain: whether a number is counted per SKU, per bin, per warehouse, per lot, or per day. A report that is correct by SKU can still be wrong by location, lot, unit of measure, or timestamp. We prevent that error by writing the grain directly into the field definition, not into a footnote.

Define Location and Time Boundaries

We state whether on-hand means a specific bin, a warehouse, a region, or every location. We also distinguish event time from dashboard refresh time. A receipt can be physically present before it is posted, and a report can refresh after either event. Our definition needs to say which state it represents.

Receiving and putaway are the events that move stock into a bin, so we name them as posting events in the entry. Our warehouse management workflows cover barcode-based receiving, pallet labels, and putaway into warehouse bins.

For traceable products, we connect the dictionary to lot tracking so we can explain whether a total includes a particular lot, serial range, expiration status, or recall hold.

Control Units of Measure and Packaging

We choose one base reporting unit for each item and store conversion factors for cases, packs, pallets, weight, and eaches. The GS1 standard reinforces why consistent package and unit definitions matter when trading partners exchange product data.

We document whether conversions happen before or after aggregation, how we round fractional quantities, and which unit the report displays. A case-level report and an each-level report can both be accurate while showing different totals if we do not identify their units.

Separate Finished Kits from Buildable Kits

We report finished kit stock separately from buildable kits. Finished-kit quantity counts assembled inventory already on hand. Buildable-kit quantity depends on available components and each component's bill-of-material ratio. We use kit rules to ensure our reports do not promise kits from components already reserved, quarantined, or assigned elsewhere.

We treat quantity and value as related but separate reporting systems. Quantity fields answer how many units we can locate, reserve, inspect, or promise. Valuation fields answer how those units are costed for accounting and financial analysis. Combining both questions in one field creates confusion when operational status changes before accounting treatment does.

We document cost method, cost layer, currency, landed-cost inclusion, valuation cutoff, and write-down treatment in their own dictionary entries. IRS Publication 538 states that, for merchandise purchased during the year, cost means the invoice price minus appropriate discounts plus transportation or other charges incurred in acquiring the goods.

We use that as a reminder to document our cost basis rather than assuming quantity multiplied by a visible unit cost equals the ledger balance.

We also avoid assuming that unavailable quantity has no value. Damaged or quarantined inventory may still carry a cost, a reserve, a repair estimate, or a pending disposition decision.

Our landed cost reporting should therefore reconcile to its own valuation logic while our operations reports remain clear about physical and sellable quantities.

If finance works in QuickBooks Online, Arka's QuickBooks integration handles the accounting entries required for cost allocations, which is worth documenting in the valuation entries of the register.

We use one shared register for report fields, formulas, and ownership. We can store it in a spreadsheet, reporting specification, or governed data catalogue, as long as every dashboard and export points back to the same approved definition.

Example: A Completed Entry for Available Inventory

Here is one filled-in entry. The values for Technical Field Name, Source System Of Record, and Approver are example placeholders; replace them with your own.

We use the NIST definition of a data element, the smallest named item of data that conveys meaningful information. In our register, that means each field gets one definition, one purpose, and enough context for another team to reproduce the number.

We recommend starting with the fields used in inventory by location, available-to-promise, open orders, receiving, and valuation reports.

To see what those reports can look like in practice, read how one Arka customer tracks availability and value by product family and monthly stock movements. Once the register is approved, we turn it into practical filters, views, and dashboards.

A dictionary only helps while it stays current. We roll it out in a few small steps, then control every change the same way.

Keeping a short change log (what changed, why, who approved, and the effective date) is what lets us answer "why did this number move?" without a spreadsheet debate.

For the wider disciplines that keep stock figures trustworthy, such as audits, barcoding, and FIFO, see our guide to improving your inventory management process.

Your dictionary only works if every report follows it. Arka Inventory is cloud software for inventory, warehouse, purchase and cost management, built natively on Salesforce to make shared reporting effortless. More than 30 paying B2B customers use it.

• Live stock visibility: See inventory across warehouses and locations, with commitments and lot or serial status alongside the totals.
• Workflows that follow your rules: Commit, allocate, and reallocate inventory using the same triggers your dictionary defines.
• One place for the numbers: Custom reports and dashboards keep every team looking at the same stock figures instead of rebuilding calculations in separate spreadsheets.
• Alerts tied to your thresholds: Inventory alerts notify users when items approach or fall below minimum stock levels, and when back-ordered items arrive.
• Results you can check: A motorhome dealership cut order cycles from over 72 hours to about 39 hours, a 46% reduction, on Salesforce, and a healthcare group unified inventory, field service, and finance across 14 locations.

If your dashboards disagree today, talk to our team. We will help you build a practical field register and test it against your real orders and receipts.