How do you set up a vendor-managed inventory (VMI) program?

Setting up a vendor-managed inventory (VMI) program means defining the rules under which your supplier monitors agreed stock positions and makes replenishment decisions on your behalf. This guide covers every operational step: governance, data requirements, replenishment logic, exception handling, and measurement. Aksulit Oy builds the inventory systems that make this practical.

What is vendor-managed inventory, and how does it differ from similar arrangements?

VMI is a supply chain arrangement in which the supplier, operating under buyer-approved rules, takes responsibility for monitoring inventory and deciding when to replenish. The buyer does not raise purchase orders for routine replenishment; instead, the supplier acts when defined thresholds are crossed. Ownership of the stock is a separate question: VMI inventory can be buyer-owned from the moment it arrives, or it can remain supplier-owned until consumed — the latter arrangement is what consignment stock means in B2B. Confusing the two leads to accounting errors and unclear liability.

The table below compares VMI with consignment stock and traditional replenishment across the dimensions that matter most when choosing an approach.

Dimension Traditional replenishment VMI (buyer-owned stock) VMI with consignment
Who monitors stock? Buyer Supplier Supplier
Who raises the replenishment? Buyer raises PO Supplier proposes or auto-orders under agreed rules Supplier proposes or auto-orders under agreed rules
Stock ownership before use Buyer (on receipt) Buyer (on receipt) Supplier (until consumed or transferred)
Cash tied up in stock Buyer Buyer Supplier
Inventory risk (obsolescence) Buyer Buyer Shared or supplier, per agreement
Data sharing required Minimal High — real-time stock and consumption High — real-time stock and consumption
Admin burden on buyer High Lower, but governance overhead shifts in Lower, but billing reconciliation adds complexity

Readiness checklist before you start

VMI fails most often because the foundations are not in place before the arrangement goes live. Work through this checklist before approaching a supplier about a VMI pilot.

  • Inventory data accuracy: Your stock records must reflect physical reality. If your system-on-hand figures regularly differ from physical counts by more than a small margin, fix this first. A supplier making replenishment decisions from inaccurate data will over- or under-supply.
  • Item master completeness: Every item in scope needs a stable identifier, a defined pack size, a minimum order quantity, and a confirmed lead time. Missing master data is a leading cause of replenishment errors.
  • Consumption visibility: You must be able to report actual consumption events, not just orders received. If your only data is inbound deliveries, the supplier cannot distinguish demand from stock build-up.
  • A trusted supplier relationship: VMI requires the supplier to act on your behalf. That requires mutual trust, open data sharing, and a willingness to resolve discrepancies without blame.
  • Internal process owner: Someone on your side must own the VMI program. Without a named owner, governance gaps appear quickly.
  • IT capability for data sharing: Confirm whether your inventory system can export real-time data via API, file transfer, or a shared portal. If it cannot, resolve this before agreeing to go live.
  • Finance alignment on ownership and billing: Agree internally on how stock will be owned, how invoices will be triggered, and how discrepancies will be reconciled before you commit to a supplier.

How do you set up a VMI program? Ten steps from scope to scale

Each step below builds on the previous one. Skipping steps to move faster is the most common reason VMI pilots fail to scale.

  1. Define scope and objectives. Choose one supplier and one bounded product category with recurring, predictable demand. Document what you are trying to achieve — for example, reducing stockout frequency, cutting manual ordering time, or improving inventory turnover — and set a baseline measurement for each objective before you change anything. Avoid starting with your most complex, high-value, or irregular stock.
  2. Select your supplier partner. The right supplier for a VMI pilot has an established relationship with you, stable lead times, reliable delivery performance, and the operational capability to receive and act on real-time stock data. Discuss their internal capacity before committing. A supplier who cannot resource the planner role will not sustain the arrangement.
  3. Negotiate and document the operating model. This is the most critical step and the one most often rushed. The written operating model must cover: products and locations in scope; inventory ownership at each stage; min/max levels or reorder rules for each item; confirmed lead times and pack sizes; service level expectations; the mechanism for replenishment proposals and orders; the emergency replenishment process; liability for damaged, expired, or excess stock; how discrepancies are identified and resolved; data access rights and confidentiality; review cadence and escalation contacts; and exit terms including how stock is returned or bought back if the arrangement ends. Do not go live without this document signed by both parties.
  4. Agree on governance roles. Name a person in each role before the pilot starts. See the governance table in the next section for the full role set.
  5. Build the data sharing infrastructure. Your supplier needs a live view of your inventory position for every item in scope. Define the system of record — the single authoritative source of stock truth — and confirm how data will flow from it to the supplier. Specify message formats, update frequency, acknowledgement requirements, how duplicate events will be handled, and what happens when connectivity is lost or events arrive late. Define the reconciliation process: how often will both sides compare records, and who resolves gaps? For more on how automated tracking supports this, see how RFID inventory management works.
  6. Build the item data dictionary. For each item in scope, confirm and document: item ID and description; storage location; usable on-hand quantity; reserved and quarantined quantities (these must be excluded from the replenishment calculation); open replenishment quantity already in transit; consumption rate and consumption history; returns and adjustments; lead time; pack size; minimum order quantity; and the timestamp of the last confirmed stock event. See the data dictionary section below for the full field list.
  7. Set replenishment rules and thresholds. For each item, calculate the reorder point and, if using min–max logic, the maximum. Document the decision logic: alert only, supplier proposal requiring buyer approval, or automatic order. Make sure the supplier understands that pack size and minimum order quantity constraints apply to every replenishment proposal. See the replenishment rules section below for a worked example.
  8. Run a time-limited pilot. Go live with one location or one product category. Set a defined pilot period — typically eight to twelve weeks. Monitor replenishment accuracy, stockout events, data latency, and discrepancy frequency weekly. Hold a structured mid-pilot review with the supplier before deciding to continue.
  9. Measure against baseline and review. At the end of the pilot, compare your KPIs against the baseline you set in step one. Use the scorecard in the measurement section below. Identify which failure modes appeared, how they were resolved, and what rules need adjusting before you expand. See also warehouse analytics for purchasing decisions for methods to analyse the consumption data the pilot generates.
  10. Scale deliberately. Add locations or product categories one at a time. Each expansion should repeat the scoping, threshold-setting, and data verification steps for the new items. Do not assume that what worked in the pilot will transfer automatically to a different category or site.

Governance roles and RACI

VMI is a collaborative arrangement that requires named owners on both sides. The table below defines the core roles. Fill in actual names before the pilot goes live.

Role Party Primary responsibilities RACI on replenishment decisions
Process owner Buyer Owns the VMI operating model, approves rule changes, chairs reviews Accountable
Supplier planner Supplier Monitors stock positions daily, generates and submits replenishment proposals or orders, manages exceptions Responsible
Warehouse / site owner Buyer Maintains physical stock accuracy, confirms receipts, reports discrepancies Consulted
IT / integration owner Buyer (and/or supplier) Maintains data feeds, monitors latency and errors, manages reconciliation Consulted
Finance / billing owner Buyer Reconciles invoices against confirmed consumption or receipts, resolves billing disputes Informed
Escalation contact Both Named individual on each side with authority to resolve disputes outside normal process Accountable (for resolution)

Data dictionary and system ownership

The supplier can only make good replenishment decisions if the data they receive is complete, accurate, and timely. The following fields are required for each item in scope. Define which system is the authoritative source for each field, and confirm that the value shown to the supplier matches the system of record at the agreed update frequency.

Data field Definition Why it matters for VMI
Item ID Stable, unique identifier for the product Prevents mismatched replenishments
Storage location Physical location within the site Required for multi-location VMI; determines which threshold applies
Usable on-hand Quantity physically present and available for use The primary input to the replenishment calculation
Reserved quantity On-hand stock already allocated to a job, order, or project Must be excluded from usable on-hand; ignoring it causes over-supply
Quarantined quantity Stock held pending quality inspection or return Must be excluded; quarantined stock is not available for use
Open replenishment Quantity already ordered or in transit Prevents duplicate replenishment proposals
Consumption Quantity used or issued in a defined period Drives demand forecasting and threshold calibration
Returns and adjustments Stock returned to inventory or adjusted by count Keeps on-hand figure accurate; unexplained adjustments signal a process problem
Lead time Confirmed days from replenishment trigger to delivery Used to calculate the reorder point
Pack size Minimum quantity the supplier ships per unit Replenishment quantity must be a multiple of pack size
Minimum order quantity (MOQ) Smallest order the supplier will accept Replenishment proposals below MOQ are not actionable
Event timestamp Date and time of each stock movement or count Enables latency monitoring and reconciliation; stale events cause incorrect decisions

Designate one system as the record of truth for each field. Where data comes from multiple sources — for example, a warehouse management system for movements and an ERP for lead times — define the merge rule and confirm who is responsible for resolving conflicts.

Replenishment rules and a worked example

VMI replenishment logic should be documented explicitly for each item. The most common approach is min–max replenishment, but the mechanics of how a replenishment is triggered and approved vary. It is important to distinguish three different outcomes that a VMI system can produce:

  • Alert: The system notifies the supplier planner that stock has crossed a threshold. The planner reviews and decides whether to act. No order is created automatically.
  • Supplier proposal: The system or supplier planner generates a replenishment proposal and submits it to the buyer for approval before an order is placed.
  • Automatic order: The system creates and transmits a replenishment order without manual approval, under pre-agreed rules. This requires high data confidence and explicit agreement from both parties.

Which outcome applies depends on the configured rules, the integration in place, and the level of trust between buyer and supplier. VMI does not necessarily mean orders occur without any approval step.

Min–max replenishment: the principle

The reorder point (minimum) is calculated as the expected demand during the replenishment lead time plus a safety stock buffer. When usable on-hand falls to or below the minimum, a replenishment proposal is created. The proposal quantity is calculated to bring the inventory position — on-hand plus open replenishment — toward the maximum, subject to pack size, MOQ, and any delivery capacity constraints.

Worked example with illustrative values

The following values are illustrative. In a real program, thresholds must be derived from your actual consumption data and confirmed lead times.

  • Item: Industrial cutting disc, 125 mm
  • Average daily consumption: 8 units
  • Replenishment lead time: 3 days
  • Safety stock (2 days of demand): 16 units
  • Minimum (reorder point): (8 × 3) + 16 = 40 units
  • Maximum (target stock): 100 units
  • Pack size: 10 units
  • MOQ: 20 units

Scenario: Usable on-hand is 38 units. Open replenishment is 0 units. The inventory position (38 + 0 = 38) is below the minimum of 40, so a replenishment proposal is triggered. The gross quantity needed to reach the maximum is 100 − 38 = 62 units. Rounded up to the nearest pack size of 10, the proposal quantity is 60 units, which is above the MOQ of 20 units, so the proposal is valid. If the buyer has pre-approved automatic ordering for this item, the order is transmitted immediately. If not, the supplier planner submits a proposal for buyer approval.

Note that if reserved or quarantined stock had been included in the on-hand figure, the inventory position would have appeared higher than it actually was, and the replenishment would have been delayed. This is why separating usable on-hand from reserved and quarantined quantities is not optional.

Exception and failure handling

Most VMI problems are predictable. Document how each of the following will be handled before the pilot goes live, and assign an owner to each exception type.

Failure mode Cause Mitigation
Poor master data Incorrect lead times, pack sizes, or item IDs Audit item master before go-live; review after every replenishment discrepancy
Wrong thresholds Min/max set without reference to actual consumption data Derive thresholds from at least 8–12 weeks of consumption history; review quarterly
Unclear stock ownership Operating model does not specify ownership at each stage Document ownership transfer points explicitly; confirm with finance before go-live
Hidden reserved or quarantined stock Usable on-hand includes stock that is not actually available Ensure system separates usable, reserved, and quarantined quantities; verify in pilot
Stale consumption events Stock movements recorded late or in batches rather than in real time Define maximum acceptable data latency; monitor event timestamps; escalate breaches
Pack-size mismatch Proposal quantity not aligned to supplier pack size or MOQ Build pack size and MOQ constraints into the replenishment calculation; validate proposals before transmission
Supplier capacity limits Supplier cannot fulfil a valid replenishment within the agreed lead time Agree a capacity notification process; define the emergency replenishment path in the operating model
No exception owner Discrepancies are identified but no one is responsible for resolving them Assign a named exception owner in the governance table; include resolution SLA in the operating model
Incentives that push excess stock to the buyer site Supplier benefits from shipping more than needed; buyer does not challenge it Include inventory value and turnover in the shared KPI scorecard; review together at each cadence

Define an emergency replenishment path in the operating model. If the normal VMI process fails — due to a data outage, a supplier capacity problem, or an unexpected demand spike — both sides need to know exactly what to do and who to call. An undocumented emergency process means the buyer falls back to manual ordering under pressure, which is exactly the situation VMI was meant to avoid.

KPI scorecard and review cadence

Measure these KPIs from day one of the pilot so you have a baseline to compare against. Review them jointly with your supplier at the cadences shown.

KPI What it measures Review cadence
Stockout rate Frequency of items reaching zero usable on-hand before replenishment arrives Weekly during pilot; monthly thereafter
Fill / service level Proportion of demand met from stock without delay Weekly during pilot; monthly thereafter
Inventory value and turnover Average stock value held and how quickly it moves; signals excess accumulation Monthly
Emergency replenishments Count of out-of-process orders raised to cover VMI failures Weekly during pilot; monthly thereafter
Obsolete and excess stock Value of stock held beyond the maximum level or with no recent consumption Monthly
Forecast / consumption variance Difference between expected and actual consumption; indicates whether thresholds need recalibrating Monthly
Data latency Time between a physical stock event and the event appearing in the shared data feed Weekly during pilot; monthly thereafter
Record accuracy Agreement between system on-hand and physical count at reconciliation Monthly
Disputed transactions Count and value of replenishments or invoices disputed by either party Monthly
Administrative time Hours spent by buyer team on inventory-related tasks; should decrease as VMI matures Quarterly

Hold a structured quarterly review with your supplier covering all KPIs, threshold adjustments, master data changes, and any proposed scope expansions. Treat the data as a shared problem-solving tool, not a scorecard for assigning blame. When a KPI is off target, the first question is whether the operating model, the thresholds, or the data need adjusting.

Technology options and where Aksulit fits

The technology required for VMI does not need to be complex, but it must be accurate, timely, and reliable enough that your supplier can trust the data they see. The minimum viable stack is an inventory management system that records stock movements as they happen, separates usable, reserved, and quarantined quantities, and can share current stock positions with the supplier in a format they can act on.

Manual or batch-updated systems introduce the data latency that causes replenishment errors. Automated tracking — whether through barcode scanning, NFC, or RFID — removes the lag between a physical event and the system record. For a detailed explanation of how RFID-based tracking works in practice, see how RFID inventory management works.

Simple Storage for automated stock visibility is Aksulit’s cabinet and storage unit solution designed for exactly this use case. It tracks RFID-tagged product movements, identifies the user via NFC, and updates stock events and balances in Simple Cloud in real time. When a stock position crosses a configured threshold, the system creates an alert. Depending on the rules configured and the integrations in place, that alert can trigger a supplier notification, a replenishment proposal workflow, or — where the integration supports it — an automatic order. The outcome depends on what has been configured; Simple Storage does not automatically create a purchase order in every setup.

Stock data from Simple Cloud can be shared with your supplier via API, which means the supplier planner or their system can read current positions without manual exports or email attachments. This is the data feed the replenishment logic runs on.

For teams who need to manage inventory away from a fixed storage unit, mobile inventory management with Simple Pocket handles receiving, put-away, picking, transfers, and full inventory counts from a phone or handheld device. Events sync to Simple Cloud automatically, keeping the central stock record current regardless of where the movement happened.

Aksulit has been building inventory management systems since 2003. We are a Finnish software company based in Laukaa, and our customers include industrial maintenance operations, tool and consumable supply businesses, technical wholesale distributors, and on-site vendor-managed locations. We understand the operational realities of keeping stock accurate in busy environments, and we have built our systems around those constraints.

Frequently asked questions

Does VMI mean the supplier places orders without any approval from us?

Not necessarily. VMI can be configured so that the supplier submits a replenishment proposal that requires buyer approval before an order is placed. Fully automatic ordering — where the system transmits an order without a manual approval step — is one option, but it requires high data confidence, explicit agreement from both parties, and a well-tested integration. Most programs start with a proposal-and-approval model and move to automatic ordering once the pilot has demonstrated consistent data accuracy.

Who owns the stock in a VMI arrangement?

Stock ownership is defined in the operating model and is separate from the VMI arrangement itself. In a standard VMI setup, the buyer owns the stock from the moment it is received. In a VMI with consignment, the supplier retains ownership until the stock is consumed or formally transferred. The ownership model affects accounting treatment, insurance, and what happens to excess stock if the arrangement ends. Clarify this in the operating model before go-live. For more on the consignment model specifically, see what consignment stock means in B2B.

How long does it take to set up a VMI pilot?

The timeline depends primarily on the state of your inventory data and the complexity of the data sharing integration. If your item master is clean, your inventory system can export real-time data, and your supplier has the internal capacity to resource the planner role, a pilot can be operational within four to eight weeks. If data quality work or system integration is required first, allow additional time before going live. Rushing the setup to meet an arbitrary start date is one of the most reliable ways to produce a failed pilot.

What is the minimum technology requirement for VMI?

At minimum, you need an inventory system that records stock movements as they happen and can share current stock positions with your supplier. Spreadsheets and manually updated systems are not suitable because they introduce the data latency that causes replenishment errors. Automated tracking — barcode, NFC, or RFID — significantly improves data accuracy and reduces the administrative burden on your team.

How do we handle discrepancies between our stock records and the supplier’s records?

Define the reconciliation process in the operating model before go-live. Specify how often both sides will compare records, who initiates the reconciliation, what tolerance is acceptable, and who is responsible for investigating and resolving gaps. Discrepancies that are not resolved promptly accumulate and eventually cause replenishment errors. Assign a named owner for disputed transactions and include a resolution timeline in the operating model.

Can VMI work for a small or mid-sized business without a dedicated procurement team?

Yes, and it is often particularly well suited to businesses without a large procurement function, because the supplier takes on much of the routine monitoring and ordering work. The key requirement is not headcount but data quality and a named process owner on the buyer side. Even a part-time owner who reviews KPIs monthly and chairs quarterly reviews with the supplier is sufficient, provided the data infrastructure is reliable.

Next step

If you are ready to explore whether VMI is the right fit for your operation, the most useful first conversation is a technical one: what data can your current system share, how accurately does it reflect physical stock, and what integration options are available with your key suppliers. Discuss the technical requirements for a VMI pilot with our team, and we can help you identify what is already in place and what needs to be built before a pilot can go live.

Related Articles