How much cash is tied up in your inventory

When demand weakens and supply chains become uncertain, the management of a trading company faces competing priorities. It needs to release capital from excess inventory while keeping saleable goods available. Decisions about which stocks can be reduced and where reserves are necessary must be made at item level; the board and management need to understand how those decisions affect liquidity and profit.

The assignment concerns a trading company headquartered in Switzerland, with operations in Germany, France, Austria and Italy and three warehouse locations. Different ERP systems make it difficult to obtain a consolidated view of inventory, sales and margins. The interim CFO is tasked with working alongside purchasing, sales and logistics to establish a reliable basis for decisions and connect inventory reduction with financial management.

The risk analysis highlights geopolitical uncertainty in supply chains and weak demand in the company's sales markets. Procurement risks and sales trends are therefore assessed together. This results in two priorities: inventory management that reduces at-risk stock and stock with little remaining economic value, and reliable cash flow planning.

Understanding where capital is tied up

The team begins by consolidating item, inventory, purchasing, sales and returns data and reconciling them with the financial accounts. While AI can flag duplicates and incorrect item mappings, the operational teams check variants and units of measure; the interim CFO is responsible for a financially consistent foundation. Warehouse location, selling entity and sales country remain separate dimensions so that inventory and earnings are allocated correctly. Reserved, blocked and in-transit goods are shown separately, while intercompany effects and currencies are adjusted for the group view.

This provides a basis for examining how long an item has been held, how quickly it sells and what quantities remain on order. Inventory value, stock coverage and returns are considered alongside lead times, expiry dates and product life cycles. The original inventory age is retained after a warehouse transfer, so that older goods do not appear unproblematic simply because they have changed location.

As age alone is insufficient for replenishment decisions, an ABC analysis based on annual consumption value shows an item's financial significance, while an XYZ classification shows how regularly it is demanded. Combined with margin and supply importance, this creates a differentiated basis for order quantities and safety stocks. A rarely needed spare part requires a different assessment from a seasonal item approaching the end of its life. Comparing all three warehouses also reveals where excess stock coincides with shortages and a transfer would be worthwhile.

This analysis also identifies the causes of capital being tied up. Minimum order quantities and volume discounts can contribute to excess stock, as can unsuitable replenishment parameters or delays in releasing returns. Understanding these relationships is a prerequisite for reducing existing stock and limiting new surpluses.

Turning inventory risks into commercial decisions

Purchasing, sales and finance therefore agree transparent criteria for at-risk goods and goods with little remaining economic value. At-risk inventory includes stock above expected demand plus a justified safety stock, as well as items whose sales prospects are threatened by ageing, the end of a season or successor products. Goods with no economically viable sales prospect within the defined disposal period fall into the internal category of ‘toxic’ stock. The two risk classes are reported separately; they do not replace an accounting valuation.

AI supports the assessment by estimating demand and expected residual stock from sales history, prices, promotions, returns and life cycles. It is essential to distinguish a lack of demand from a lack of availability: an item that was out of stock could not be sold. For new or rarely sold items, comparable products and professional judgement supplement the data. Each classification must disclose its assumptions and the uncertainty of the forecast.

The interim CFO derives priorities from the capital tied up and expected net proceeds after discounts and disposal costs, enabling a commercial comparison of purchasing freezes, warehouse transfers, returns to suppliers, targeted clearance sales and delisting with the operational teams. Selling below cost can make economic sense if continued storage is expected to cause greater losses. Purchasing and sales confirm the specific action; any required write-down is assessed separately.

Procurement is adjusted in parallel. Order quantities and safety stocks reflect demand, lead-time risks and the agreed service level. Volume discounts are weighed against additional financing, storage and obsolescence costs. Smaller call-off quantities and more reliable delivery intervals are appropriate where they reduce total costs and secure supply.

Connecting cash release with profitability

Management needs to know when an inventory action actually creates financial headroom. A write-down does not release cash. A warehouse transfer benefits the company only when it enables additional sales or actually avoids a purchase. Expected cash benefits are therefore reconciled with the cash flow forecast and subsequently with actual cash movements.

The analysis extends beyond the warehouse. Receivables and marketplace payout periods determine when sales proceeds become available; supplier payment terms and early-payment discounts affect outflows. Together, these factors capture the time between paying suppliers and receiving customer payments. A rolling 13-week cash flow forecast, updated weekly, shows the effects of weaker demand, longer lead times or higher procurement costs. Thresholds are agreed with management at which orders, safety stocks or investments must be adjusted.

Inventory reduction must also take profitability into account. Deducting cost of goods sold from net sales after discounts and returns gives gross profit. After deducting the defined variable costs of platforms, payment processing, shipping, returns and marketing, the contribution margin remains. Analysis by item, country and sales channel thus shows which sales contribute to covering fixed costs. This makes it possible to assess where a promotion is commercially viable and where expected cash inflows come at the expense of profit.

Applying AI selectively in planning and warehouse operations

The value of an AI application must be measured against these commercial objectives. Before a model supports replenishment, its forecast is compared with a simple seasonal forecast using historical periods excluded from model development. Alongside forecast error, the assessment considers systematic over- or underestimation and its consequences for inventory and availability. Only after approval by the responsible specialists does the model support proposals for order timing, quantities and transfers, taking account of existing stock, open orders and lead times. Exceptions are reviewed by the operational owners; any subsequent automated execution requires defined limits and approval rules.

Further applications are possible in the warehouse. Demand frequency and items commonly ordered together can support storage assignments that shorten picking routes, provided weight, dimensions and storage conditions permit. Unusual transactions or repeated inventory adjustments prompt targeted recounts. For returns, prioritisation by value, sales prospects and anomalies helps direct condition checks and subsequent processing. Resaleable goods should be released promptly; clear responsibilities, scan confirmations and timely postings keep the process reliable.

Power BI brings the information together for management. Operational reservations and warehouse orders remain in the ERP or warehouse management system. Inventory accuracy, picking errors and returns processing time are among the measures used to assess whether applications deliver the expected benefit. For each application, the CFO compares the expected benefit with the costs of implementation, interfaces, licences, operations and internal working time. One-off cash release, recurring savings and additional contribution margin are assessed separately.

Embedding implementation in management routines

In a pilot using business cases from all three warehouses, the team first checks data quality, forecasts and proposed actions. Rollout follows approval by the responsible specialists. The pilot already establishes how improvement will be recognised. AI models are monitored during operation; confidential data remain in approved environments.

The management dashboard considers capital tied up, profitability and product availability together. Every metric receives a binding calculation method, data source and responsible owner. Targets depend on product group, lead time and service promise. Consistent periods, currencies and tax bases ensure comparability; missing consumption data and actual demand of zero require separate treatment. The metric definitions in the appendix provide the technical foundation.

In a weekly decision meeting, each inventory action is assigned an owner, a deadline and an expected effect on profit and cash. The interim CFO brings together the interests of purchasing, sales, logistics and finance, prepares the necessary decisions and tracks implementation. This gives management a basis for resolving competing priorities deliberately and intervening when performance deviates from expectations.

At the end of the assignment, the internal team takes ownership of the metric definitions, data checks and decision processes after employees have applied the methods to their own business cases. The success of the assignment is measured by verified cash release, the economic performance of inventory and compliance with the agreed service level.

Appendix: financial and operational management

The following definitions provide the foundation for financial and operational management in the assignment. Risk classes and targets are set for the company concerned; they are not universal industry standards.