The question before the conclusion.
- Lens
- inventory
- Journey
- 8 decision points
- Reading
- 6 min
A salesperson calls the warehouse: ‘how many units of item 4412 do I have left?’
Answer: sixty.
They sell sixty. Two days later, the order ships incomplete. There were indeed sixty units physically present, but twelve were reserved for another customer and fifteen were part of a backorder already promised.
No one lied. Three people used the word ‘inventory’ for three different things.
The three inventories
Physical inventory. What is in the warehouse, measurable and countable. This is the quantity the warehouse knows.
Available-to-sell inventory. Physical inventory less everything already committed: reservations, orders being picked, and quantities blocked for a dispute or inspection. This is what sales needs, and the figure least often calculated correctly.
Projected inventory. Available inventory plus confirmed incoming purchase orders, with their dates. This makes it possible to say: ‘I do not have it today, but I can commit it to you for Tuesday’.
A distribution business that does not distinguish these three concepts lives with a permanent disadvantage: it promises what it does not have or rejects a sale it could have made.
The real cost of ambiguity
The lost revenue from a stockout is visible: a lost sale and an unhappy customer. It is discussed in meetings.
What no one discusses is the sale incorrectly refused. A salesperson who does not trust the displayed inventory adds a mental safety margin. They say no, or announce a longer lead time than necessary, and the customer goes elsewhere. That loss leaves no record: it appears in no indicator, no one reports it, and it happens every day.
That is why the reliability of displayed inventory has greater commercial value than it is given credit for. A salesperson who trusts the screen sells more—not because there is more inventory, but because they stop censoring themselves.
What must be decided—and is not an IT matter
What is deducted from available inventory? Does entering an order reserve stock, or must the order first be approved? Does a quotation in progress reserve stock? For how long?
Who may reserve, and for how long? Without an expiry rule, reservations accumulate: after a year, a substantial share of inventory is reserved for dead opportunities. With some humour, this is called phantom inventory—it is on the shelf, but no one can sell it.
What happens with an order that can be served only in part? Ship what is available, wait until it is complete, or ask the customer? The answer depends on the customer and product—but it must be written somewhere, otherwise every salesperson decides differently and the warehouse bears the consequences.
What is included in projected inventory? A confirmed purchase order, yes. An order placed but not confirmed? Goods in transit? Goods undergoing customs clearance? Each has a different probability and lead time.
These four decisions can be made in one meeting between sales, procurement and the warehouse. They determine the value of everything displayed afterwards.
The particular case of distribution: several models under one roof
What makes B2B distribution distinctive is that the same company often runs three models in parallel.
For part of its catalogue, it buys into stock and sells from that stock. For another part, it sells back-to-back: the customer order triggers a supplier order. For a third, it acts as an intermediary: it never takes possession of the goods, but arranges and invoices a service.
These three models do not have the same availability rules, lead-time commitments or margin structure. Applying the same rules to all of them produces false promises in at least two cases out of three.
The question to ask, item by item or family by family, is: which model does it follow? Many companies discover through this exercise that they do not know—and that sales and procurement do not give the same answer.
Credit exposure: the control that must come before, not after
A related issue, resolved at the same point.
In many organisations, customer credit exposure is checked at invoicing, or worse, during debt collection. By then the goods have left. The debate is no longer ‘should we deliver?’ but ‘how do we recover the money?’.
The useful control happens when the order is taken: current exposure, payment terms and any overdue amounts. It is not a hardening of policy—it allows the salesperson to negotiate in full knowledge of the facts, before making a commitment.
One observation from experience: companies introducing this control almost always fear that they will block sales. What generally happens is something else—they discover customers they had served without margin for years.
What a system genuinely contributes here
Nothing spectacular, and a great deal of daily value:
show everyone the same figure, with its definition;
distinguish physical, available and projected inventory without manual calculation;
apply reservations and expiry according to the rules decided;
link a customer order to its purchasing requirement when the model calls for it;
check credit exposure automatically at the right time;
and show margin by order at decision time, not three months later.
That last point changes behaviour the most. A salesperson who sees the margin while building an offer makes different trade-offs. Not because they are being watched, but because they know.
