Vendor-managed inventory gets pitched as a service you provide. It is more useful to think of it as a position you occupy. The supplier who manages the shelf sees demand before anybody quotes on it, and stops competing on price for routine replenishment.
It also transfers risk onto you, which is why plenty of distributors run it badly and lose money doing it.
What VMI actually means
You take responsibility for keeping agreed items available at agreed levels at the customer's site. You decide when to replenish, not them. In exchange you get visibility, a contracted position, and usually a term commitment.
The variants differ mostly in who owns the stock and when title passes.
Consignment. You own the stock sitting on their shelf. Title passes on consumption. Most attractive to the customer, most exposed for you, and it lives or dies on the accuracy of the usage count.
Customer-owned VMI. They buy on delivery, you still manage levels and replenishment. Far less capital at risk. Most industrial VMI is really this.
Managed crib. You run a defined storage area on their site, sometimes staffed. Highest service, highest cost, only works at volume.
Where it goes wrong
Consumption is invisible. If you cannot see what was used until you walk the site, you are guessing. Guessing means either stockouts, which destroy the whole proposition, or overstock, which destroys the margin.
Shrinkage. On consignment, anything that leaves the shelf without being recorded is yours. Industrial cribs leak more than anyone admits.
Slow movers. The customer wants everything stocked. Every low-turn item is capital sitting still. The negotiation over what is genuinely stocked versus ordered is the negotiation that determines whether the contract makes money.
Aged stock on age-limited items. This one catches distributors of rubber goods and rated equipment specifically. Consignment stock that sits past its usable life is a write-off you funded.
Why tagging changes the arithmetic
The whole model turns on knowing what is on the shelf and what is installed, without sending somebody to count.
Tag the bin and you get consumption at the moment of use rather than at the next site visit. Tag the item and you also get the other half: what is currently in service, how old it is, and when it comes due. That is the difference between replenishing a shelf and forecasting demand.
For age-limited items it also does something a bin count never can. It tells you which stock is ageing before it expires, so you can rotate it into the next order rather than write it off.
Negotiating it properly
- Define the stocked list narrowly and review it quarterly. The list only grows if nobody looks at it.
- Agree min and max per item in writing, with a named process for changing them.
- Set a shrinkage threshold. Below it, you absorb. Above it, there is a conversation. Without a number this becomes an annual argument.
- Cap ageing exposure on consignment. Either the customer buys stock that passes an agreed age, or you have the right to rotate it out. Do not leave this open.
- Put a term on it. VMI takes a year to pay back the setup. A contract terminable at thirty days is a favour, not a partnership.
- Agree who sees what. You will hold data about their consumption and equipment condition. Write down what you may do with it.
The number to watch
Not revenue. Inventory turns on the stocked list.
Revenue under VMI usually goes up simply because you captured orders that were already happening. That tells you nothing about whether the arrangement works. Turns tell you whether you are funding a warehouse on somebody else's site.
If turns on the managed list are below what you run in your own warehouse, the stocked list is too wide, and the fix is a conversation rather than more stock.