Why article-level tenders beat basket discounts
A generous headline discount can hide an expensive agreement. The precision of the specification decides the quality of the price.
Most framework agreements for indirect categories are negotiated the same way: the supplier presents a price list, the buyer negotiates a discount on it, and the agreement is signed at "list minus X percent." It feels like a result. X is a number, the meeting ends well, and procurement reports a saving.
The supplier, meanwhile, has kept control of the only thing that matters: the list. A discount on a price you do not control is not a price — it is an invitation for the supplier to decide your cost structure. List prices can be raised. Products can be substituted for near-identical articles at different list positions. And most importantly, the discount can be placed where you are not looking.
The arithmetic of the basket
Consider an illustrative example. A supplier offers 15 percent off the full catalogue. Within your actual purchasing pattern, however, a handful of high-volume articles dominate the spend — as they nearly always do. If the list prices of exactly those articles sit 30 percent above the market, the agreement is a loss dressed as a saving. The 15 percent headline is real on paper and irrelevant in practice, because the average is computed over a catalogue you do not buy evenly.
This is not sharp practice at the margins; it is the standard economics of catalogue pricing. Suppliers know their own price-volume distribution with precision. Buyers who negotiate at basket level are negotiating blind against someone who can see.
Precision reverses the asymmetry
The alternative is to specify the tender at the level where money actually changes hands: the article. A tender at category scope, specified article by article, tells suppliers exactly what the bloc buys — which articles, in which volumes, to which requirements. Suppliers no longer price a catalogue; they price the demand itself, article for article, in competition with each other.
Three things change immediately. Bids become comparable — two offers on the same specified articles can be evaluated line by line, with no averaging tricks available. Bids become honest — there is no tail of unexamined articles in which to hide margin. And negotiations become factual — the discussion is about the price of a specific article in a specific volume, not about the theology of a discount percentage.
Why almost nobody does it
If article-level tendering is so clearly better, why is it rare? Because it is expensive to do alone. It requires knowing your own consumption at article level — which means classifying thousands of order lines and invoices, matching equivalent products across suppliers, and keeping that picture current. For a single company, the analytical cost often exceeds the negotiating gain on any one category.
This is exactly the cost a bloc removes. In Aitenda, the analysis is done once, on the combined data of all members — orders, invoices and enriched product data from suppliers — with AI-assisted classification and product matching doing the volume work that no manual team could. The result is a specification each member could not have afforded to produce alone, applied to a volume no member could have offered alone.
Article-level precision is not a technicality of tender design. It is the difference between borrowing the supplier's version of your costs and owning the facts yourself.