The Kraljic matrix
How buyers classify what they buy: profit impact against supply risk, giving four boxes. Buyers classify categories, not suppliers, and they want the strategic box kept small.
While a supplier is quietly classifying its customers, the customer’s procurement team is doing the same to the supplier, deciding whether it belongs in the strategic box or the irritating-little-cost box. The tool they use is almost certainly Peter Kraljic’s, published in the Harvard Business Review in 1983 as Purchasing Must Become Supply Management and on the desk of every serious buyer since.
What are the four boxes?
The matrix sorts every purchase on two axes: how much it affects the buyer’s bottom line, and how risky it is to secure. Non-critical items, low on both, are the paperclips of the business, bought on ruthless efficiency and increasingly by machine. Leverage items matter to the bottom line but come from plentiful suppliers, so the strategy is to make suppliers compete, squeeze the price and remind each of them they are replaceable. Bottleneck items cost little but could halt operations, so the buyer manages them nervously, hedging and hunting for alternatives. The strategic box, high impact and genuinely hard to source, is where the buyer reaches for partnership and the long-term relationship. It is small on purpose.
Why does the buyer want the strategic box small?
Strategic relationships cost the buyer investment, compromise and dependence, and they reduce the thing procurement prizes most: optionality. The book’s image is an ocean of interchangeable suppliers around a tiny island of partners, with the buyer measured on growing the ocean and shrinking the island. The supplier’s growth depends on the opposite.
Which box is a supplier in?
The question is wrong. A buyer classifies categories, one line at a time, not whole suppliers. As an illustration, an industrial supplier’s spare parts might sit in leverage, bought by tender against four rivals, while its condition-monitoring service sits in bottleneck because nobody else can read the customer’s machines. Same supplier, two boxes, two sets of people. The figure lets either offer be moved.
Four boxes, and the two ways out.
The buyer's play: make suppliers compete, squeeze the price, and make sure each one knows it is replaceable. Operational buyers run the tender to the policy they were given.
The way out: raise supply risk honestly, by becoming hard to replace: proprietary methods, deep integration, insight the customer cannot get elsewhere. One of the two boxes where AI can already run the sourcing with no human in the loop.
Who in procurement matters?
Procurement is not one job. A few strategic procurement leaders decide which categories are strategic in the first place. Below them, many more operational buyers execute that policy: they run the tender, apply the guidelines and squeeze the agreed suppliers. The operational buyer is not being difficult; they are doing what the matrix told them to do. That is why reaching the strategic level, where the box can actually change, matters so much in mapping the decision-making unit.
How does an offer move?
The boxes are perceptions on two axes, and perceptions move. Profit impact rises when the offer is connected to something the customer’s strategy depends on, which is the big picture rather than the specification. Supply risk rises, honestly, when the supplier becomes genuinely hard to replace. It takes years. Mark Davies sets out the climb in Are you a strategic supplier?
There is a sharper edge now. The non-critical and leverage boxes are where buying is being automated, first by catalogues and e-tendering, now by AI sourcing that drives the price down with no human in the loop. In the book’s words: “the buyer of the future is a machine, and it will not take your calls.” The strategic box is the human one.
Set out in full in: Creating Customer Value Propositions with AI (Kogan Page, 2027), chapter 5.
Related terms
Decision-making unit (DMU)
Everyone on the customer side who decides, influences, uses, pays for or can veto a purchase, mapped as seven roles and their relationships rather than job titles.
Big picture and small picture
The small picture is what the customer asks for, in the tender. The big picture is what its strategy needs. The gap between the two is where a supplier adds value.
Machine buying
B2B purchasing done by software on the buyer's side. Around a third of B2B commerce by 2030, and the transactional layer of selling with it.
The commodity trap
The gravity in business markets that pulls every offer towards sameness until price is the only thing left to compare, made stronger now that AI has stripped the scarcity out of capability.
Customer needs analysis
The diagnosis. Contextual intelligence about the customer's world, analytical depth about the problems underneath the problems, and clarity about who feels them.
See the method run on one real account in three days: the One Page Proposition.