Published 17 September 2026 · 6 min read
The Kraljic matrix is the most widely taught tool in procurement and one of the most poorly applied. Peter Kraljic published it in Harvard Business Review in 1983 and the framework has outlived almost everything published alongside it. Here is what it is, how to build one from your own data, and where it misleads.
The four quadrants
Two axes: profit impact (how much you spend, how much it affects cost or quality) and supply risk (how hard it would be to replace the supplier).
- Leverage — high spend, low risk. Many capable suppliers, lots of money. Competitive tendering works here, and this is where most quick savings live.
- Strategic — high spend, high risk. Few suppliers, critical to operations. Manage the relationship, not the price. Squeezing here is how you end up with a supply failure.
- Bottleneck — low spend, high risk. Cheap but hard to replace. The goal is to reduce the risk: qualify a second source, hold stock, redesign the specification.
- Routine — low spend, low risk. The long tail. Minimise the cost of buying rather than the price of the item — frameworks, P-cards, catalogues.
How to build one from your spend data
- Classify your spend first. The matrix plots categories, not suppliers or transactions. Without a taxonomy you have nothing to place on it.
- Profit impact comes straight from the data. Total annual spend per Level 2 category. Split high/low at the median, or wherever the natural break is.
- Supply risk is a judgement. Nothing in your ERP tells you this. Score each category on number of capable suppliers, switching cost, lead time, and whether a failure stops production.
- Plot and argue. Do it as a workshop. The disagreements are the value — they surface assumptions nobody had written down.
Where it misleads
- It is a snapshot. A leverage category with one supplier going bust becomes strategic overnight. Revisit annually.
- Supply risk gets scored by the person who owns the category, who has an incentive to call their spend strategic. Score it in a group.
- It implies four strategies for hundreds of categories. Use it to prioritise, not to prescribe.
- The axes hide the tail. Routine is one box on a chart and often 60% of your supplier count.
Why most attempts stall
Not the framework — the data. Teams block out a workshop, then discover they cannot answer “what do we spend on this category annually” without two weeks of manual work, and the exercise quietly dies.
Classified spend turns the matrix from a project into an afternoon. One axis is then a database query and the other is the conversation you actually wanted to have.
Further reading
- Building a spend cube — where the profit-impact axis comes from.
- Supplier rationalisation — what to do with the leverage quadrant.
- Tail spend management — what to do with routine.