Kraljic Matrix: The Purchasing Portfolio Model Explained

By Brian Tr · Updated Oct 10, 2026

The Kraljic matrix is a 2x2 purchasing portfolio model that sorts everything a company buys by two measures: profit impact and supply risk. The result is four groups, strategic, leverage, bottleneck, and non-critical items. Each group calls for a different sourcing approach, so buyers stop treating paper clips and rare components the same way.

This guide covers where the model came from, how to score your own spend, what to do in each quadrant, a template you can copy into a spreadsheet, and where the matrix falls short.

Key Facts

• What it is: a purchasing portfolio model that classifies purchased items into four categories on two dimensions, profit impact and supply risk.
• Who created it: Peter Kraljic, then a director in McKinsey & Company's Düsseldorf office, in "Purchasing Must Become Supply Management", Harvard Business Review, 1983.
• The four groups: strategic (high impact, high risk), leverage (high impact, low risk), bottleneck (low impact, high risk), non-critical (low impact, low risk).
• Core idea: match your sourcing strategy to the item. Use buying power where the market is deep, build partnerships where supply is thin, and cut effort where neither matters.
• Main weakness: the scores are judgment calls, and the matrix is a snapshot that goes stale as markets move.

Where the Kraljic matrix came from

Peter Kraljic published the model in 1983 in Harvard Business Review. The HBR page for the article lists him as a director in the Düsseldorf office of McKinsey & Company, and the title makes the argument: purchasing had to grow from a clerical function into supply management, a strategic job.

The matrix he built is now treated as a standard diagnostic in the field. A peer-reviewed study of purchasing portfolio models describes it as the most recognized and commonly used purchasing portfolio model, and notes that it classifies items into non-critical, bottleneck, leverage, and strategic categories along two dimensions: profit impact and supply risk.

Kraljic also said how to measure each axis. According to that same study, he suggested judging strategic impact by the volume purchased or total cost, and judging supply risk by the complexity of the supply market: shortages, technological change, substitutes, entry barriers, logistics costs, and monopoly or oligopoly conditions.

The two axes

Profit impact

Profit impact asks how much this item matters to your results. Think about how big the spend is, what share of your product's cost it represents, and how much it affects quality or the customer's experience. A component that makes up a third of your cost of goods has high profit impact. A box of staples doesn't.

Supply risk

Supply risk asks how hard it would be to get the item, or get it again if something goes wrong. The factors Kraljic listed point at the structure of the supplier market: how many suppliers exist, how scarce the material is, whether substitutes work, how tough it is for new suppliers to enter, and whether one or two suppliers dominate. Lead times, geography, and how specialized the item is belong here too.

Put them together and you get four boxes.

The four quadrants

Strategic items

These are the purchases your business can't do without and can't easily replace. Maybe you rely on a custom component from one qualified supplier, or on a cloud platform your product runs on. The usual approach is a long-term relationship: shared forecasts, joint planning, and a backup plan for the day the supplier fails. This is where a formal supplier relationship management program earns its keep. Some companies also consider making the item in-house, or qualifying a second source over time.

Leverage items

Leverage items carry real spend, but plenty of suppliers sell something close enough. Here you hold the power. Consolidate volume, run competitive bids, negotiate framework agreements, and switch suppliers when price or service slips. Standard raw materials, packaging, and commodity services often land here.

Bottleneck items

A bottleneck item doesn't cost much, but a shortage can stop production or a project. The goal isn't a lower price, it's availability. Typical moves include holding safety stock, qualifying an alternative item or supplier, signing longer supply agreements, and redesigning the product to use something easier to get. The aim over time is to push these items toward the non-critical box.

Non-critical items

Office supplies, standard fasteners, and routine maintenance goods sit here. Spend is small and suppliers are easy to swap, so the best strategy is to spend as little buyer time as possible. Standardize the catalogue, use purchasing cards or e-procurement, and automate reorders.

Kraljic matrix vs. other portfolio tools

People often confuse the Kraljic matrix with other 2x2 or ranking tools. They answer different questions.

A few useful contrasts:

• Kraljic vs. BCG. Both plot a portfolio on two axes and prescribe a different play for each box. But BCG looks outward at what you sell, and Kraljic looks inward at what you buy. If you've used the BCG matrix to decide where to invest in your product lines, think of Kraljic as the same habit applied to your supply base.
• Kraljic vs. ABC analysis. ABC ranks items by one number, usually how much money flows through them. That's useful, but it can't see risk. A cheap part with one supplier sits in the C pile and still shuts down your line. Kraljic adds the risk axis that ABC lacks, and ABC is a fast way to get the spend data you need for the profit-impact axis.
• Kraljic vs. supplier segmentation. The matrix sorts items; segmentation sorts the companies you buy them from. Most teams use both: run Kraljic first, then decide which suppliers in each quadrant deserve a deeper relationship. A supplier that sells you three strategic items needs different attention than one that sells you pens.

Pair it with Porter's Five Forces too. Supplier power is one of the five forces, and the supply-risk axis is a close cousin: it asks how much power the supplier side holds over you.

How to build a Kraljic matrix in five steps

Step 1: List what you buy

Pull a year of spend from accounting or your purchasing system. Group items into categories rather than individual part numbers. A category might be "corrugated packaging" or "payroll software," not 40 line items. Aim for 20 to 100 categories for a mid-sized company.

Step 2: Score profit impact

For each category, rate it on factors such as annual spend, share of product cost, and effect on quality or customer experience. Use a simple 1 to 5 scale. Weight the factors if one clearly matters more, and write the weights down so the next person can reproduce your scores.

Step 3: Score supply risk

Rate each category on the supplier-market factors: number of qualified suppliers, availability of substitutes, how hard it is to switch (see switching costs), lead times, and supplier concentration. Again, 1 to 5.

Step 4: Plot and assign quadrants

Choose a cutoff, often 3, and place each category in a box. Then check the borderline cases by hand. If a category scores 2.9 on risk but your only supplier just announced a plant closure, it's a high-risk item whatever the arithmetic says.

Step 5: Set a strategy and review it

Write one action per quadrant, name an owner for every strategic and bottleneck category, and set a review date. Markets change, so revisit the scores at least once a year and after any major disruption.

A worked example (hypothetical)

Suppose a mid-sized company that makes commercial coffee machines scores five purchase categories. The numbers are invented to show the mechanics.

The interesting row is the pressure sensor. It's a small cost, so it never appears in a "top spend" report. But if the one supplier misses a delivery, the assembly line stops. That's the kind of item a spend-only view hides, and the matrix surfaces it.

Kraljic matrix template

Copy this table into a spreadsheet. Fill in the scores, then let a formula assign the quadrant (for example, high impact is a score of 3 or more).

If you need to compare options inside one quadrant, a decision matrix works well for scoring candidate suppliers against weighted criteria.

Examples by business function

• Manufacturing: Raw metals and plastics are often leverage items. A sole-source specialty component is strategic. A rarely-ordered spare part with one supplier is a bottleneck.
• IT: Your core cloud platform or ERP system is strategic, since switching is slow and costly. Commodity hardware like laptops is leverage. Single-use software licenses for small teams are non-critical.
• Marketing: An agency that holds your brand knowledge may be strategic. Print and swag are leverage or non-critical. A niche data provider with no equivalent could be a bottleneck.
• HR: Payroll and HRIS vendors tend toward strategic. Recruiting agencies are often leverage because many compete. Background check vendors can be bottleneck items in tight hiring windows.
• Facilities: Utilities in a regulated market are bottleneck or strategic. Cleaning and office supplies are mostly non-critical.

If you sell into large buyers, it helps to know how their procurement teams think. Many sort vendors with logic like this, which is why our guide on working with corporate procurement is worth a read. A vendor that looks like a commodity gets squeezed, so the case for being seen as strategic matters.

Strengths and limitations

The matrix is popular because it's simple, it uses data you probably already have, and it forces a conversation about risk that spend reports skip. It also gives a shared vocabulary: "that's a bottleneck item" means the same thing to finance and operations.

But it has real limits. The Chartered Institute of Procurement & Supply lists several: the model ignores the supplier's point of view, gives no guidance on moving items between quadrants, and can make allocating items to a quadrant complex and difficult. A research paper on the model also points out that the factors behind the ratings aren't equally important, so the weighting matters (University of Maribor Press).

Other cautions:

• Scores are subjective. Two buyers can rate the same category differently. Agree on definitions and weights before scoring.
• It's a snapshot. A supplier merger or a new tariff can move a whole category in a month.
• It can hide total cost. A cheap supplier in a leverage category may cost more in quality problems than a pricier one.
• Categories interact. Squeezing one supplier on leverage items can damage the relationship you need for a strategic one.

Do's and don'ts

Do:

• Score at the category level, not part by part.
• Write down your scoring rules and weights.
• Test borderline items against reality, not just the formula.
• Assign a named owner to every strategic and bottleneck category.
• Aim to move bottleneck items toward non-critical through substitution or redesign.
• Review at least annually and after any supply shock.

Don't:

• Use the same tactic everywhere. Squeezing a sole-source supplier invites trouble.
• Treat the cutoff as exact. A 2.9 and a 3.1 aren't different worlds.
• Spend negotiation time on non-critical items. Automate them.
• Forget the supplier's side. A supplier who loses money on you won't prioritize you in a shortage.
• Confuse it with the value chain. The value chain maps your activities, while Kraljic maps your purchases. They work well together, since inbound logistics and procurement are part of the chain.