TL;DR: The Kraljic matrix is a 2x2 grid that sorts what you buy by profit impact and supply risk into four quadrants: non-critical, leverage, bottleneck and strategic. Each quadrant gets a different buying approach, so you stop negotiating hard over paper towels and stop ignoring the one supplier who can stop your line. Score categories, not suppliers, and redo the scores at least once a year.
The Kraljic matrix is a purchasing portfolio model that classifies purchase categories on two axes, profit impact and supply risk, and assigns a buying strategy to each of the four resulting quadrants. Peter Kraljic, then a director at McKinsey & Company in Dusseldorf, introduced it in "Purchasing Must Become Supply Management" in the September 1983 issue of Harvard Business Review. Low impact and low risk items are non-critical: make buying them cheap to process. High impact, low risk items are leverage items: use competition to cut price. Low impact, high risk items are bottleneck items: secure supply first. High impact, high risk items are strategic: build long-term relationships and contingency plans. Below is a scoring rubric you can use in a spreadsheet and a worked example with eight categories from a 40-person machine shop.
What are the four quadrants of the Kraljic matrix?
The four quadrants of the Kraljic matrix are non-critical, leverage, bottleneck and strategic items. Each one tells you where to spend buyer time and what a good outcome looks like.
| Quadrant | Profit impact | Supply risk | Goal | Typical actions | Typical SMB examples |
|---|---|---|---|---|---|
| Non-critical | Low | Low | Cheap, fast buying | Catalogs, purchasing cards, delegated approval, fewer suppliers | Office supplies, cleaning, standard fasteners |
| Leverage | High | Low | Lowest total cost | Competitive RFQs, volume consolidation, annual price reviews | Commodity raw materials, packaging, freight |
| Bottleneck | Low | High | Continuity of supply | Safety stock, qualify a second source, longer agreements, redesign out | OEM spare parts, certified test labs, niche components |
| Strategic | High | High | Long-term value and resilience | Joint planning, multi-year contracts, risk reviews, make-or-buy checks | Critical inputs, specialist subcontractors |
Non-critical items are low impact and plentiful. The goal is to reduce the administrative burden, for example by delegating purchasing or using catalogs (Wikipedia summary of the model). For many of these purchases the cost of processing the transaction exceeds the value of what is bought, so the win is fewer steps, not a lower unit price.
Leverage items matter to your margin but have many qualified suppliers. This is where buying power works: bundle volume, run a competitive quote, switch when the numbers say so.
Bottleneck items are small in spend but hard to replace. A $400 part with a single source can idle a $500,000 machine. Pay for security here: buffer stock, a second qualified supplier, or a design change that removes the dependency.
Strategic items are both expensive and hard to replace. They need senior attention, longer agreements and a written contingency plan. Kraljic's article also asks you to weigh your buying power against the supplier's strength for these items and pick one of three stances: exploit (use strong buying power to lock in price and terms), balance (a middle path), or diversify (reduce risk by finding alternative suppliers or substitutes), as summarised by Mind Tools.
How do you score profit impact and supply risk?
Score each purchase category from 1 to 5 on three criteria per axis, average them, and treat 3.0 or above as "high". The rubric below is ours, built from the two dimensions Kraljic defined; the point is consistency, so keep the same rubric every year.
Profit impact asks how much the category moves your profit and product:
| Criterion | 1 | 3 | 5 |
|---|---|---|---|
| Share of total third-party spend | under 2% | 5-15% | over 30% |
| Effect on product quality or customer | none | noticeable | defines the product or a customer requirement |
| Margin effect of a 10% price change | negligible | visible in quarterly margin | material to annual profit |
Supply risk asks how hard the supply market is:
| Criterion | 1 | 3 | 5 |
|---|---|---|---|
| Qualified suppliers you could buy from | 6 or more | 2-3 | 1 |
| Time and cost to switch | days, no requalification | weeks, a trial run | months, customer or certification approval |
| Disruption exposure | stock items, short lead time | some lead time or regional concentration | long lead time, single region, scarce input |
The model defines supply risk by the complexity of the supply market: monopoly situations, barriers to entry and technological innovation. Mind Tools gives the same idea in plain terms: risk is high when an item is scarce or its availability could be hit by government instability or natural disasters, and profit impact is high when the item adds significant value to your output.
Two scoring rules prevent most bad results:
- Score categories, not suppliers. In its original form the matrix maps purchases, not suppliers. One distributor can sell you a leverage item and a bottleneck item, and a single supplier score blends them into a misleading average.
- Do not use spend as the whole vertical axis. Spend is one of three profit impact criteria above, on purpose. Using spend alone pushes high-volume, low-margin categories to the top and hides cheap items that are critical to quality.
A worked example: eight categories at a small machine shop
Here is the matrix applied to a hypothetical 40-person CNC machine shop with $808,000 a year of third-party spend across eight categories. The figures are illustrative; use your own spend export and the rubric above.
| Category | Annual spend | Share of spend | Profit impact | Supply risk | Quadrant |
|---|---|---|---|---|---|
| 6061 aluminum bar | $410,000 | 50.7% | 4.3 | 2.0 | Leverage |
| Outsourced anodizing | $120,000 | 14.9% | 3.7 | 4.0 | Strategic |
| Certified titanium bar | $95,000 | 11.8% | 3.7 | 3.7 | Strategic |
| Cutting tools and inserts | $85,000 | 10.5% | 3.3 | 1.7 | Leverage |
| LTL freight | $40,000 | 5.0% | 2.3 | 1.7 | Non-critical |
| OEM spindle repair | $38,000 | 4.7% | 2.0 | 4.7 | Bottleneck |
| Shop and office supplies | $11,000 | 1.4% | 1.0 | 1.0 | Non-critical |
| Calibration lab services | $9,000 | 1.1% | 1.7 | 3.7 | Bottleneck |
How one row was scored, outsourced anodizing:
- Profit impact: share of spend 14.9% scores 3; quality effect scores 5 because customer drawings specify the coating and a bad batch scraps finished parts; margin effect scores 3. Average (3 + 5 + 3) / 3 = 3.7.
- Supply risk: two qualified finishers within trucking distance scores 4; switching needs first-article approval from customers, scoring 4; lead time and single-region concentration score 4. Average 4.0.
- Result: strategic, even though anodizing is only 15% of spend.
What the shop does with the result:
- Leverage (aluminum, cutting tools, 61% of spend): run a competitive RFQ for aluminum bar every 12 months with at least three mills or service centres, and consolidate cutting tools onto one or two distributors for volume pricing. This is where most of the savings sit.
- Strategic (anodizing, titanium): quarterly reviews with both finishers, a 12-24 month agreement with agreed capacity, and qualify the second anodizer on one part number now so a switch takes weeks, not months.
- Bottleneck (spindle repair, calibration): keep one spare spindle cartridge on the shelf, sign a service agreement with response times, and book calibration dates a quarter ahead. Accept paying more here.
- Non-critical (freight, supplies): a single office supplies account with a monthly card statement, and a freight broker or platform so nobody negotiates individual shipments.
Notice what the matrix changed: without it, the shop would spend its negotiating energy on the biggest line (aluminum) and its worry on nothing. With it, a $9,000 calibration contract gets a plan because losing it would stop shipments that need certificates.
How does a small business use the Kraljic matrix with little spend?
A small business uses the Kraljic matrix by scoring 8 to 20 categories, not hundreds of suppliers, and by assuming its buying power is weaker than the textbook suggests. Three adjustments make it fit a small company:
- Use fewer categories. Group your spend export into 8-20 categories that a person would recognise. More detail adds work without changing decisions.
- Push your own risk scores up a notch when you are a small customer of a big supplier. Project management educator Mike Clayton makes this point in the video below: if you are a small account to a large supplier in your strategic quadrant, supply risk may be higher than the matrix shows, because you are the customer they drop first.
- Prefer "diversify" over "exploit" for strategic items. With $100,000 of annual spend you rarely have the leverage to dictate terms. Qualifying a second source, holding stock and keeping specifications standard are usually cheaper than trying to win a power struggle.
For leverage items, small companies get the most from the simplest tool: a clean, comparable quote from several suppliers at least once a year. If you run those quotes over email and spreadsheets, AuraVMS is RFQ and purchase-order software for small and mid-sized businesses: suppliers respond without creating an account, quotes line up side by side with automatic L1/L2/L3 price ranking, and anonymous bidding keeps suppliers from anchoring on each other. AuraVMS starts at $4.99/month, billed annually.
Mike Clayton walks through the four Kraljic quadrants with the strategy each one needs, then gives practical tips: start early, treat the matrix as changing over time, and remember that being a small customer of a large supplier raises your real supply risk.
How do you build a Kraljic matrix step by step?
You build a Kraljic matrix in four phases, the same structure Kraljic proposed: classify purchases, analyse the supply market, position each item, and plan actions.
- Classify. Export 12 months of spend from accounting, group it into categories and score profit impact.
- Analyse the market. Score supply risk per category. Ask the people who know: production, quality and whoever handles shortages.
- Position. Plot the scores. For strategic items, judge whether you or the supplier holds more power and choose exploit, balance or diversify.
- Plan actions. Write one owner, one action and one date per category. A matrix without actions is a slide.
Kraljic matrix vs ABC analysis: what is the difference?
ABC analysis ranks categories or items by spend or consumption value only, usually into an A group (about the top 80% of value), B and C. The Kraljic matrix adds the second dimension, supply risk, and replaces raw spend with profit impact.
| Kraljic matrix | ABC (Pareto) analysis | |
|---|---|---|
| Dimensions | Profit impact and supply risk | Spend or usage value |
| Output | Four quadrants with a strategy each | Three priority bands |
| Catches low-spend critical items | Yes, as bottleneck items | No, they land in C |
| Effort | Needs judgment and cross-team scoring | Mechanical from a spend export |
| Best use | Sourcing strategy and risk planning | Inventory control, first spend triage |
In the example above, ABC analysis would put calibration services in the C band and forget them. The Kraljic matrix flags them as a bottleneck. Use ABC for a quick first cut of the spend data, then Kraljic to decide how to buy. Our comparison of supplier evaluation models such as 10Cs, QCDD and Kraljic covers how the matrix sits next to supplier scoring tools.
What are the limitations of the Kraljic matrix?
The main limitations of the Kraljic matrix are that it is a snapshot, it depends on subjective scores, and it looks only at the buyer's side of the relationship.
- It goes stale. Kraljic intended it for periodic planning. A leverage category can turn strategic within months when suppliers exit or a tariff hits. Rescore yearly and whenever a market shifts.
- Scores are judgment calls. Gelderman and Van Weele's 2003 paper in the Journal of Purchasing and Supply Management on "measurement issues" in Kraljic's model is largely about this problem. A written rubric and two or three scorers reduce it.
- It is often misapplied to suppliers. Score categories; use supplier tiering when you need to rank the suppliers themselves.
- It ignores the supplier's view. A supplier may see you as a small, unattractive account. Supplier preferencing models fill that gap.
- It does not say how to buy. It says what kind of approach fits. You still need sourcing tactics, such as deciding between single, sole and dual sourcing for a bottleneck item, or a weighted scoring matrix when you compare quotes for a leverage item.
FAQ
What is the Kraljic matrix used for?
The Kraljic matrix is used to decide how to buy each category of goods and services: where to push for price, where to secure supply, where to partner and where to simply automate. It is a core tool in category management and strategic sourcing.
Is the Kraljic matrix for suppliers or for purchase categories?
It was designed for purchase categories. Applying it to suppliers mixes categories with different risks into one score. Classify categories first, then look at which suppliers serve each one.
What is the difference between leverage and strategic items?
Both have high profit impact. Leverage items have many qualified suppliers, so competition lowers cost. Strategic items have few suppliers or high switching costs, so the priority is a stable, long-term relationship and a contingency plan.
Why are bottleneck items important if spend is low?
Because one missing low-cost item can stop production or shipments. Bottleneck items deserve safety stock, a second source or a service agreement even though they rarely show up in spend reports.
How often should you update a Kraljic matrix?
At least once a year, and again whenever a supply market changes: a supplier exits, lead times jump, a tariff or regulation lands, or your product changes.
Who created the Kraljic matrix?
Peter Kraljic, a McKinsey & Company director, published it in the September 1983 Harvard Business Review article "Purchasing Must Become Supply Management".