On this page
Key takeaways
- Centralized procurement means one team buys for the whole company. Decentralized means each unit buys for itself. Most real organizations mix the two.
- Centralizing pays when spend is similar across sites and volume gives buying power. It hurts when needs are local and speed matters.
- Decide by category, not for the company as a whole. Commodities often centralize well. Site-specific services often do not.
- Centralizing and enforcing are separate. A central team that cannot serve requests quickly will be bypassed.
- Use a scorecard of a few factors to make the choice visible, and revisit it each year.
Free PDF: Centralize or not decision scorecard. Get it below
Take a group with six factories. Each buys its own cleaning supplies, safety gloves and office paper from local suppliers, and nobody has ever compared prices across plants. A finance analyst pulls the numbers one afternoon and finds the same glove priced at two different levels in two plants forty kilometers apart. The group then centralizes glove buying, saves money, and a month later a plant manager complains that his gloves now arrive late. Both facts are true, and that is the whole debate.
The example is invented, but it shows why the question has no universal answer. Each model fixes one problem and creates another.
Centralized procurement is a model in which a single team makes the buying decisions for the whole organization. Decentralized procurement leaves those decisions with individual business units or sites. A hybrid model assigns each category to whichever approach suits it.
What does each model do well and badly?
| Factor | Centralized | Decentralized |
|---|---|---|
| Price | Volume is pooled, so buying power is stronger | Each unit negotiates alone, so prices vary |
| Speed | Requests queue at one team, so response can slow | Local buyers act quickly on local needs |
| Control and compliance | One set of rules, easier to audit | Rules differ by unit and are harder to check |
| Supplier base | Fewer suppliers, easier to manage | More suppliers, more duplication |
| Local fit | Distance from users, risk of poor fit | Close to users, better fit for local needs |
| Cost to run | Needs a capable central team | Needs capable buyers in every unit |
The table describes tendencies, not guarantees. A badly run central team can be slower and dearer than a good local buyer.
What happens when buying is not controlled?
Spend outside the agreed process is the usual price of too little structure. Our guide to maverick spend covers that problem directly, including how to measure it. This page is about the structural choice that can reduce it, and the two topics are related but different: you can centralize and still have maverick spend if the central route is too slow.
How do you decide for each category?
Score each category on a handful of factors. The scorecard in the download below uses the six here. It is our own tool, built from common practice, and you should adjust the weights to your business.
| Factor | Points toward central | Points toward local |
|---|---|---|
| Spend similarity | Same item or service needed at many sites | Needs differ site by site |
| Volume benefit | Suppliers give real discounts for volume | Little price gain from pooling |
| Risk | Failure is costly or regulated | Failure is easy to fix |
| Speed need | Orders are planned | Orders are urgent and unpredictable |
| Local supply | Suppliers operate nationally | Good suppliers are only local |
| Expertise | Specialist knowledge is scarce | Local buyers know the market well |
Add up how many rows point each way for a category. A clear majority gives you the answer. A split result suggests a hybrid, where the central team sets the contract and local buyers place the orders.
Free PDF
Centralize or not decision scorecard
A six factor scorecard to decide, category by category, whether to centralize, keep local or run a hybrid, with a pilot plan and service level draft.
- Six factor scoring grid for each category
- Hybrid pattern chooser
- Service level draft for the center
- Before and after measure sheet
What does a hybrid look like in practice?
- Central contract, local ordering. The center negotiates the agreement and each site orders from it. This keeps the price and removes the queue.
- Central strategy, local sourcing. The center sets standards and a preferred list. Sites choose within it.
- Central for strategic categories, local for the rest. IT, energy and logistics buy centrally. Site services and small consumables stay local.
- Shared services for transactions. Orders and invoices run through one team, while category strategy sits with business leads.
McKinsey's survey of more than 300 procurement leaders found two thirds have split procurement into strategic and transactional tracks. That is the hybrid idea at company scale: judgment work in one place, speed work in another.
How do you make a change stick?
- Start with one or two categories. Pick ones with similar needs across sites and a clear price gap.
- Agree service levels. Tell sites how fast the center will respond. A promise of "within two days" is better than a promise of "quickly".
- Measure before and after. Record price, cycle time and complaints for three months before the change.
- Protect the exception route. Sites need a fast, visible way to ask for something unusual. If the route is closed, they will go around it.
- Review after six months. Keep what worked, change what did not, and say so openly.
Deloitte's 2025 survey found siloed ways of working was the top barrier to delivering value at 57 percent. A move to central buying that ignores those silos tends to move the conflict, not remove it. The procurement change management guide covers how to bring people with you.
What mistakes do teams make?
- Centralizing everything at once. The central team drowns and service drops.
- Counting the savings but not the delay. A cheaper glove that arrives late shuts a line, and that cost never reaches the report.
- Treating the choice as permanent. A growing company may centralize at 300 staff and loosen at 3,000.
- Forgetting the people. Local buyers lose authority in a central model. Tell them what their role becomes before the announcement.
For the wider model choices, see the procurement operating model guide, and for team design read procurement organization structure. The overall plan sits in the procurement strategy pillar.
Sources
- McKinsey procurement report, as reported by Digital Commerce 360: more than 300 procurement leaders, November 2025.
- Deloitte, Global CPO Survey 2025: more than 250 CPOs in 40 countries.
- The comparison tables and the six factor scorecard are our own working tools, not research findings.
Frequently asked questions
What is centralized procurement?
It is a model where one team makes the buying decisions for the whole organization. It pools volume for better prices and applies one set of rules, at the cost of speed and local fit.
What is decentralized procurement?
It is a model where individual business units or sites make their own buying decisions. It gives speed and local fit, but prices vary and suppliers multiply.
Is centralized or decentralized procurement better?
Neither wins everywhere. Centralize categories with similar needs and real volume benefit. Keep local the categories that need speed or local supply. Most organizations use a hybrid.
What is a hybrid procurement model?
A hybrid assigns each category to the approach that suits it. A common form has the center negotiate contracts while sites place the orders against them.
How do you move from decentralized to centralized procurement?
Start with one or two categories with similar needs and a clear price gap. Agree service levels, measure before and after, protect a fast exception route and review after six months.