Category Tree in Procurement: How to Structure and Maintain Spend Categories

In procurement, reliable analysis starts with a simple question: What are we buying?

The answer is often more difficult than expected. Suppliers use different descriptions, invoices contain inconsistent information, accounting codes reflect financial reporting rather than procurement needs, and similar purchases may be recorded in different systems.

A category tree gives procurement a consistent structure for grouping products and services. This matters because procurement cannot assign category ownership, analyse spend or prioritise sourcing opportunities without first organising the spend data.

In this article, you will learn what a category tree is, how it supports spend analysis and category management, and how the Procurement Management Team can design and maintain one.

LHTS framework connection

Role: Procurement management
Supporting role: Tactical procurement
Process: Procurement governance, spend classification, spend analysis, category management and sourcing prioritisation
Level: Basic
Related course: Spend Analysis

Quick answer: What is a category tree?

A category tree is a hierarchical structure used to group the products and services purchased by an organisation.

It normally starts with broad spend areas and then divides them into increasingly detailed categories and subcategories. Procurement uses the structure to classify transactions, assign category responsibility, analyse spend and identify sourcing opportunities.

A category tree should reflect how procurement needs to understand and manage external spend—not simply copy the organisation chart, supplier list or financial chart of accounts.

What does a category tree look like?

A category tree moves from broad classifications to more detailed purchasing areas.

A simple example could look like this:

Level 1: Indirect spend

• Level 2: Information technology
  • Level 3: Hardware
    • Laptops
    • Monitors
    • Servers
  • Level 3: Software
    • Business applications
    • Security software
    • Software as a service
  • Level 3: IT services
    • Consultancy
    • Support
    • Application development

The lowest useful level is sometimes called the leaf category. This is normally the level where transactions are classified and where procurement can perform meaningful analysis.

The number of levels depends on the organisation. A smaller company may need three levels, while a complex international organisation may need four or five. More levels do not automatically create better analysis. The structure should only be as detailed as the organisation can classify, own and maintain reliably.

Why procurement needs a category tree

Raw spend data rarely arrives ready for procurement decisions. The same type of purchase may appear under several descriptions, suppliers or accounting codes.

For example, laptop spend could appear as:

• Computer equipment
• End-user devices
• IT hardware
• Office computer
• Portable PC
• Supplier product number 4321

Without classification, these transactions may look unrelated. Once they are mapped to the same category, procurement can see the combined spend.

A well-designed category tree helps procurement answer questions such as:

• How much do we spend in each category?
• Which suppliers receive that spend?
• Which business units are buying similar products or services?
• Where is spend fragmented?
• Which categories lack contracts?
• Where could suppliers be consolidated?
• Which categories should receive sourcing resources?
• Who is responsible for each category?

The category tree therefore turns transaction data into a management structure.

How the category tree connects to procurement management

The category tree is primarily a procurement management responsibility because it creates the structure within which category managers and buyers work.

The Procurement Management Team should decide:

1. The purpose of the category structure.
2. The level of detail required.
3. The principles used to define categories.
4. Who owns each part of the tree.
5. Who may request or approve changes.
6. How historical spend will be handled when categories change.
7. How the structure will be implemented across procurement systems.

Category managers should contribute specialist knowledge and propose improvements. Tactical buyers can identify classification problems during sourcing activities, while operative buyers may identify incorrect data during purchase-order and invoice processing.

However, local users should not change the structure independently. Uncontrolled changes quickly reduce comparability and confidence in procurement reports.

Defining the category, conducting spend analysis and assigning category responsibility are also core elements of category management work. The LHTS article on the key activities of a category manager identifies spend analysis and category definition as early priorities.

Where the category tree fits in the procurement process

The category tree supports several connected procurement processes.

Spend data collection

Procurement collects transaction data from sources such as purchase orders, invoices, contracts, payment systems and supplier master data.

Data cleansing

Supplier names, descriptions, currencies, duplicate records and incomplete information are corrected or standardised.

Spend classification

Transactions are assigned to categories in the approved tree. Classification may be performed manually, through rules, by suppliers, or with software and machine-learning support.

Spend analysis

Procurement analyses spend by category, supplier, business unit, geography, contract status and other dimensions.

Category management

Category managers use the classified information to define category scope, identify opportunities, prepare strategies and monitor performance.

Sourcing prioritisation

Procurement management uses category data to decide where market analysis, sourcing projects, negotiations or supplier-development activities should be initiated.

The category tree is therefore not only a reporting tool. It connects procurement data with responsibility and action.

Category tree or standard classification system?

An organisation can create its own category tree, adopt an external standard or combine the two approaches.

Custom category tree

A custom category tree reflects the organisation’s own products, services, market structure and management responsibilities.

Its main advantage is relevance. Categories can be designed around the decisions procurement needs to make.

Its main disadvantage is maintenance. The organisation must define the categories, document their meaning, manage changes and map historical data.

UNSPSC

The United Nations Standard Products and Services Code is a global hierarchical system for classifying products and services. It can support procurement optimisation and company-wide spend visibility.

An organisation can adopt UNSPSC directly or map its internal categories to UNSPSC codes. Mapping allows the company to maintain an internally relevant category tree while retaining a connection to an external standard.

CPV

The Common Procurement Vocabulary provides a standard classification system for public procurement in the European Union. Its purpose is to standardise how contracting authorities describe the subject of procurement contracts.

CPV may be relevant to public-sector sourcing, but it is not automatically the best internal management structure for every organisation.

HS codes

The Harmonized System classifies goods for customs tariffs and international trade statistics. Its international foundation uses six-digit codes.

HS codes are essential for customs and landed-cost work, but they serve a different purpose from a procurement category tree. A procurement category may contain several HS codes, and one HS classification may need to be mapped differently depending on the organisation’s category strategy.

Read the LHTS article on HS codes for a deeper explanation of how customs classification affects RFQs, landed cost, contracts and international sourcing.

GTIN and EAN

A Global Trade Item Number uniquely identifies an individual trade item. GTIN-13 is also commonly called EAN. These identifiers help distinguish specific products, but they do not group procurement spend into strategic categories.

This is an important distinction:

GTIN identifies the item. A category tree classifies the type of purchase.

How to design a procurement category tree

1. Define the decisions the tree must support

Start with purpose rather than category names.

Determine whether the structure must support:

• Spend analysis
• Category ownership
• Sourcing pipelines
• Contract coverage
• Supplier management
• Risk reporting
• Sustainability reporting
• Budget planning
• External benchmarking

A category tree designed only for reporting may differ from one intended to assign global category responsibility.

2. Analyse existing spend data

Review current transactions, supplier descriptions, accounting codes, purchase-order structures and existing local categories.

This identifies the language already used by the organisation and shows where classification problems are likely to occur.

3. Create a clear hierarchy

Start with broad areas and move towards useful analytical detail.

Categories at the same level should represent a reasonably similar degree of detail. Avoid placing a broad category such as “Professional services” beside a narrow category such as “Temporary laboratory technicians.”

4. Define every leaf category

Each leaf category should have:

• A clear name
• A short definition
• Examples of included purchases
• Examples of excluded purchases
• A category owner
• Relevant external-code mappings where needed

Definitions are essential because category names alone are often open to interpretation.

5. Test the structure with real transactions

Apply the draft category tree to a representative sample of invoices or purchase-order lines.

Testing will reveal overlaps, missing categories and descriptions that users interpret differently.

A category tree that looks logical in a workshop may still be difficult to use with real spend data.

6. Assign governance

Define who can request, review and approve changes.

A practical model is:

• Category managers propose changes.
• Procurement data specialists assess classification consequences.
• Finance and system owners assess reporting and technical consequences.
• The Procurement Management Team approves structural changes.

7. Implement the tree across systems

The approved structure may need to be reflected in:

• Spend-analysis platforms
• Procure-to-pay systems
• Sourcing systems
• Contract repositories
• Supplier portals
• Data warehouses
• Reporting tools

Where systems use different coding structures, controlled mapping tables may be required.

Practical example: Classify the purchase, not the supplier

Assume that one technology supplier provides:

• Laptops
• Software subscriptions
• Implementation consultancy
• Maintenance services

A common mistake is to classify all invoices from that supplier as “IT hardware” because the supplier was originally selected to provide computers.

The result is misleading spend data.

The purchased item or service should drive the category assignment:

• Laptops belong under IT hardware.
• Subscriptions belong under software.
• Implementation belongs under IT consulting or implementation services.
• Maintenance belongs under IT support or maintenance services.

Supplier classification and spend classification are different dimensions. One supplier can serve several categories, and one category can contain several suppliers.

How to maintain the category tree

A category tree is not a one-time project. It must remain stable enough to support comparisons but flexible enough to reflect meaningful business changes.

Procurement management should use three types of review.

Regular data-quality review

Review unclassified transactions, miscellaneous categories, unusual mappings and categories with unexpectedly high or low spend.

Scheduled governance review

Conduct a structured review, normally once or twice per year, with category managers, procurement analysts, Finance and system owners.

Event-driven review

Assess the tree when the organisation introduces a major new technology, completes an acquisition, changes its operating model, enters a new market or creates a new category-management responsibility.

Changes should be documented in a change log. The organisation must also decide whether historical transactions will be reclassified. Without this decision, reports before and after the change may not be comparable.

Common mistakes when building a category tree

Copying the chart of accounts

Financial accounts are designed for accounting and budget control. They do not always provide the market or supplier perspective needed by procurement.

Copying the organisation structure

Business units change more frequently than purchasing markets. Categories should describe what is purchased rather than which department buys it.

Creating too many categories

Excessive detail makes classification expensive and unreliable. A category with very little spend or no clear owner may not need to exist as a separate category.

Creating categories that overlap

When users cannot determine which of two categories should be selected, classification becomes inconsistent.

Using supplier names as categories

Suppliers may provide several products and services. Categories should represent the purchase, not the supplier.

Failing to define ownership

A category without an owner is unlikely to be analysed, managed or updated properly.

Changing the tree without managing history

Renaming, merging or splitting categories without remapping historical data can make trends appear that are caused by structural changes rather than real purchasing behaviour.

Treating external standards as automatic answers

UNSPSC, CPV and HS provide valuable structures, but they were created for particular purposes. Procurement management must decide how each standard supports the company’s own analytical and management needs.

Learn more through the Spend Analysis course

The category tree provides the classification structure, while spend analysis turns the classified data into procurement insight.

The Learn How to Source Spend Analysis course explains how procurement collects, cleanses, classifies and analyses expenditure data. It also explains how spend analysis supports category management and data-driven procurement decisions.

The course is a natural next step for buyers, category managers and procurement leaders who want to understand how raw purchasing data becomes actionable category information.

Frequently asked questions

What is a category tree in procurement?

A category tree is a hierarchy used to organise purchased products and services into categories and subcategories. It supports spend analysis, category ownership, sourcing and procurement reporting.

Who should own the category tree?

Procurement management should own its governance. Category managers, Finance, procurement analysts and system owners should contribute to its design and maintenance.

How many levels should a category tree have?

There is no universal number. The structure should contain enough detail to support procurement decisions without creating categories that the organisation cannot classify or manage consistently.

Is UNSPSC a category tree?

UNSPSC is a standard global classification hierarchy for products and services. It can be used as a procurement category tree or mapped to a company-specific structure.

What is the difference between a category tree and a chart of accounts?

A category tree groups spend according to procurement and supply-market needs. A chart of accounts groups transactions according to financial-accounting requirements.

Should categories be based on suppliers?

No. Categories should normally be based on the product or service purchased. A supplier may deliver products and services belonging to several categories.

How often should the category tree be updated?

Data quality should be monitored continuously. Formal governance reviews can be conducted once or twice per year, with additional reviews when significant business or market changes occur.

Conclusion

A category tree is the foundation for structured procurement information. It gives the organisation a common language for describing spend, assigning category responsibility and identifying sourcing opportunities.

The Procurement Management Team should ensure that the tree is relevant, clearly defined, properly governed and consistently applied. External standards such as UNSPSC can provide useful reference points, but the final structure must support the organisation’s own procurement decisions.

The practical next step is to review the current category tree against actual transaction data. Look for unclassified spend, overlapping definitions, excessive use of miscellaneous categories and categories without a clear owner. These gaps show where the structure or its governance needs improvement.

About Learn How to Source

Learn How to Source (LHTS) is an online platform based in Sweden, offering a range of procurement courses accessible globally. It serves as a community where procurement experts share their knowledge through online courses, designed for various experience levels from introductory to expert. Courses are concise, about 30 minutes each, and cover different aspects of procurement, tailored for different buyer roles. The courses focus on practical knowledge, presented by seasoned professionals, and include quizzes and certificates. They can be accessed from any device, emphasizing microlearning for flexibility and efficiency.

Note. Illustration to the blogpost “Product classification systems available to categorize spend” was created with Chat-GPT on September 15, 2024.