Category Segmentation

In addition to spend data and contracts addressable spend categories should be classified according to strategic importance. This will help us in the future to prioritize strategic sourcing activities based on suppliers/category criticality to the company.

The Kraljic Portfolio Model is a commonly used method for this purpose though there are many other models and almost all of them mimic the Kraljic model this or that way. Those other models just change the x and y axis criteria mainly based on the priorities of the categorization exercise.

The Kraljic portfolio model helps map out category segmentation in two dimensions:

Profit Impact: volume or value purchased, impact on supply chain “value-add”, business growth potential or dependency

Supply Risk/Criticality: product availability, number of suppliers, ease of switching a supplier, availability of substitutes

Determine profit impact by answering the questions below:
Is the category total value important in the company’s total spending?
Do the client’s end customers perceive that this category adds significant value?
Does the category differentiate the end product significantly?
Would a category failure affect the client’s end customer satisfaction?

Determine supply risk by answering the questions below:
What is the market internal competition?
Can you easily switch to another category?
What is your buying power for this category?
What is the bargaining power of sellers?
Can new entrants be easily found and invited to tender?

In the end category segmentation is all about the approach we will take in supplier relationship management and understanding the type of value category/supplier provides. Hence we can also determine which and how many resources to allocate for supplier segments.

After the segmentation is done we have a strategic direction for each category:
1. Leverage products
Leverage products allow the company to exploit its full purchasing power through tendering, target pricing and product substitution

2. Strategic products
The company should be maintaining good relationships with strategic partners

3. Bottleneck products
Bottleneck products should be handled by volume insurance, vendor control, security of inventories and backup plans.

4. Routine products
Routine and non-critical products require efficient processing, product standardization, order volume and inventory optimization

Another approach to take after the segmentation is complete is to shift categories to neighbour segments. It is called Moving in the Matrix:

- Leverage products -> Strategic products
Develop a strategic partnership
Exploit buying power

- Strategic products -> Leverage products
Accept the locked-in partnership
Maintain strategic partnership
Terminate undesirable partnership, find new supplier

- Bottleneck products -> Routine products
Reduce dependence and risk, find other solution
Accept the dependence, reduce the negative consequences

- Routine products -> Leverage products
Pooling of requirements
Individual ordering, pursue efficient processing

Total Cost of Ownership

Total Cost of Ownership (TCO) is a concept used to determine direct and indirect costs and on that basis compare suppliers’ proposals/bids or segment suppliers for strategic sourcing analysis.

In our case TCO will be used in two phases – it will help to determine cost drivers for each category and estimate savings ranges.

TCO generally could be characterized by eight components:
1. Acquisition (Purchase Price, Freight, Engineering, Installation, Customization)
2. Operating (Labor, Utilities, Fuels)
3. Training (Software Upgrade, Personnel Training)
4. Quality (Cost of Returns, Durability of Goods)
5. Maintenance (Spare Parts, Maintenance Labor)
6. Warehouse (Storage)
7. Environment (Electricity Consumption, Electricity Savings)
8. Salvage (Salvage Value (Positive or Negative) of the assets at the end of its life)

So TCO is not just about the unit price reductions. TCO can be reduced by demand management, improved process efficiencies and better supplier management. While focusing on achieving TCO savings we can look closely at e.g. delivery lead time, days in inventory, parts shortage or defect level per part.

Acquisition cost may not be a representation of the total cost of a commodity, though very often acquisition and operating costs are the only ones taken into consideration when choosing the supplier. The following example shows why:

Total Cost of Ownership

At first glance the acquisition cost of the first item from Supplier 1 seems to be the highest but when we analyze TCO for this item across all three suppliers we see that actually
Supplier 1 provides us with the best offer as TCO for his offer is the lowest. Therefore Supplier 1 delivers better value.

Contracts review

In order to better understand where the main savings opportunities lie the existing contracts are to be reviewed. This exercise will help to determine which categories represent potential savings, track expiring contracts to identify new negotiating opportunities, and identify cost savings opportunities.

Within each contract, the following information can be found:
Current contract status
Contract expiry date
Spend covered by contract
Product specification
Internal customers
Agreements to lock the prices
History of price increase
Contract change notices
Other terms and conditions

At the high level phase of the spend analysis, there are three main items that must be checked before starting opportunities assessment:
1. Contract expiry date
2. Penalty for early termination
3. Terms and conditions
These items help us to identify addressable spend and prioritize categories for which contracts are expiring in near future.

General steps of contract review:
Identify existing contracts
Identify contract termination penalties
Identify contract renewal cycle
Identify recently sourced contracts
Analyze total annual spend and relevant contracts to determine the proportion addressable through strategic sourcing initiatives
Evaluate total cost of ownership to determine cost drivers
Evaluate project duration
Evaluate resources required
Evaluate supplier switch costs
Evaluate potential disruption to operations (risks)
While evaluating contracts we usually build the following table in order to capture the key data.

which later when completed transforms into timeline for savings summary.

This summary helps us to identify when new negotiation opportunities become available and whether the benefits of early termination outweigh its costs.

What is addressable spend?

Today’s post is short but it is a really important one.

Addressable Spend is defined as spend that can be impacted through sourcing activities.
In order to determine if spend is addressable one may require to conduct interviews with internal clients referencing to the preliminary spend data analysis.
We should also determine “sourceability” of each category e.g. taxes, depreciation, transfer pricing, royalties, wages cannot be regulated.

The following questions may help to determine spend addressability:
1. Is the relationship with this supplier crucial regardless of the opportunity cost?
2. Is there a preference for suppliers for special reasons?
3. What are the political factors to be considered when assessing opportunities?
4. Is there a management preference for suppliers? (Esp. in private sector)
5. Are there upcoming projects that would hinder the sourcing activities? E.g. a CPU manufacturer was selected earlier for the launch of a new mobile product. This cannot be re-sourced because the whole product technical design is based on certain CPU parameters.
6. Are there any requirements for quality, service level? I.e. only a certain suppliers can meet the requirements.

Preliminary Data Analysis. Five sample spend data views and the insights each offers

So we cleansed, normalized and categorized spend data. Now we have a better visibility of spend. We can do a high-level data scan to understand how spend is allocated across commodity categories, suppliers, customers, contract vehicles and purchasing methods. The data scan is really helpful as we can better understand spend across the various spend parameters.

The data scan is also important as we can spot and correct missed data gaps and redundancies. We will also be able to outline addressable and unaddressable spend. After that redundant and unaddressable spend data can be eliminated.

Key spend observations:
Spend by category is the most fundamental observation which provides a spend overview to strategic sourcing team. Examining top categories also reveals whether the categorization structure is too wide or too narrow. For example, if the top 5 categories represent 90% of total purchased spend, the categorization structure might be too wide to be able to provide insights on category focus.

Top vendors/suppliers representing most purchase spend:
Examining spend by supplier reveals the degree of spend fragmentation by category, as well as potential opportunities for improving negotiating strength. If spend is highly fragmented, it is of value to look deeper to understand the reasons why. For example, there may be many suppliers because of the abundance of small businesses or decentralized procurement organization model. If spend is concentrated with a relatively few number of suppliers, exploring different ways to reallocate spend or expanding our supplier base may improve negotiating strength.

Spend data can also be viewed across various dimensions, such as spend by vendor, spend by location, spend by cost centre, spend by type of contract, and spend by commodity.

Here are five sample spend data views and the insights each offers for understanding spend:

Sample Spend View #1: Commodity Spend by Organizational Unit
This view helps to understand which part of the organization is generating the greatest commodity spend. Such data view is useful for prioritizing commodity requirements by organizational unit and identifying key stakeholders within the user community. Organizational views should consider contracting offices managing the procurement process, as well as end users.

Sample Spend View #2: Spend to Transaction Ratios
Comparing the ($) amount of spend to the number of supplier transactions is one indicator of Procurement efficiency. If we see numerous transactions made for a relatively small amount of spend, it is a clear indicator that most likely there may be opportunities for process automation and streamlining.

Sample Spend View #3: Commodity Purchase Price by Supplier
Variations in purchase price across suppliers and across multiple contracts with the same supplier can reveal opportunities for unit price reductions. We may also want to ask the broader question: “what is the real price of the commodity purchased?” Analysis of the supplier cost structures, purchasing processes, and other price drivers may help to identify means to further reduce purchase prices.

Sample Spend View #4: Supplier Management Workload
Another view of spend fragmentation looks at the size of annual spend and the implications for acquisition workload. If the annual spend with a large number of suppliers is very low, this could be driving up administrative costs. It may add value to consider re-allocating or consolidating spend.

Sample Spend View #5: Historical Spend Compared to Current Budget
Comparing each organizational unit’s historical spend to the current year budget is a good way to validate the integrity of the Spend Analysis. If significant discrepancies emerge, it is advised to re-check the calculations and investigate reasons behind such discrepancies.

Data Categorization. Mapping Data into Categories

Now that we have cleansed data we need to prepare it for the analysis by organizing the data into logical commodity categories. This commodity structure forms the foundation of our spend analysis. A preliminary data analysis can be performed at this stage across different dimensions, such as spend by organizational unit and by supplier.

The categorization structure must be explicit enough to ensure clear spend visibility. Each category must be clearly defined. A good categorization structure has the following features:
- Category titles are unambiguous and mutually exclusive
- Hierarchical structure (i.e. Class and Commodity)
- Products/services appear in one category; categories each have only one parent
- Products/services are grouped according to dominate usage in world market

A commonly used categorization structure is UNSPSC.
UNSPSC, also known as “United Nations Standard Products and Services Code”. That is a globally used classification hierarchy for products and services.

UNSPSC is an open standard taxonomy.

Open standard means:
- There are publicly available specifications
- Those standards and taxonomy are not proprietary (i.e. available to all to implement, no license fees for usage, no restrictions on sharing codes with partners)

Taxonomy means:
- A classification system of products and services bought
- A hierarchical tree structure which enables “drill down” and “roll up” analysis

There are obvious benefits of using UNSPSC:
1. Segments for raw materials, industrial equipment, components and supplies, end-use products, and services are available in over 20,000 categories
2. Collaboration with customers or suppliers through the use of a common classification system
3. “Roll up” and “drill down” analysis better identification opportunities
4. Responsive to the marketplace

Strategic sourcing team has to decide how much detail they would like to drill down in each category at this high level stage of the spend analysis.

So then we classify spend data based on the defined categorization structure. Additional information can be brought into consideration to gain insights into appropriate commodities. This includes researching existing contracts, as well as looking at categorization standards within the relevant commodity industry.

Contract research
For larger transactions, we may revisit the actual contract documents to better understand line-item purchases. These purchases can then be grouped into logical commodity categories.

For smaller transactions, where the cost of detailed contract research outweighs the benefits, it could be a better idea to extrapolate appropriate commodities based on information gathered from the analysis of larger contracts. For example, if large suppliers under a given FSC code typically provide a certain product or service, it can be assumed that small suppliers with the same FSC code provide similar product or service.

To supplement the contract research we may also review other documents from the purchasing paper trail such as purchase orders and invoices.

Industry research
Insights on commodity structure can also be gained by looking at how top suppliers group their products and services. Typically, product/service categories converge within an industry (particularly those that are more mature) to reflect a relatively stable classification structure.

It is also recommended to look at standard industry classification systems such as FSC or NAICS to understand how they breakdown spending into different categories.

Industrial products and services categorization standards:

FSC – federal supply codes are used by the United States government to describe the products, services, and research and development purchased by the government.

NAICS – North American Industry Classification System is used by business and government to classify business establishments according to type of economic activity (process of production) in Canada, Mexico and the United States

Cleansing and Normalizing Data

The next step after we collected data is to review data set for completeness, cleanse and normalize data.

Very often collected data is incomplete. It is recommended to address data incompleteness with the following options:
- Return to existing data sources to try to find missing data
- Seek out new data sources that may be able to fill in the gap
- Use data extrapolation and other statistical estimation techniques to fill critical gaps (though it could be an arguable approach as the output will not present factual data but only approximate numbers)

On the other hand, data redundancies should be identified and eliminated. Data overlap is common especially when combining data from multiple sources. This can be achieved by reviewing data records side by side and identifying any duplicates. Assistance from the original data sources may be required to resolve any records in question.

After that the entire data has to be scanned in order to identify errors such as:
- Mismatched data fields (e.g. a supplier name is listed in the line item description)
- Spelling errors
- Data miscodes (e.g. an incorrectly assigned FSC code)
- Other questionable data (e.g. a purchase date of 2080)

To confirm and correct the most significant data errors, it may be necessary to refer to the original documents (contracts, agreements). The original data sources should be consulted regarding any major uncertainties about the data.

Another concern is data consistency. Data inconsistencies are inevitable when combining data from numerous sources. The most common issues include:
- Inconsistent units of measure
- One data source might truncate dollar figures to the nearest thousand ($1,000 = $1,000,000), whereas another data source might write out the entire figure.
- Inconsistent supplier names
- A single supplier like Dell Computers, for example, might also be listed as Dell Comp., DELL, Dell, or Dell Inc.
- Parent-child relationships not being captured
- Companies treated as different companies, such as Smith International vs. Wilson vs. Van Leeuwen Pipe & Tube
- Recent mergers or acquisitions or company name changes not being captured – such as Southern Bell Company vs. BTE, Suncor vs. Petro Canada

Because this can be a very time-consuming task, a general rule of thumb is to focus on standardizing the 20 percent of suppliers, who typically account for 80 percent of the total spend.

Information must be validated with budgets, data owners and other sources. Comparing historical spend by business unit to current year spend budgets provides an excellent check of analysis integrity.

Data Requirements for Spend Analysis

After we defined Scope and Objectives of the spend analysis we have to define Data Requirements.
Defining data at the beginning of the process insures that we are consistent in the process of data collection across different sources.

At a minimum the following data elements are to be collected:
1. Commodity Purchased – line-item detail about the product/service
2. Date of Purchase
3. Purchase Amount
4. Purchase Quantity
5. Requisitioners – the requisitioners or business units that generated the requirement
5. Procurement Department – the procurement office handling the contract
6. Supplier Name
7. Purchase Method – the contract or other method (e.g. purchase card, purchase order) used to make an acquisition

Now we know what kind of data we are looking for. Let’s identify Data Sources. It is better to start this exercise by speaking with the main stakeholders directly involved with procurement transactions. There are two types of stakeholders we may need to talk to internal and external stakeholders.

Internal Stakeholders:
1. Procurement – employees who manage supplier contracts and track procurement data.
2. Finance/Accounting – employees who track expenditures through financial management systems.
3. Requisitioners/Business Units – employees who submit the initial requests or task orders leading to an acquisition.

External Stakeholders are Suppliers, or the product/service providers whose sales records can be used to validate spend.

When communicating with the main stakeholders to request for necessary data it is important to make explicit and detailed data requests.

Data requests may cover:
- A brief overview of the Spend Analysis initiative and its benefits
- An overview of how spend data will be used
- A description of the data requested, including specific data elements

- The time period for which data is needed
- The preferred format for receiving data, such as spreadsheet and database
- A proposed deadline for data collection

Usually procurement and reporting systems are ok for high level spend analysis but still there could be inaccuracies and data gaps which should be covered later depending on the approach taken and criticality of the data missing:

Available data sources can typically be uncovered through an understanding of the internal processes and policies followed for purchasing.

The typical data sources include:
1. Financial systems (e.g. A/P, supplier master file, GL)
2. Purchasing and AP modules of ERP systems (e.g. SAP MM and AP, Oracle Financials)
3. Purchasing systems (e.g. Ariba), Purchasing card management reports
4. Materials management systems (e.g. MRP)
5. Maintenance systems (e.g. Maximo)
6. Manual systems (e.g. spreadsheets)
7. Hard copy invoices
8. Purchase orders and contracts

As soon as we have data sources identified we can proceed with Data Collection.
There are various methods to collect data. In some cases it is recommended to utilize all of the following methods in order to ensure data accuracy, but this should be decided case by case depending on the data quality and availability, and budget available:

1. Manual – collate spreadsheets
2. Download from A/P and other systems such as SAP
3. Queries from data warehouse
4. Third party applications, such as Spend Radar, Iasta and IBM’s Emptoris Spend Analysis,

High Level Spend Analysis. Defining Scope and Objectives

Let’s look in more detail into high level spend analysis. The High Level Spend Analysis usually consists of four main steps:

1. Data extraction and normalization
2. Data mapping into data categories
3. Contracts and spend review
4. Indentification and prioritization of opportunities

Before we decide which types of data we want to include in our spend analysis we should consider the following three parameters which will guide us through data collection and analysis:

1. Organizational scope
2. Time Frame
3. Types of Spend

1. Organizational scope
A Spend Analysis can be conducted at an enterprise-wide level or within a specific department or division. There are three different factors that may affect organizational scope:

a. Known location(s) of commodity spend
b. Organizational complexity
c. Available resources

A very good illustration of complex organization (where you have to consider the above factors) is Microsoft. This giant employs more that 92 000 people in 112 countires and its operations are supported by 160 procurement employees in 33 countries managing $16 billion of spend.

To define the scope we should also address these key questions:

a.What is being purchased?
The goal is to get the greatest detail about the specific products or services being purchased. Typically, commodity categories with line-item descriptions of what falls within each category are required. E.g commodity category “IT hardware” may include – laptops, hard drives, network cards and etc.

b. Who is purchasing these products or services?
In addition to understanding how the procurement department processes each spend request, it is crucial to get information about the way the end user requests products or services.

c. From whom are the products or services being purchased?
E.g. details on the specific suppliers who are delivering the products or services requested.

d. How are these products or services being purchased?
E.g. the contract or off-contract vehicles being used to make these purchases.

2. After the organizational scope has been defined we have to decide on the Time Frame to include in the analysis. We should determine the start and end dates of the period to collect the data. As a general rule, the data for the most recent full fiscal year should be used. Less data makes it difficult to find patterns and understand factors such as seasonality. Thus, it is important to include at least 24 to 48 months of data, which is sufficient enough to identify seasonality, outliers and other important implications.

3. The types of spend on which the analysis will concentrate are another critical parameter.
For example, should the following be included?
- Off-Contract Purchase Card Expenditures
- Research & Development Spend
- Grants Spend

The issue of whether to cover spend with all suppliers, or to focus on a specific category such as small businesses is to be taken into consideration as well.

In addition to the key spend questions, there may also be other specific objectives associated with Spend Analysis. For example, the following objectives were established to guide the IT Spend Analysis initiative:

Provide an accurate and detailed buildup of the total spend for the IT Equipment and Services commodity.
Develop IT commodity segments from a sourcing perspective
Understand the importance and relative priority of various IT segments
Understand current sourcing practices as evidenced through suppliers, contracts, buying offices, and other entities connected to purchasing
Develop a spend baseline for performance measurement

Spend Analysis Objectives

There are four main Spend Analysis objectives which help to:

1. Develop an organization-wide view of spend that is comprehensive, detailed, and accurate (e.g. how much we spend on stationary, meals, printing equipment, and further going down to specifics: how much we spend on pencils, pens and staples; water, business lunches, meeting snacks, coffee; printers. toners, paper. services)

2. Identify opportunities for reducing total costs by coordinating purchasing across the organization (e.g. transportation department buys computers from supplier A and finance department from supplier B, is there a reason for that? If not let’s buy computers from one supplier only and use it as a leverage to negotiate discounts or other perks)

3. Establish a spend baseline against which to measure and compare future spend activity (e.g. next year we can have a look and see if we are buying computers cheaper than the last year or not)

4. Prioritize strategic sourcing initiatives for further analysis and implementation (usually if there is a big spend volume there are several sourcing waves based on priorities and impact on the business those spend categories have, let’s say if we manufacture aircrafts parts supply is critical for the business and their cost impacts the sales price but the price we pay for the office stationary is not critical so we can deal with this category later)

As a result of the Spend analysis we will have visibility into our total spend across different dimensions, such as commodities, customers, suppliers). We will identify opportunities to improve purchasing efficiency and optimize spend across the organization. The company will also gain a better control over spend by getting a reliable baseline against which to measure future spend and performance. On-going monitoring of spend performance and internal compliance with approved sourcing strategies is another benefit to mention.

We should distinguish between the two types of spend analysis:

1. A High-Level Spend Analysis covering all categories of spend is performed as part of an initial Opportunity Assessment to identify the best commodities, or spend categories, on which to focus Strategic Sourcing efforts.

2. Once a commodity is identified for Strategic Sourcing, a Detailed Commodity Spend Analysis/Category Analysis is performed to understand the specific purchasing history and patterns for that commodity/category.