News Stay informed about the latest enterprise technology news and product updates.

Optimizing the sourcing process through spend analysis

Through spend analysis, a purchasing organization can perform supply base rationalization, improve contract compliance, reduce item proliferation, eliminate maverick spending and identify supply chain risk.

Procurement executives have the goal of securing products and services of the highest quality at the lowest possible cost for their operations. In the current economic climate, procurement executives are also responsible for managing supply risk for components in their supply chain. But to accomplish these objectives, they need insight into what their company buys, from whom and for how much. One way to gain this insight is through spend analysis.

Spend analysis is the process of gaining 20/20 visibility into what was spent within the organization over a certain time period, on which items, through which suppliers and at what purchase prices. Spend analysis also presents time-based trends of this information by metrics such as organization unit, geography, commodity and source category. With visibility into such information, purchasing managers can identify opportunities to save money and manage risk.

Some analysts claim that an organization can save 2% to 5% of its total spend by identifying and executing opportunities such as supply base rationalization or contract compliance. To show a similar gross margin increase, a typical manufacturer would need to increase its revenues by about 15% -- a tall order. Insights gained through spend analysis can also help companies negotiate lower purchase prices with suppliers and limit margin erosion.

Obstacles to spend analysis

Organizations need to start their spend analysis with data that is accurate and complete. Most companies expect that their ERP systems have accurate and complete spend data, and that this data resides in a few easily accessible systems. Unfortunately, such assumptions are optimistic for the following six reasons:

Data typically resides in multiple heterogeneous source systems, making it difficult to aggregate into a single, consolidated view. For example, companies growing through M&A may have disparate source systems. Depending on the maturity of sourcing processes within an organization, there may be several elements of relevant data residing in spreadsheets and legacy systems.

  • Different codes are often used to describe the same supplier or commodity across different systems. For example, a supplier called HP may be 'coded' as HP in one set of transactions and as Hewlett-Packard in another. As a result, an aggregate spend with a supplier may be calculated inaccurately, and an opportunity for leverage with that supplier may be lost.
  • The item code in most ERP systems does not relate to any standard classification. Consequently, it becomes difficult to aggregate similar and/or equivalent data. For example, components with similar specifications but sourced from different suppliers are typically defined as separate item codes. As a result, it is difficult to aggregate the purchases of such components in order to understand the dollar amount of purchase of that type of component across multiple suppliers, as well as see the price variance of similar components across suppliers. Without being able to aggregate such information (other than by arduous manual means), it is difficult to identify opportunities to save money by combining spend across commodities, locations, suppliers and programs.
  • Few ERP systems identify relationships between suppliers. For example, your ERP system may not tell you that Lab Safety Inc. is a subsidiary of WW Grainger. If it did, you'd realize that you are spending a lot more money with WW Grainger than you originally thought, and you could use that information to gain negotiating leverage.
  • Data is rarely classified at the right level of granularity. For example, at level 1, a component may be identified as category "valve." If this category comes up as a "problem" category to take action on, it may be that there are several suppliers that represent this category. However, the problem may be more specific in this category to a level 3 granularity such as a ball valve (which may fall under a level 2 sub-category of rotary valve). Tracking the issue to this category and correlating to the supplier responsible for this sub-category would then give a specific actionable item for negotiation.
  • Few ERP systems contain information, such as minority status of suppliers or debarred list, that can help you to take advantage of tax breaks or meet certain regulatory compliance requirements.

Hence, the first step in analyzing spend is to ensure that the data is complete, centralized, accurate and prepared for such analysis. Once the data is ready, spend analysis – which is typically a two-phase process – can be started.

Phase 1: Prepare data for spend analysis

Procurement organizations have historically used highly manual means, such as crunching data in spreadsheets, to prepare data for spend analysis. But due to the sheer volume of purchasing transaction data, they could perform this step manually only for limited categories (representing about 5% to 20% of the company spending) and that, too, with limited accuracy. This made it hard to extract much efficiency out of the sourcing process.

In addition, since additional purchasing transaction data is created every quarter, such analysis must be done on an ongoing basis. Manual methods make it even harder to prepare this data on an ongoing basis. Technology such as SAP BusinessObjects Spend Performance Management has allowed organizations to easily prepare the data for all commodities in a short time with a high accuracy rate. SAP BusinessObjects Spend Performance Management consists of the following steps:

  1. Extracting information from source systems: Spend data often exists as individual transactions in various source databases such as ERP systems and eProcurement applications; travel and expense data from travel management companies; and procurement data from purchasing card providers such as banks. This information must be extracted from the various sources and aggregated. In some of Bristlecone's spend analysis engagements, we have aggregated data from more than a dozen systems, including multiple instances of the same ERP system. This is enabled by an easily deployable starter kit (included) that addresses both SAP and non-SAP source systems.
  2. Normalization and enrichment: Extracting spend transaction data into a common location isn't enough. The data must then be cleansed, normalized, classified and enriched.
  3. Cleansing and normalization ensure that similar supplier names (such as HP or Hewlett-Packard or HP Systems) are recognized and converted to a common supplier name. Also, it ensures that the subsidiary organizations roll up to the right parent organizations.
  4. Each item should be mapped to a proprietary classification taxonomy that improves on standards-based classification taxonomies such as United Nations Standard Products and Services Code (UNSPSC). Once an item is mapped to an industry standard code, it is easy to aggregate spend information for similar components from two different suppliers or even roll up that information for a product family.

    Note: During this step, other supplier-related information such as parent company name, revenues, credit rating, Standard Industry Code and diversity status can be added. This type of data enrichment allows for deeper and richer spend analysis, for example, by aggregating total spend with a supplier, including all its subsidiaries, for better negotiating leverage.

Bristlecone has completed this process in about three months for organizations with about 10 source systems, more than 2,500 suppliers and more than 1 million transactions.

Phase 2: Spend analysis and identifying opportunities

Once data has been cleansed, normalized, classified and enriched, it can be analyzed to identify "low-hanging" cost-saving opportunities. For example, spend analysis may show that the company is buying the same item from more than one supplier. Seeing that, the company could consolidate spend for that item under one supplier. The company might also find that certain items are being purchased through suppliers that are not preferred or not on contract, or merely at higher prices, and could then install processes to ensure compliance with purchasing policies and eliminate maverick spend.

Other cost-saving opportunities identified through spend analysis might include reducing item proliferation and ensuring compliance with contracts to reduce purchase price variance. The data could also be used to identify and analyze risk in the supply chain by understanding which commodities could become single-sourced or end up with fewer suppliers relative to company policies if the supplier base is consolidated. In certain scenarios, by understanding the relationships of tier 1 suppliers to their component suppliers, one can influence the spend and identify the systemic risk in the supplier community within one's ecosystem.

About the author: Naresh Hingorani is the Integrated Sourcing and Procurement Practice Area Leader at Bristlecone, a global supply chain consulting firm. Strategic purchasing initiatives require deep insights into current spend, which requires a combination of technology, a proven framework and subject-matter expertise and must be conducted as an ongoing program to ensure that savings results "stick." Bristlecone brings all three elements together in a project and managed service program and leverages SAP's Data Standardization and Enrichment as well as Spend Performance Management technology to help clients reduce their purchasing spend.

Dig Deeper on SAP SCM

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.