Successful spend analysis needs a sustained application of resources; a commitment to refreshability; the sponsorship of one or more high-level executives such as the CEO, chief financial officer (CFO) or chief purchasing officer (CPO): and the participation of all relevant stakeholders. It also necessitates the adoption of the proper systems and processes to repeat the expenditure review on a regular basis, making it possible to uncover monthly, quarterly or year-over-year differences that can spark new cost-saving initiatives.
Whether using a license-based model requiring in-house installation of application hardware and software or an external solution hosted by an application service provider (ASP), implementation of a comprehensive spend analysis program involves five components. These include:
* Spend data capture software to gather and consolidate spending data from across A/P, PO and p-card systems, data business units and geographies
* Normalization software to eliminate discrepancies among supplier names, addresses, currencies, dates and other data compiled from multiple sources
* Classification software to code all products and services according to a uniform scheme to ensure commonality for the same or similar items
* Analytical and reporting tools typically interactive OLAP (OnLine Analytical Processing) systems with flexible reporting outputs
* A data warehouse to act as a central repository for data collected from all relevant enterprise systems
Spend analysis can be used by companies that are doing paper-based purchasing as well as those that have embraced e-procurement, but an enterprise using paper-based systems must also convert its paper data into electronic format. This is an extra step, but it offers an incentive to make the inevitable transition to electronic purchasing.
Once the infrastructure is in place, managers must implement proper procedures to ensure consistency across the enterprise, as well as comprehensive results. Processes must be standardized across all business units; the system must be configured to look at both past spending patterns and projected demand, typically one to three years in the future; software must be customized to generate the precise metrics the organization needs; and so on. Then the system can analyze corporate spend along dimensions such as commodity, class, part, part level, parametric attributes, supplier and division.
The Importance of Classification
The success of a spend analysis effort will be closely tied to the classification scheme employed. Granular, hierarchical classification of data is essential, not only to ensure successful standardization across the procurement environment, but also to aid search precision and permit aggregation and drill-down to any level of analysis.
With a common set of product identifiers that are part of a hierarchical taxonomy, individual purchases can be identified by highly specific descriptors ("safety glasses") or rolled up into more generic categories ("shop supplies" or "industrial supplies"). This allows users to evaluate expenditures at exactly the level most useful to their business or their purpose, including the identification of families of products that can be combined into contractible groups to negotiate a single source of supply and volume discounts.
The classification system that appears to be emerging as the universal standard is the United Nations Standard Products and Services Code (UNSPSC), the first global coding convention for products and services. Developed by the United Nations in association with business information provider D&B, the UNSPSC system arranges the entire universe of products and services into more than 13,000 hierarchical categories according to a five-level numbering system. This hierarchical structure and depth of detail make it possible to achieve the level of specificity required to effectively pinpoint a given product set, as well as to create meaningful product groupings.
Under the antenna category, for example, the UNSPSC assigns unique codes to communications, radio, automotive, satellite, aircraft, broadcast, microwave, television and radar antennas, enabling users to zero in on the precise product of interest. All antenna types share the same segment, family and class codes, facilitating search analysis and roll-up. The fifth level of the UNSPSC coding system allows for customization within the organization, tailoring the categorization hierarchy to user requirements.