Why is Typical Vendor Spend Data not Useful?
Typical spend data is not useful to provide business insight and drive informed decision making.
In many large global operations, spend data originates from different people, in different organizations, in different locations, even using different vernacular or languages, and all with different requirements. Further, it is typically coded from a financial perspective.
While General Ledger (GL) and cost center are helpful for the finance organization, they do not provide the views needed to drive procurement decisions. For example, in an accounting system, an item might be classified as a three-year depreciable asset, but not as a laptop computer. This is not the most useful view for purchasing professionals. They need information that is organized—by commodity, by vendor, by business unit, by geography—to show spending patterns. They need to understand how the company purchases, not how the accounting department records expenses.
Once organizations achieve spend visibility by vendor “family” and commodity, the data is sometimes not granular enough to provide business insight and drive informed decision-making. Detailed product information and attributes are trapped inside cryptic line item descriptions and need to be structured and mapped against a more granular commodity structure.
In pursuit of company-wide spend visibility, enterprises are confronted with three key problems
• The data available is poor and ill-suited to driving procurement decisions
• Enriching the quality of the available data has been challenging and costly
• Business dynamics are constantly changing
Why is information quality typically poor?
The first barrier to good information quality is that spend data is dispersed. It is scattered across multiple, disconnected accounting systems (example, Accounts Payable (AP), Enterprise Resource Planning (ERP), corporate purchasing cards, eProcurement systems, and electronic funds transfers) and detailed product information exists across a variety of formats (example, XML, HTML, PDF, Microsoft®Word, text files, spreadsheets, databases). However, even once the data is aggregated, the second barrier is that the data is often of poor quality: unstructured, incomplete, inaccurate, or not at the right level of detail.
There are multiple drivers:
• Data entry errors.
Spending data is often recorded inconsistently with errors, duplicates, and misspellings, leaving a large amount of unclassified, “other” spend.
• Nomenclature variability.
The usage of language, nomenclature, and terminology varies by different organizations and groups responsible for the data, making it difficult to achieve a consistent view.
• Duplicate vendor codes.
Typically, vendor names are spelled various ways and there is no link between parent corporations and subsidiaries making it difficult to get a picture of total spend with any one vendor. Even in a single AP system, for example, individual suppliers might have more than one unique code assigned to them, making it difficult to compile total spending by supplier.
All of this leads to poor vendor management.