Where is the data?Rarely is spend data all in one system. On top of that, important spend data may exist on legacy and external systems, which only makes getting the data more difficult. To build an accurate picture of spend, all relevant data will be needed. Locating the relevant data sources is just the first task. Data will also need to be extracted into a usable format, so there may be a requirement for IT to get involved.
Quality of the spend dataOnce you’ve got hold of the spend data and are confident that it covers the vast majority of spend, how do you make sense of it? It may be in different languages, currencies and use different names or codes for similar items. Data entered manually can contain typos and inconsistencies. Can you identify data for a specific calendar year, or does some data refer to a company’s financial year? Is the data detailed and consistent enough to be analysed? It’s not straightforward.
Cleansing the dataThe tasks to extract, cleanse, classify and normalise spend data should not be underestimated. Traditional, rules-based approaches to cleansing and classifying spend data can be time-consuming and may introduce inconsistencies if carried out manually. Automated approaches may be quicker but exceptions will undermine completeness, and therefore accuracy.
Use of leading-edge technologyMarket leaders in spend analysis use artificial intelligence and machine learning software to help detect and correct inaccurate or incomplete spend data and format errors. This software uses “fuzzy” matching to compare spend data against its own database to ensure that similar item descriptions, company names and other information are identified and corrected automatically. The categorised spend data is presented in customisable output for immediate business use, not just a generic presentation of spend data.
Spend analysisOnce the data is cleansed, normalised and available in a customised spend data classification and reporting structure, there should be clear visibility of total aggregated spend on goods and service across all parts of an organisation. Now the analysis can begin, which should lead to:
- Ability to effectively identify and act upon patterns in spend data to uncover additional strategic sourcing opportunities
- Increased ability to identify cost savings or supplier consolidation opportunities
- Improved forecasting and better informed decision-making processes through reliable and accurate spend data input
- Access to benchmark data and insights
- Potential for operational efficiencies and improved supplier management
- Greater collaboration within procurement teams