Data Preparation
Provide access to normalized, rationalized, and simplified data (e.g. with all frequencies expressed in kHz as real numbers) for analytical use.
Improvement Planning
Practices-Outcomes-Metrics (POM)
Representative POMs are described for Data Preparation at each level of maturity.
- 2Basic
- Practice
- Audit the data to identify errors, gaps, anomalies, and inaccuracies that need to be corrected.
- Outcome
- There is clarity on the quality of data available and what needs to be done to improve its collection and to correct it.
- Metrics
- Number of data collection problems identified per data set.
- Number of data quality issues identified per data set.
- 3Intermediate
- Practices
- Put in place a workflow (sequence of data preparation steps) to address and correct data errors, anomalies, and inaccuracies.
- Validate the effectiveness and efficiency of the data preparation process and adjust it as required.
- Outcome
- Processes are in place to resolve any issues that have been identified in the collection of data and its suitability for analytics.
- Metric
- Number of workflows and processes in place to ensure the accuracy and quality of data.
- 4Advanced
- Practice
- Build on the effectiveness of the data preparation process by using advanced techniques and tools to transform the data for analysis.
- Outcome
- Advanced and automated data preparation techniques including warehousing are used — to ensure a very high level of data quality and the appropriate transformation prior to analysis.
- Metrics
- Percentage automation of data preparation.
- Percentage of raw data that is transformed for analysis.
- 5Optimized
- Practice
- Use research and feedback from data analytics to continually improve how data is prepared.
- Outcome
- Research and feedback identify new approaches for data preparation.
- Metrics
- Number of improvements implemented per data set per year.
- Percentage cost savings delivered by data preparation improvements.