OpenExplore™
Overview
OpenExplore™ is a 2-week effort that provides an analytic foundation for performance measurement by investigating your data -- often sourced from a fact table in a star schema data mart. During the first days of the engagement, OpenBI works closely with the business team to identify metrics and dimensions that are important to company performance. Using the R and Weka open source packages, we apply statistical and machine learning techniques with data visualization to characterize key performance attributes. After sharing initial findings, we iterate with the business team to produce the final analysis, at which time we craft a report and present our findings to management. The output of an OpenExplore™ is often the input that directs the design of a subsequent management dashboard implementation. At project conclusion, we deliver pertinent statistics/graphics in html/pdf formats, along with a report interpreting results and making recommendations for subsequent steps.
|
Examples of analytical domains include:
- Customer Segmentation
- Profitability Factors
- Marketing Mix
- Purchase Propensity
- Retention Factors
- Underwriting Risks
|
|
|
Deliverables
The main OpenExplore™ deliverables are a series of HTML pages and pdf files that detail study findings, as well as a management report that catalogs the performance measures, business dimensions, and go-forward recommendations.
|
|
| Management Report |
- Details high-level performance metrics, time dimensions, business dimensions, and characterizations/discriminations mined from the data
- Summary of findings of mining and graphical analyses
- Recommendations for go-forward performance measurement
- Results of modeling/mining exercises, as appropriate
|
| Reports & Graphs |
- HTML and pdf files detailing tabular and graphical statistics produced from the analyses
- Tabular and graphical time series performance measures characterized/conditioned on important business dimensions
|
| Data |
- Client provided dataset
- Up to 12 core metrics
- Up to 8 dimensions
- Up to 100MB of data
|
|
|