Human Resources

1. What are my staffing requirements in the future based on my business plan?

Where will we need to deploy/redeploy our resources around the world?

Human Resources Simulation: Large organizations need to manage their human resource populations to remain effective and competitive.  A department HR requirement is a function of the service provided and the employee productivity; both change over time.  CMU-RC used a Systems Dynamics software tool to simulate the future of an enterprise incorporation the above factors as well as the aging of the employee base and the advancement of employees through ‘the ranks’ based on historical or proposed promotion rates.

Finance

2. What cash flow management opportunities exist for us to more effectively manage spending?

Spend Analysis: CMU-RC utilized advanced data mining and predictive techniques on a global company’s US accounts payables data to highlight spend patterns useful for cash flow management opportunities.

Quality

3. What is the financial risk for future time periods as a result of product claims? 

Where (geographically) will the magnitude of these claims occur?

Predictive Warranty Modeling:  CMU-RC has developed a set of models designed to help a client company understand likely future warranty claims against its products.  The failures are related to environmental factors (weather and soils), but claim rates also vary by local communication channels (e.g. neighborhood conversations, newspaper articles, etc.).  A series of neural network models combined sales and claims data with weather, soils and demographics data to produce a failure prediction model; these data were also combined to predict the local life-cycle of the claims activity; and the claim dates were combined with the previous data to predict the sequencing of regional life-cycles around the nation.

SUGI 29 Presentation (PDF)
 

4. What is really going on in unclassified text fields called “other”?

Early Warning Detection from Call Center and free text data: CMU-RC conducted an advanced text mining project of call center/trouble ticket data to identify potential improvement for overall service level quality. We have also applied advanced text mining, clustering, and sequencing of highly unstructured comment fields on a product to gain additional insight into product performance.

SUGI 30 Presentation (PDF)
 

R&D

5. How can new Intellectual Property be evaluated by a new patent attorney?

Web application: A small company with external I. P. support needed rapid evaluation of uniqueness some new concepts.  CMU-RC produced an application to organize potentially relevant patent numbers (and pre-grant applications) obtained from searching the USPTO.  Components provide maps of claims; clustering of abstracts, claims, and descriptions based on words used; linked extracts of citations, inventors and assignees.

Sales

6. How can we prioritize expenditures on loyalty improvement based on their impact to future profitability?

Customer Relationship Management: CMU-RC has conducted a project for a fortune 100 company that utilizes customer loyalty data and retention attraction results to predict their impact on bottom line profitability.

7. What opportunities to up-sell and cross-sell exist within our current customer base? What can we learn about our current customers that helps target new customers?

Advanced Market Segmentation: CMU-RC applied advanced data mining and neural networking technology on a large multi national corporation’s internal data to score similar clients for potential sales and to target or up-sell services to existing clients.

Marketing & Business Unit

8. What initiatives are our competitors undertaking that will impact their market position?

Competitive Analysis: CMU-RC created a set of competitive predictive models to examine customer attraction patterns and to measure the impact of marketing efforts. We also have developed solutions that can monitor competitor websites, or identify new concepts that appear on WWW or text based publications.

9. How can data mining be understood and used by people without extensive experience?

Visualization: The client requested that visual data mining tools be explained with real business data and interesting relationships demonstrated using flash presentations for introduction of data mining to managers.

10. How can we use external data to broaden the scope of our analysis of internal data?
What might we infer about  new customers, new geographies, or new competitive environments?

 

Inclusion of External Datasets: CMU-RC has also been very successful locating and working with data from the US and around the world. We have helped corporations combine external data sources, such as Census, satellite, geographic/map, third party purchased data, foreign economic statistics, etc., with their own internal corporate information and transaction data, like sales or claims data, to expand predictive power or model insight.

11. What impact will a rumored competitor moving into our market have?

Site Selection: A client was presented with the rumor that a competitor would locate a facility in the region.  CMU-RC developed a gravity model based on ten years of competitive activity (i.e., other competitive sites being established) From this, the client was able to put a boundary on the negative impact to various confidence levels.

12. What impact will price changes have on the current volumes in various markets?

What effects does manufacturing efficiency have on profitability?

Economic Profit Optimization: A business line manager was confronted with shrinking economic profit and wants to understand what factors are most significant in the decline.  Factors considered were:

  ·  Rising raw material costs related to the cost of oil

  ·  Manufacturing capability (equipment breakdowns affected ability to produce at times)

  ·  Marketplace dynamics (the total market consisted of nine end-user markets with different growth rates and price elasticities)

CMU-RC developed a model for each end-user market covering a four year period by quarter.  Externally obtained projections of oil costs were incorporated to project the economic profit for the next two years as a function of price.  This was used to develop a pricing strategy by end-user market to optimize the economic profit.

Operations & Supply Chain

13. Will e-commerce selling cannibalize existing value added business offering?

Supply Chain Analysis: CMU-RC was engaged to help evaluate a new business model for a global corporation to understand the financial impact and to determine any potential for cannibalization of down stream products when the company launched a new service higher up the supply chain.

14. How do we use knowledge about our business drivers to better plan for future resource utilization?

Predictive Model for Healthcare:  CMU-RC created a predictive model that utilizes outpatient and emergency department activity data as a mechanism to predict the near term of inpatient volumes and doctor specialty mix to serve as an early warning device that can recognize significant increases or decreases in patient volumes.

15. How can we negotiate better rail-rate agreements?
What information should our negotiators have to get optimal rates?
What shipping costs should we expect to incur to deliver to new customers?

Supply Chain Logistics:  A client uses a lot of rail transportation, which has been increasing in cost as congestion in major cities increases.  The client was looking for ways to negotiate better rate contracts.  Rail shipment detail for the last three years was used to build a predictive model based on route, carrier, competitive position, and escalation clauses.  The existing contracts were scored against the model and those that were indicated to be overpriced were given to the negotiators with an indication of what in the contract seemed too high.