Powerhouse Successes
Powerhouse Incorporated is a data mining vendor with whom we've established a partnership. Tom Breur is certified as a Powerhouse Consultant. Powerhouse is a data mining tool based on an innovative use of Information Theory. We feel that this tool is the most significant innovation in Artificial Intelligence since the invention of the Neural Network.
But where Neural Networks require extensive preprocessing and are slow to train, Powerhouse eliminates the need to preprocess and builds models blazingly fast. It has an extremely easy-to-use interface making it very, very easy to learn.
Project cycle times from raw data to prediction (or exploration) are slashed, making miners much more productive. We feel that using Powerhouse we're at least 5 (!) times as productive as any other tool we've ever used before (most major vendors).
The Powerhouse Company Brochure
| Market: Mutual Funds Application: Sales Rep Attrition |
|
|---|---|
| Need/Problem | Marketing group's entire yearly bonus was at risk unless they could identify which of their more than 300,000 sales reps were most likely to churn. |
| Before Powerhouse | Software vendor delivered an attrition model which proved to have no predictive power. Lost 9 months. |
| After Powerhouse | Identified at risk reps accounting for $4.4 billion in decreased sales with more than 93% accuracy. Savings 2004: > $250 million. |
| Market: Mutual Funds Application: Revenue Growth |
|
| Need/Problem | The company wanted to identify sales reps with maximum revenue growth potential. |
| Before Powerhouse | No analysis was used. |
| After Powerhouse | Identified sales reps who would account for growth in the target period. Revenue Growth: $642 million. |
| Market: Consumer Products Application: Competitive Threat, Brand Loyalty |
|
| Need/Problem | The company needed to increase consumption in existing stores and to expand into new markets without costly missteps or poor selections. |
| Before Powerhouse | No analytics to identify trends in consumptions by market or un-served market segments. |
| After Powerhouse | New markets, and product segments identified, new products developed. Year 1: > $2 million in revenue. |
| Market: National Media Company Application: Telemarketing |
|
| Need/Problem | A wealth of data and no cost effective method of selecting variables. |
| Before Powerhouse | Initial data set: 250 million records, 500 variables. No way to analyze. |
| After Powerhouse | Selected best 8 variables in 45 minutes. Saved: $2 million in data costs. |
| Market: Wireless Carrier Application: Reducing Churn |
|
| Need/Problem | Needed to identify key intervention point and activity in customer's churn cycle. |
| Before Powerhouse | Lack of analytics relied on guess work. Provided no visibility into churn. |
| After Powerhouse | Reduced churn by 0.3%, translates to 7,500 customers worth. $3.75 million in revenue. |