Upload
open-analytics
View
306
Download
0
Tags:
Embed Size (px)
Citation preview
On the “Moneyball” – Building the Team, Product, and Service to Rival
Breaking Biases
http://www.youtube.com/watch?v=yGf6LNWY9AI
Evolution of an Idea
The White House Harvard Technology Healthcare
• Reduce Turnover 45% - 75%
• Improve Quality of Care and Patient Experience
1996 2001 20142010
Big Data and Preferences
How to Get a Job at GoogleFEB. 22, 2014
Continue reading the main story
Big Data and Software Development
Big Data and Hiring in Hospitals
By year, all facilities combined (non-RNs only) Year Number of employees Numer of terminations Turnover rate Cost to replace
2009 5217 1121 21.49% $ 13,452,000.00
2010 4616 1020 22.10% $ 12,240,000.00
2011 4557 1040 22.82% $ 12,480,000.00
2012 4497 1103 24.53% $ 13,236,000.00
By year, all facilities combined (RNs only) Year Number of employees Numer of terminations Turnover rate Cost to replace
2009 2326 533 22.91% $ 15,457,000.00
2010 2210 468 21.18% $ 13,572,000.00
2011 2102 389 18.51% $ 11,281,000.00
2012 2070 497 24.01% $ 14,413,000.00
Note: The Joint Commission estimates the cost to replace for RNs to range between $46,000 and $64,000. For purposes of this analysis, we have estimated conservatively at $29,000. Along the same lines, our conservative estimate for the average cost to replace non-RN positions is $12,000. These costs to replace include lost productivity due to ramp-up time for a replacement hire as well as recruiting costs and hiring costs.
Total 2012 Estimated Cost of Turnover: $27,649,000
2
Sample Success
60 days 90 days 180 days 360 days
Rate Rate Rate Rate
Pegged Recommended Hires 5.7% 8.6% 8.6% 25.7%
Pre-Pegged Period 25.0% 27.3% 35.2% 48.9%
Pegged Improvement 77.1% 68.6% 75.7% 47.4%
Turnover Reduction Achieved
Keys to Big Data
• Set Clear Objectives• Prediction?• Segmentation?• What is the Business Goal?
• Define success and figure out how to measure it accurately
• Understand the difference between intended and unintended consequences
Keys to Big Data (Cont’d)
• Carefully map out the work flow for data collection
• Be cautious in presenting results – assume your conclusions are wrong until you know they aren’t
• Define success and figure out how to measure it accurately!!!!
Q&A
Questions?