Building Data Champions Combining Internal and External Data

Building Data Champions Combining Internal and External Data to Find Jobs and Fill Them Philip S. O’Donnell VP Data Science and Analytics – philip. odonnell@adeccogroup. com EMSI Conference 2019 - Sept. 17, 2019, 1: 30 p. m. - North Cape Bay

What are We Talking About – Introduction – Real World Examples – Getting Started with Data

Introduction to Me and My Work

Who Is Adecco Group? The largest global staffing company you probably haven’t heard of…

My Journey Management Consulting Balanced Scorecard Collaborative Manager Strategic Planning Redbox-Coinstar Senior Manager Analytics T-Mobile – Handset Financing and Insurance Director Data Science Lee Hecht Harrison - Global VP Data Science & Analytics Adecco Group – Americas Lead a team of Data Scientists, Data Engineers, Business Analysts and Visualization Specialists

What is Data Science? Bigger Traditional Analytics but Learning + Storytelling Most Needs are not Complex (AI) 6

We Manufacture Data • Data is Both Raw Material and Finished Goods • Build or Buy • Adoption and Quality are #1 Goals • “[Noun]Champion” by Data. Champions 7

External Data Gives Context Transactions Local Reality Surveys Broad Estimates Internal External 8

Caveats Adecco is EMSI API-Driven Our team does not build AIs for Recruiting or Search and Match 9

Real World Examples 10

High Level Data Landscape Branches Sources Prospects Metrics Time Pay Candidates Clients Quality Cost Submissio Job Orders ns Placement s 11

Goal: Internal + External Data > Either Fill More Jobs Pay Higher Wages Drive Data Quality Avoid Bias Internal External

Goal: Balance Hiring Speed and Cost Time to Fill by Compensation Talent Supply by Compensation Internal External 13

Goal: Hold Ourselves Accountable Time to Fill by Job Title Internal Local Employment By Job Title External 14

Goal: Keep People Working Longer Temp Turnover in Market Internal External 15

Goal: Help Companies Forecast Better Job Order Mix in Market Internal Online Job Posting Mix in Market External 16

Goal: Guide Reskilling Pay Rates by Skills Internal Skill Supply by Compensation External 17

Goal: Market Rate or Higher Average Temp Wages Paid Average Wages Paid in Market Internal External 18

Goal: Reflect the Reality Of Wages Relative Temp Wages Paid In Market Cost of Living Index Internal External 19

Goal: Set Compensation Expectations Time to Place Job Supply by Compensation Internal External 20

Goal: Standard Reference Hierarchies Branches / Billing Codes MSAs / SOC Codes Internal External 21

Goal: Get More Accurate Data SOC Codes Demand Supply Data Internal External 22

Goal: Ensure Accurate Data Products Internal Labor External Labor Market Estimates Internal External 23

Moving Forward With Data 24

Getting Started With Data Don’t Hire Data Scientists First Start Simple - Add Complexity Later Serve People Who Value You Consistent Quality Matters 25

What Data Don’t We Use? Don’t: Model using Protected Classes Do: Validate Models to Avoid Bias 26

Thank you for your attention 27