Upgrade to Pro — share decks privately, control downloads, hide ads and more …

A Study of the Organizational Dynamics of Software Teams

A Study of the Organizational Dynamics of Software Teams

This is the slides for a talk given at ICSE 2018

8e81db9f29d2543ada5fac546f99e023?s=128

Michael Hilton

June 03, 2018
Tweet

Transcript

  1. THE ORGANIZATIONAL DYNAMICS OF SOFTWARE ENGINEERS Michael Hilton Carnegie Mellon

    University @michaelhilton Andrew Begel Microsoft Research @abegel
  2. BACKGROUND Software engineers change teams frequently at Microsoft. Why do

    they do this? How does it happen? What are the costs of changing teams? ­ Loss of tacit knowledge ­ Loss of relationships that support collaboration ­ Team newcomers require mentoring ­ Monetary costs can be as high as 200% of employee’s annual salary.
  3. UNFOLDING MODEL OF EMPLOYEE TURNOVER Organizational Science has studied employee

    turnover for over 100 years! We consider decisions to leave and to join separately. Job dissatisfaction Shock Pre-existing plan No pre-existing plan Unsolicited job offer
  4. Corporate reorganization Change in product focus New manager Cancelled project

    Moving due to geography Shipping deadline Receiving a promotion Receiving an unsolicited job offer Changing job role Layoffs are coming WHAT EVENTS CAUSE ENGINEERS TO CONSIDER SWITCHING TEAMS?
  5. WE FOCUS ON INTERNAL TURNOVER AT MICROSOFT

  6. STUDY METHODOLOGY Interviews (for depth) Surveys (for generalization) Human Resources

    Data Analysis (for triangulation)
  7. STUDY METHODOLOGY 8 semi-structured interviews with Microsoft engineers Surveys (for

    generalization) Human Resources Data Analysis (for triangulation) Varied by division and tenure at the company
  8. STUDY METHODOLOGY 8 semi-structured interviews with Microsoft engineers 374 survey

    respondents with Microsoft engineers who had changed teams at least once at the company. Human Resources Data Analysis (for triangulation) 7% random sample. 20% response rate. 80% male, 18% female, 2% other/decline to state 80% at corporate HQ. 20% at satellite offices. Average tenure: 8.4 ± 5.5 years Average industry experience: 12.2 ± 6.1 years. Varied by division and tenure at the company
  9. STUDY METHODOLOGY 8 semi-structured interviews with Microsoft engineers 374 survey

    respondents with Microsoft engineers who had changed teams at least once at the company. 90,000+ Microsoft engineer employee data records Average tenure 8.9 ± 6.0 years (similar to survey) 7% random sample. 20% response rate. 80% male, 18% female, 2% other/decline to state 80% at corporate HQ. 20% at satellite offices. Average tenure: 8.4 ± 5.5 years Average industry experience: 12.2 ± 6.1 years. Varied by division and tenure at the company
  10. SURVEY DEMOGRAPHICS 0% 5% 10% 15% 20% 25% 30% 35%

    Just myself 2-3 4-6 7-10 11-15 16-20 >20 Team Size 0% 5% 10% 15% 20% 25% 0 1 2 3 4 5 6 Years under Manager (avg)
  11. WHY DO ENGINEERS WANT TO LEAVE THEIR TEAMS? Reasons grouped

    by clustering analysis Percent Change is coming (technology, charter, re-org, turnover) 52.6% Seeking new challenges or location (role, location, challenges) 39.0% Dissatisfaction with manager (priorities, goals, person, actions) 31.6% The grass is always greener on the other side (novelty, escape) 12.3% Not a good fit (bored, no need for my skills) 5.3% Poor team dynamics (dysfunctional, no career growth) 4.4%
  12. WHY DO ENGINEERS CHOOSE TO JOIN PARTICULAR TEAMS? Reasons grouped

    by clustering analysis Percent Liked new team and/or technology (exciting, manager) 85.8% Coworker asked me to join (new team, old team) 37.8% Joined for better opportunities (location, domain, lack of other options) 24.5% Followed my manager (former or current) 14.6%
  13. TAKEAWAYS Seek new challenges 1 Exercise agency 2 Follow people

    3 Change happens 4
  14. TAKEAWAY: ENGINEERS SEEK NEW CHALLENGES Junior employees more often leave

    teams to find new challenges. Senior employees more often leave teams because they don’t like their manager. Project recommenders may inadvertently pigeon- hole engineers based on their past work instead of their future goals.
  15. TAKEAWAY: EXERCISE AGENCY Employees who voluntarily leave or join teams

    are more satisfied with the result, yet take longer to onboard than those who change teams involuntarily. 72% 75% 24% 22% 4% 3% Leave Join Turnover Consent Voluntary Involuntary It's Complicated
  16. MY MOST RECENT TEAM CHANGE WAS GOOD FOR ME 2%

    5% 5% 2% 4% 9% 8% 4% 9% 17% 29% 11% 34% 30% 26% 39% 51% 39% 32% 45% TECHNICALLY SOCIALLY POLITICALLY OVERALL Disagree Mostly Disagree Neutral Mostly Agree Agree
  17. PERCEIVED ONBOARDING TIME AFTER A MOVE 0% 5% 10% 15%

    20% 25% 30% 5-6 months 3-4 months 1-2 months 3-4 weeks 1-2 weeks Up to 1 week No time at all
  18. PEOPLE LIKE A NEW CHALLENGE, EVEN IF IT TAKES LONGER

    TO ONBOARD Join reasons grouped by clustering analysis Percent Liked new team and/or technology (exciting, manager) 85.8% Coworker asked me to join (new team, old team) 37.8% Joined for better opportunities (location, domain, lack of other options) 24.5% Followed my manager (former or current) 14.6% Leave reasons grouped by clustering analysis Percent Change is coming (technology, charter, re-org, turnover) 52.6% Seeking new challenges or location (role, location, challenges) 39.0%
  19. PEOPLE WANT CONTROL OVER WHEN THEY MOVE. OFTEN, THEY’LL MOVE

    JUST BEFORE A RE-ORG. Leave reasons grouped by clustering analysis Percent Change is coming (technology, charter, re-org, turnover) 52.6%
  20. TAKEAWAY: FOLLOWING PEOPLE Employees who found their next teams by

    word of mouth were more satisfied with the result. 0% 10% 20% 30% 40% 50% 60% 70% Word of mouth Internal career web site Manager found position for me How Engineers Found New Teams to Voluntarily Join
  21. PEOPLE FIND OUT ABOUT NEW POSITIONS FROM THEIR SOCIAL NETWORK

    Social networking is especially important when employees need to move further away organizationally. People follow their friends, co-workers, and (good) managers to new teams. Join reasons grouped by clustering analysis Percent Liked new team and/or technology (exciting, manager) 85.8% Coworker asked me to join (new team, old team) 37.8% Joined for better opportunities (location, domain, lack of other options) 24.5% Followed my manager (former or current) 14.6%
  22. CHANGE HAPPENS Engineers move often. This improves skills and diversity

    of the overall population. Managers should prepare for employees to leave teams and encourage them to “graduate.” 0% 5% 10% 15% 20% 25% 30% 35% 1 2 3 4 5 6 7 8 9 >10 Number of team changes at Microsoft (avg)
  23. ORGANIZATIONAL LEVELS TRAVERSED WHEN CHANGING TEAMS Employees who leave because

    they don’t like their manager are more satisfied with their moves and move further away organizationally. 0 20 40 60 80 4 5 6 7 8 9 10 11 12 13
  24. TAKEAWAY: MAKE CHANGE HAPPEN Managers have a big impact on

    their employees’ everyday experience. Employees that don’t like their managers should move immediately. Don’t wait. Who you work with is more important than what you work on.
  25. TAKEAWAYS 1. Seeking New Challenges ­ People like a new

    challenge, even if it costs more onboarding time. ­ Recommender systems should avoid pigeon- holing people based on their past work. 2. Exercising Agency ­ People want to control when they change teams. ­ This is especially important during re-orgs. 3. Following people ­ People follow their friends, co-workers, and (good) managers to new teams. ­ People find out about new positions from their social network. 4. Change happens ­ Managers: embrace your employees’ ”graduation.” ­ Employees: Managers have a big impact on your everyday experience – if you feel a change is necessary, don’t wait. Want to learn more? Read our paper: http://andrewbegel.com/papers/orgdynamics.pdf