Chmyr Olha. WHY TEAMS ALWAYS F*CK UP WITH THEIR OWN ESTIMATES
Download as PPTX, PDF0 likes51 views
The document discusses common pitfalls in project estimation, highlighting the importance of a structured approach and awareness of emotional traps. It emphasizes defining subtasks, accounting for uncertainties, and avoiding overly optimistic estimates due to the planning fallacy. Strategies for improving estimation accuracy include gathering multiple estimates, breaking down tasks, and considering both productive and non-productive time.
Chmyr Olha. WHY TEAMS ALWAYS F*CK UP WITH THEIR OWN ESTIMATES
- 1. WHY TEAMS ALWAYS F*CK UP WITH THEIR OWN ESTIMATES OLHA CHMYR, PROJECT MANAGER, SPD-UKRAINE (PitchBook)
- 2. ● 5 years of Project Management ● worked for startups, small/medium companies, and FMCG corporations ● beautiful creatures and coffee lover ● Project Manager at SPD-Ukraine (PitchBook) ABOUT ME:
- 3. WHO I AM
- 4. AGENDA ● Theory of estimating ● Emotional traps in the estimation process
- 6. GIVE ME YOUR ESTIMATE! #NOESTIMATES
- 7. TIME ESTIMATES PROCESS DEFINE THE TARGET DIVE INTO DETAILS ● split into obvious subtasks; ● do not get brought away with detailed detailing. 1 2
- 8. DEFINE RISKY SUBTASKS ESTIMATE TIME FOR DETAILED TASK ● include non-productive hours; ● include time for code review, integration and other dev activities. 3 4 TIME ESTIMATES PROCESS
- 9. LET’S TALK MORE ABOUT NUMBERS AND UNKNOWNS ● Ask for at least 2 estimates – min and max (it is better to have even three – min, max and most likely); ● More unknowns, less confidence
- 10. UNKNOWNS AND UNKNOWN UNKNOWNS ● Do we understand the spec well? ● How many negative cases are covered in the spec? ● How will we test the system? ● How will the system behave at the moment of failure?
- 11. ● Can the architecture be scaled? ● What are the risks associated with this solution and to the choices about to be made? ● How will we monitor the system? ● Is it possible to complete a task more straightforwardly while still accomplishing a business task? UNKNOWNS AND UNKNOWN UNKNOWNS
- 12. YOU CAN BREAK THE ANALYSIS PARALYSIS ● Can it last for a half a year? ● Can it last for a week? ● Can it last for a month? ● Can it last for a day?
- 13. ESTIMATE THE QUALITY OF TIME ESTIMATES K = T(p)/T(o)
- 14. HOW TO AVOID MISTAKES ● do not get brought away with detailed detailing; ● include risky tasks and increase their estimate; ● provide more specificity in the task; ● split into the subtasks until everyone feels well comfortable about them;
- 15. HOW TO AVOID MISTAKES ● calculate individual coefficient of error estimation; ● include the blocking dependencies between the tasks of different devs. ● investigate tasks which are hard to estimate; ● include non-productive time expenditures;
- 16. PLANNING FALLACY Planning fallacy – cognitive distortion of planning
- 17. Roger BuehlerAmos Tversky and Daniel Kahneman PLANNING FALLACY
- 18. 70% did not meet the deadline 80% were sure about the deadline THE MAIN REASONS FOR MISTAKES
- 19. When we thought over something for ourselves, it seems to us that this will happen more likely. THE MAIN REASONS FOR MISTAKES
- 20. Detailed plan leads to the more optimistic estimate THE MAIN REASONS FOR MISTAKES
- 21. THE MAIN REASONS FOR MISTAKES Hot/Cold empathy gap
- 22. THE REASON OF PLANNING FALLACY Focalism
- 23. HOW TO LIVE ON
- 24. Willis Everett McNelly HOW TO LIVE ON
- 25. Short and unrealistic terms Additional time HOW TO LIVE ON
- 26. HOW TO LIVE ON External estimate
- 27. If everything is as usual, how much time do you need to finish the task How did the events go, which of them usually lead to the timing shift? HOW TO LIVE ON
- 28. PLEASE, DON’T BURN OUT