SlideShare a Scribd company logo
Want a better estimate ?
                     Alexandre Cuva
  Coach Agile, CSM, CSPO, LTM3 HSPTP
                   PMDay Bucharest - 2012
Alexandre Cuva
Email :                    Organizational Coaching
alexandre.cuva@altran.co   (Management 3.0, Scrum)
m                          Team Coaching
Twitter: @cuvaalex         (Scrum, XP, Kanban)

Blog: http://agile-        Technical Coaching
alexcuva.blogspot.com/     (TDD, BDD, C#, Java, Groovy)

Phone: +41 78 715 8309     Agile Training
                           (Management 3.0, Agile, Scrum, XP)
Want better estimation ?
Observation

The complexity is growing fast
Observation

The current standard management system, does not provide
satisfaction to all.
My Actual Mission

Help companies to make this sentence True for them


“Organizations can become learning networks of
individuals creating value, and the role of leaders
should include the stewardship of the living rather
than the management of the machine.”

http://www.stoosnetwork.org
How to have an Accurate Estimation ?
Common Scenario
How long do you
think this project
                      I think in 6
will take ? we need
                      month we
to finish in 3
                      will be done
month.




     Didn’t hear
     me I say it
     should be
     done in 3
     month !!!
Problem of estimation

                   From this picture




                   Estimate the size of
                   my submarine ?
3 points estimation

The 3 point estimate          1. Take the worst
belongs to the time              scenario
management knowledge          2. Take the best scenario
area. It can also be used     3. Take the median
during Cost Estimation.
A three point estimate is a
better estimate, compared     E = (O+M+L)/3
to a single point estimate.
Single point estimate
simply gives you a single
number scenario
Things that influence Estimates
Project Size            Culture
                        Tools
Type of Software
                        Team Cohesion
Personnel Factors       Management

Programming Languages   …

Complexity
Constraints
Turnover
Experience
Are estimate important ?

                               Scheduling



      Defining
      Iteration

                  Estimation


                                      Prioritization




                  Break Work
What is estimate?


                  Estimate                                            Target   Commitment




Source : Want a better estimates ? Stop Estimating by Joel Semeniuk
Estimation Challenges

Provides an 95% Confident estimation for:
 How old I am ?
 How many bugs will have on the next project ?
 What is the distance between the earth and the
 moon ?
 What is the % surface of water on Earth ?
Question: What make you are confident at
95%
Important Laws of Nature

       Parkinson’s Law
            The amount of time which one has to perform a task is the amount of time
             it will take to complete the task.


       The Student Syndrome
            Start as late as possible


       Exponential Error Aggregation
            The more detailed – the further from truth you will be




Source : Want a better estimates ? Stop Estimating by Joel Semeniuk
5 law to Ponder from Joel Semeniuk

        ESTIMATION LAW


Source : Want a better estimates ? Stop Estimating by Joel Semeniuk
Estimate Law #1

                  Don’t estimate if you
                  can measure
Estimate Law #2

Estimate in a way human are good at !
Compare instead estimate unit
Estimate Law #3

         Measure things that are Measurable
                   Eg Start-Finish Time
                           versus
                Actual versus Planned Effort
Estimate Law #4

                   To mitigate the risk of
        Vision     incorrect effort
       Roadmap
                   estimations we reduce
                   the precision of our
       Release     estimates according
       Iteration
                   to how much we
                   know about what
                   we’re estimating.
         Day
                   This in turn helps us
                   to be more accurate.
Cone of Uncertainty
Estimation Law #5


      Use the different Metrics for different Estimates
GOOD PRACTICE TO START
Simple Practices that Work
             Use Sizing for Requirements


                   Measure Velocity


          Decompose into Task Just in Time in
                       Hours


               Calibrate and Recalibrate


          Use cycle Time to Predict Expected
                       Duration
Use Sizing for Requirement
Example : Ski trail difficulty
              Numeric Values
              ½, 1, 2, 3, 5, 8, 13, 20, 40, 50, 100

              Green Trail : ½, 1, 2, 3 (16)
              Blue Trail : 5, 8 (8)
              Red Trail : 13, 20 (8)
              Black Trail : 50, 100 (8)
Measure Velocity

The amount of product backlog that a team can handle in one single
sprint
Quantified in story points
Story point is an arbitrary measure to quantify the required effort to
finish an user story. Namely, how hard the story is. Loosely based on
Fibonacci series.


             Business Solutions
Size Stories

Story should fit in one Iteration
Story too complex should be split on multiple simple one
Fixed Cycle Time
              Remember Law #1 : Measure instead guessing




           Measure the time to resolve
Summary

Estimation Law
  1.   Don’t estimate if you can measure
  2.   Compare instead estimate unit
  3.   Measure things that are Measurable
  4.   Reduce the precision of our estimates according to how much we know
  5.   Use the different Metrics for different Estimates

Good Practices
  •    Sizing for requirement
  •    Hours for small task
  •    Velocity
  •    Size Story
  •    Fixed cycle time
Resources
Want better estimation ?
Alexandre Cuva
Email :                    Organizational Coaching
alexandre.cuva@altran.co   (Management 3.0, Scrum)
m                          Team Coaching
Twitter: @cuvaalex         (Scrum, XP, Kanban)

Blog: http://agile-        Technical Coaching
alexcuva.blogspot.com/     (TDD, BDD, C#, Java, Groovy)

Phone: +41 78 715 8309     Agile Training
                           (Management 3.0, Agile, Scrum, XP)

More Related Content

Want better estimation ?

  • 1. Want a better estimate ? Alexandre Cuva Coach Agile, CSM, CSPO, LTM3 HSPTP PMDay Bucharest - 2012
  • 2. Alexandre Cuva Email : Organizational Coaching alexandre.cuva@altran.co (Management 3.0, Scrum) m Team Coaching Twitter: @cuvaalex (Scrum, XP, Kanban) Blog: http://agile- Technical Coaching alexcuva.blogspot.com/ (TDD, BDD, C#, Java, Groovy) Phone: +41 78 715 8309 Agile Training (Management 3.0, Agile, Scrum, XP)
  • 5. Observation The current standard management system, does not provide satisfaction to all.
  • 6. My Actual Mission Help companies to make this sentence True for them “Organizations can become learning networks of individuals creating value, and the role of leaders should include the stewardship of the living rather than the management of the machine.” http://www.stoosnetwork.org
  • 7. How to have an Accurate Estimation ?
  • 8. Common Scenario How long do you think this project I think in 6 will take ? we need month we to finish in 3 will be done month. Didn’t hear me I say it should be done in 3 month !!!
  • 9. Problem of estimation From this picture Estimate the size of my submarine ?
  • 10. 3 points estimation The 3 point estimate 1. Take the worst belongs to the time scenario management knowledge 2. Take the best scenario area. It can also be used 3. Take the median during Cost Estimation. A three point estimate is a better estimate, compared E = (O+M+L)/3 to a single point estimate. Single point estimate simply gives you a single number scenario
  • 11. Things that influence Estimates Project Size Culture Tools Type of Software Team Cohesion Personnel Factors Management Programming Languages … Complexity Constraints Turnover Experience
  • 12. Are estimate important ? Scheduling Defining Iteration Estimation Prioritization Break Work
  • 13. What is estimate? Estimate Target Commitment Source : Want a better estimates ? Stop Estimating by Joel Semeniuk
  • 14. Estimation Challenges Provides an 95% Confident estimation for: How old I am ? How many bugs will have on the next project ? What is the distance between the earth and the moon ? What is the % surface of water on Earth ? Question: What make you are confident at 95%
  • 15. Important Laws of Nature Parkinson’s Law The amount of time which one has to perform a task is the amount of time it will take to complete the task. The Student Syndrome Start as late as possible Exponential Error Aggregation The more detailed – the further from truth you will be Source : Want a better estimates ? Stop Estimating by Joel Semeniuk
  • 16. 5 law to Ponder from Joel Semeniuk ESTIMATION LAW Source : Want a better estimates ? Stop Estimating by Joel Semeniuk
  • 17. Estimate Law #1 Don’t estimate if you can measure
  • 18. Estimate Law #2 Estimate in a way human are good at ! Compare instead estimate unit
  • 19. Estimate Law #3 Measure things that are Measurable Eg Start-Finish Time versus Actual versus Planned Effort
  • 20. Estimate Law #4 To mitigate the risk of Vision incorrect effort Roadmap estimations we reduce the precision of our Release estimates according Iteration to how much we know about what we’re estimating. Day This in turn helps us to be more accurate.
  • 22. Estimation Law #5 Use the different Metrics for different Estimates
  • 24. Simple Practices that Work Use Sizing for Requirements Measure Velocity Decompose into Task Just in Time in Hours Calibrate and Recalibrate Use cycle Time to Predict Expected Duration
  • 25. Use Sizing for Requirement Example : Ski trail difficulty Numeric Values ½, 1, 2, 3, 5, 8, 13, 20, 40, 50, 100 Green Trail : ½, 1, 2, 3 (16) Blue Trail : 5, 8 (8) Red Trail : 13, 20 (8) Black Trail : 50, 100 (8)
  • 26. Measure Velocity The amount of product backlog that a team can handle in one single sprint Quantified in story points Story point is an arbitrary measure to quantify the required effort to finish an user story. Namely, how hard the story is. Loosely based on Fibonacci series. Business Solutions
  • 27. Size Stories Story should fit in one Iteration Story too complex should be split on multiple simple one
  • 28. Fixed Cycle Time Remember Law #1 : Measure instead guessing Measure the time to resolve
  • 29. Summary Estimation Law 1. Don’t estimate if you can measure 2. Compare instead estimate unit 3. Measure things that are Measurable 4. Reduce the precision of our estimates according to how much we know 5. Use the different Metrics for different Estimates Good Practices • Sizing for requirement • Hours for small task • Velocity • Size Story • Fixed cycle time
  • 32. Alexandre Cuva Email : Organizational Coaching alexandre.cuva@altran.co (Management 3.0, Scrum) m Team Coaching Twitter: @cuvaalex (Scrum, XP, Kanban) Blog: http://agile- Technical Coaching alexcuva.blogspot.com/ (TDD, BDD, C#, Java, Groovy) Phone: +41 78 715 8309 Agile Training (Management 3.0, Agile, Scrum, XP)