Project Retrospectives A kinder, gentler, more productive way to learn from past mistakes

Topics Typical Post-mortem Project Retrospectives are are Different Planning and and conducting a retrospective Words Words of of Wisdom Summary Norman Kerth Copyright 2005 Software Quality Consulting Inc. Slide 2

Typical Post-mortem Post-mortems often fail fail because they: occur occur with with no no planning and and little little preparation are are viewed as as forums for for exacting retribution and and venting frustration are are led led by by a someone who who may may not not have have good good meeting facilitation skills skills may may take take only only an an hour hour or or two twoat at most most often often result result in in no no changes to to behavior As As a result, ROI ROI is is very low lowsince little is is learned and and nothing changes Copyright 2005 Software Quality Consulting Inc. Slide 3

Project Retrospectives Are Different Project Retrospectives are are events not not meetings Some key key attributes: They They are are planned Everyone comes comes prepared Trained facilitator plans plans events events and and leads leads discussions Focus Focus is is on on acquiring wisdom Typically held held off-site and and take take about about 2 days days (yes, (yes, days) days) By By making an an investment to to learn from past mistakes, organizations can can reduce likelihood they will will repeat same mistakes on on next project Copyright 2005 Software Quality Consulting Inc. Slide 5

Wisdom What is is wisdom? accumulated knowledge trait trait of of utilizing knowledge and and experience with with common sense sense and and insight ability ability to to apply apply knowledge, experience or or understanding with with common sense sense and and insight What group of of people do do we we often associate with wisdom? How do do people acquire wisdom? Source: WordNet 2.0, 2003 Princeton University Copyright 2005 Software Quality Consulting Inc. Slide 6

Business Motivation Why spend time on on Project Retrospectives? It It is is much much easier easier to to identify identify another s another s foolishness foolishness than than to to recognize recognize one s one s own. own. When When project project teams teams work work together, together, they they learn learn stuff stuff Some Some stuff stuff is is not not important important and and some some stuff stuff is is very very important important Important Important stuff stuff has has potential potential to to become become wisdom wisdom Acquiring Acquiring wisdom wisdomoften often comes comes directly directly from from experiences experiences whether whether good good or or bad bad End End of of a project project is is best best time time to to openly openly discuss discusswhat happened happened in in order order to to acquire acquire wisdom wisdom We We don t don t want want to to make make same same mistakes mistakes again again Kerth, N. L., Project Retrospectives A Handbook for Team Reviews, Dorset House, 2001. Copyright 2005 Software Quality Consulting Inc. Slide 7

Business Motivation Project Retrospective Goals: Collect factual information on on project performance Improve process, procedures, management, culture, etc. etc. Capture collective wisdom Repair Repair damage to to project team Enjoy Enjoy the the accomplishment Get Get the the story story out out Kerth, N. L., Project Retrospectives A Handbook for Team Reviews, Dorset House, 2001. Copyright 2005 Software Quality Consulting Inc. Slide 8

Project Retrospective Overview Norm s Prime Directive Regardless of of what we we discover, we we must understand and and truly believe that that everyone did did the the best job job he he or or she she could, given what was was known at at the the time, his his or or her her skills and and abilities, the the resources available, and and the the situation at at hand. Norman Kerth Kerth, N. L., Project Retrospectives A Handbook for Team Reviews, Dorset House, 2001. Copyright 2005 Software Quality Consulting Inc. Slide 9

Getting Started Select an an experienced facilitator Identify participants Facilitator interviews participants Identify required preparation Plan events so so everyone can can participate Copyright 2005 Software Quality Consulting Inc. Slide 10

Getting Started Select an an experienced facilitator who who was was not not associated with with the the project in in any any way way is is ideally ideally a disinterested third third party party is is not not biased biased has has good good meeting management skills skills Facilitator can can make make or or break break the the retrospective Identify participants Everyone directly involved with with project, including managers Explain Retrospective Process Get Get commitment to to participate Copyright 2005 Software Quality Consulting Inc. Slide 11

Getting Started Facilitator interviews participants To To get get a sense sense of of what what happened To To identify any any sensitive issues, personalities To To prepare to to handle these these issues issues should they they come come up up Identify required preparation Each Each participant asked asked to to recall recall what what happened from from their their perspective Each Each participant asked asked to to bring bring artifacts important to to them them and and be be prepared to to discuss their their significance to to project Copyright 2005 Software Quality Consulting Inc. Slide 12

Getting Started Plan events so so everyone can can participate Off-site locations are are best best Comfortable meeting room Breakfast and and lunch lunch brought in in each each day day Evening team-building activities Prizes Prizes and and fun fun Copyright 2005 Software Quality Consulting Inc. Slide 13

Day 1 Day Day 1 Objectives Reinforce idea idea that that retrospective is is not not about about finding fault, fault, but but for for learning how how to to do do it it better better next next time time Create Create a feeling of of safety safety so so people people feel feel comfortable speaking openly and and honestly about about what what happened Day Day 1 Agenda Agenda Establish Establish ground ground rules rules Create Create Safety Safety Exercise Exercise Define Define Success Success Exercise Exercise Lunch Lunch Discuss Discuss Artifacts Artifacts Exercise Exercise Project Project Timeline Timeline Exercise Exercise Evening Evening Events Events Copyright 2005 Software Quality Consulting Inc. Slide 14

Establish Ground Rules When someone is is speaking, we we will will not not interrupt them We We will will accept everyone s opinion without judgment We We will will talk talk from our our own own perspective and and not not speak for for anyone else else There will will be be no no jokes made at at the the expense of of anyone else else When someone is is holding the the designated object, then then only only that that person may may speak While everyone is is encouraged to to contribute, your your participation in in discussions and and exercises is is optional Kerth, N. L., Project Retrospectives A Handbook for Team Reviews, Dorset House, 2001. Copyright 2005 Software Quality Consulting Inc. Slide 15

Create Safety Exercise Purpose To To create create an an atmosphere where where everyone feels feels comfortable talking openly about about important issues issues Process Assess safety with with a safety safety scale scaleand and vote vote by by secret secret ballot Based Based on on the the results of of the the vote, vote, facilitator determines if: if: Everyone Everyone feels feels reasonably reasonably safe safe talking talking openly openly Or Or additional additional steps steps are are needed needed to to increase increase level level of of safety safety Bottom Line We We can t can t begin begin until until everyone feels feels safe safe talking openly and and honestly about about their their experiences Kerth, N. L., Project Retrospectives A Handbook for Team Reviews, Dorset House, 2001. Copyright 2005 Software Quality Consulting Inc. Slide 16

Safety Scale 5 Hey, no problem, I feel comfortable saying anything. 4 I ll say most things but a few things might be hard to say. 3 I ll share some things but keep some things to myself. 2 I m not going to say much. I ll let others bring up issues. 1 I won t let managers know what I really think. Kerth, N. L., Project Retrospectives A Handbook for Team Reviews, Dorset House, 2001. Copyright 2005 Software Quality Consulting Inc. Slide 17

Create Safety Exercise The The secret ballot is is held and and facilitator tallies results Mostly 4 s 4 sand 5 s, 5 s, there s no no problem - retrospective can can continue Mostly 2 s 2 sand 3 s, 3 s, there is is a problem and and facilitator needs to to address it it Copyright 2005 Software Quality Consulting Inc. Slide 18

Create Safety Exercise The The team forms natural affinity groups Team Team members silently move move toward people with with whom whom they they have have worked closely on on this this project Each Each group group finds finds a corner corner and and they they brainstorm ideas ideas for for how how to to increase safety Each group brings presents items discussed within group Once whole team reaches consensus, a second secret ballot is is taken and and hopefully, results mostly 4 s 4 sand 5 s Copyright 2005 Software Quality Consulting Inc. Slide 19

Define Success Exercise How do do you you define success? A successful project is is one one on on which which everyone says: says: I I wish wish we we could could do do that that over over again again the the very very same same way. way. Was this this project a success? Kerth, N. L., Project Retrospectives A Handbook for Team Reviews, Dorset House, 2001. Copyright 2005 Software Quality Consulting Inc. Slide 20

Define Success Exercise Some industry data 31% 31% of of projects are are canceled before before they they ever ever get get completed 53% 53% of of projects cost cost 189% 189% of of their their original estimates On On average, only only 16% 16% of of software projects are are completed on-time and and on-budget For For small small software companies, 78% 78% of of their their software projects are are deployed with with at at least least 74% 74% of of their their original features and and functions The Chaos Report, prepared by the Standish Group, 1994. Copyright 2005 Software Quality Consulting Inc. Slide 21

Project Facts List List some key key objectives achieved by by Project Team For For example: You You completed the the project (it (it wasn t cancelled) You You delivered X% X% of of all all promised features You You delivered ### ### bug bug fixes fixes in in addition etc. etc. Copyright 2005 Software Quality Consulting Inc. Slide 22

Artifacts A Project Retrospective is is like like an an archeological dig Most Most people tend tend to to only only remember what what occurred during during the the last last part part the the project We We need need to to uncover and and discuss evidence of of what what actually occurred throughout the the entire entire project Artifacts are are evidence About About a week week in in advance of of the the retrospective, each each participant is is asked asked to to search for for important artifacts related to to the the project Copyright 2005 Software Quality Consulting Inc. Slide 23

Examples of Artifacts Documents: Project Schedules Marketing Requirements Software Functional Specs Specs Architecture and and Design documents Significant E-mails Napkins with important architecture drawings Coffee cups, pizza boxes, a can can of of bug spray, or or other artifacts that that represent long hours of of work Copyright 2005 Software Quality Consulting Inc. Slide 24

Discuss Artifacts Exercise Artifacts are are placed on on table Each person discusses artifact they brought and and its its significance to to them on on this this project After all all artifacts are are discussed, group votes for for the: the: Most Most significant Most Most unusual etc etc Prizes awarded by by facilitator Copyright 2005 Software Quality Consulting Inc. Slide 25

Create Project Timeline Exercise Each group works in in a separate area Inclusive not not consensual activity Each group identifies significant events that that occurred during project: Write Write event event on on post-it post-it note note along along with with approximate dates dates Anyone who who thinks thinks an an event event was was important, creates a post-it post-it for for it it Use Use artifacts to to stimulate your your memory Copyright 2005 Software Quality Consulting Inc. Slide 26

Evening Activities Evening Activities planned to: to: Help Help repair repair damage to to team team Foster Foster sense sense of of belonging Have Have fun fun and and see see people in in different light light Focused around dinner dinner and and after after dinner dinner activities Copyright 2005 Software Quality Consulting Inc. Slide 27

Day 2 Day Day 2 Objectives Glean Glean wisdom from from Project Timeline Day Day 2 Agenda Review Results from from Day Day 1 Adjust Adjust ground ground rules rules if if necessary necessary Mining for for Gold Gold Exercise Lunch Lunch Complete Mining for for Gold Gold Exercise Closing Thoughts and and Reflections Copyright 2005 Software Quality Consulting Inc. Slide 28

Mining for Gold Exercise Five Five Lists: What What worked well well that that we we don t don t want want to to forget forget What What did did we we learn learn (Wisdom) What What should we we do do differently on on the the next next project What What still still puzzles us us What What do do we we need need to to discuss in in more more detail detail Copyright 2005 Software Quality Consulting Inc. Slide 29

Closing Thoughts and Reflections Are Are there any any issues that that need to to be be discussed that that we we haven t discussed so so far? far? Was this this retrospective worthwhile? What do do you you think will will happen as as a result of of this this exercise? Words of wisdom from a Jedi Master Copyright 2005 Software Quality Consulting Inc. Slide 30

Pay it Forward If If you you found found this this talk talk to to be be of of value Norm Norm Kerth Kerthis is a highly highly respected consultant who who developed the the Project Retrospective technique. Sadly, Sadly, he he was was in in a serious car car accident and and suffered a disabling brain brain injury. injury. As As a result, result, he he cannot work work and and lives lives on on a very very limited income. You You can can help help recognize his his contribution to to our our industry by by sending a small small donation. Checks can can be be made made payable to to Norm Norm Kerth KerthBenefit Fund Fund and and sent sent to to Norm Norm Kerth KerthBenefit Fund Fund c/o c/o Process Impact, 11491 11491 SE SE 119th 119th Drive, Drive, Clackamas, OR OR 97015-8778. You You can can also also visit visit Karl Karl Weiger s website www.processimpact.com/goodies.shtml for for more more details about about contributing to to the the fund. fund. Thanks. Copyright 2005 Software Quality Consulting Inc. Slide 31

Software Quality Consulting offers workshops on: on: Writing Writing and and Reviewing Reviewing Software Software Requirements Requirements Building Building Realistic Realistic Project Project Schedules Schedules from from Software Software Requirements Requirements Software Software Verification Verification & Validation Validation Accurate Accurate Estimating Estimating and and Scheduling Scheduling Using Using Yellow Yellow Sticky Sticky Method Method Predictable Predictable Software Software Development Development Peer Peer Reviews Reviews and and Inspections Inspections Improving Improving the the Effectiveness Effectiveness of of Testing Testing Risk Risk Management Management for for Embedded Embedded Software Software Development Development Get Get more more information information at atwww.swqual.com/training/on_site.html Sign Sign up up for for my my monthly e-newsletter Visit Visitwww.swqual.com/newsletter/Subscribe.htm Copyright 2005 Software Quality Consulting Inc. Slide 32