SM CMM Integration, IDEAL, and SCAMPI are service marks of Carnegie Mellon University. ® Capability Maturity Model, Capabi...
What is this all about? <ul><li>In 1993 a highly reputed oilfield services company hired me into their Austin offices to w...
Peter’s Story <ul><li>One other addition to the team was Pete B. He came preceded by his record as a PM: 12 projects in th...
In The Two Years… <ul><li>I worked with Pete in many things. </li></ul><ul><li>Until one day, already friends, I challenge...
Will Pete’s Secrets Work for You? <ul><li>Well, they might.  </li></ul><ul><ul><li>your managers really manage your teams ...
Agenda <ul><li>Mike’s Pool Party (a parable of sorts) </li></ul><ul><li>Task Assignment and the On Time Fallacy </li></ul>...
Mike’s Pool Party <ul><li>Every year, in June, Mike Grandstand III holds a grandiose, larger than life pool party.  </li><...
Mike’s Pool Party is Exclusive <ul><li>Mike only invites friends, or friends of his friends. (Although lately a malicious ...
Invitation Hi Friend! You are invited to my pool party, on Saturday July 3 rd , at Great Hills and River Lane by the swimm...
Where the Heck is That Pool??? <ul><li>Wanda might have gotten the invitation by mistake, BUT SHE IS RESOLVED to go. </li>...
It’s Only Twenty Five Miles Away <ul><li>Now that’s a relief, because her car will not make it too far </li></ul>
She Gets Turn x Turn Instructions <ul><li>She asks for time of arrival at each intersection. </li></ul><ul><li>She factors...
The Day is Here <ul><li>Wanda starts her car, resolved to follow the plan to perfection. </li></ul><ul><li>She reads the m...
She Is Early  <ul><li>Now what? </li></ul><ul><ul><li>Should she speed towards the next intermediate point, stretching the...
Pete’s Projects <ul><li>In Pete’s Projects the planning phase included all stakeholders. </li></ul><ul><ul><li>mainly the ...
Pete Pill of Wisdom I <ul><li>Pete did not blindly follow the estimates of his teams, nor did he fail to consider the nois...
Pete the Risk Manager <ul><li>Pete did a great job of understanding his teams as persons and professionals </li></ul><ul><...
Probability: Laws of Randomness <ul><li>Pete took the estimates as the mean value of a distribution. From his data he had ...
Progress Through the Tasks <ul><li>The Project can be considered as a network of tasks. If we never return to a previous t...
Critical Path <ul><li>Given three tasks A, B, C, where </li></ul><ul><ul><li>A and B both must precede C, but can be done ...
How Pete Sees Critical Path Critical Path is still BCFG but the worst case scenario is ugly! 3  7  11 2  3  4 1.7 2 2.4 2....
More on Randomness <ul><li>If we believe that the estimate points to the mean value, then there is equal chance of being l...
Why Are we Never Early? <ul><li>Because we never let this happen. </li></ul><ul><li>When was the last time a programmer ca...
What Happens If Early is Impossible There is only chance of being late! 3  7  11 2  3  4 1.7 2 2.4 2.5  3  3.5 1.7  2  2.4...
How Pete Manages Early Delivery <ul><li>Pete has three tools of the same kind to ensure some of the tasks (if not all) com...
Pete’s Tools <ul><li>Walkthroughs </li></ul><ul><ul><ul><li>with walkthroughs Pete forces “good” insight and feedback of s...
How Pete Uses Peer Reviews total task duration  0%  40%  80%  100% Walkthrough FTR Inspection Insert good feedback and str...
Walkthroughs <ul><li>Author may distribute materials to review before review meeting </li></ul><ul><li>Reviewers examine m...
Technical Reviews <ul><li>Review leader identifies primary focus of review </li></ul><ul><ul><li>specific questions to add...
Inspections <ul><li>Author and moderator distribute materials for review </li></ul><ul><ul><li>work product under inspecti...
Conclusions <ul><li>Aiming to be on time is a guarantee of lateness if you do not allow for early completions on certain t...
Use Peer Reviews
Any Questions? thank you for your time [email_address] www.liveware.com
Upcoming SlideShare
Loading in …5
×

Dont Be On Time

437 views
355 views

Published on

Why not being early makes you late, and the way to solve it.

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
437
On SlideShare
0
From Embeds
0
Number of Embeds
14
Actions
Shares
0
Downloads
5
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide
  • Module 12: Tying It All together Module Length: 85 minutes (60 minutes for the presentation and 25 minutes for the exercise) Module Summary This module brings together discussions on relationships, representations, and equivalent staging. Note There is one exercise at the end of this module (Exercise 8: Map Statements to Process Areas). State that this module will integrate and summarize the key concepts we’ve learned in the course.
  • s
  • Establishing the critical path is essential to optimizing the time through a schedule. We need to know where the schedule is most constrained, to be able to modify it and remove those constraints. Talk about slack and be sure people understand the chart here.
  • Establishing the critical path is essential to optimizing the time through a schedule. We need to know where the schedule is most constrained, to be able to modify it and remove those constraints. Talk about slack and be sure people understand the chart here.
  • Establishing the critical path is essential to optimizing the time through a schedule. We need to know where the schedule is most constrained, to be able to modify it and remove those constraints. Talk about slack and be sure people understand the chart here.
  • Walkthroughs are informal. The author(s) generally calls the meeting and selects the team for the review. Material may or may not be circulated prior to the review. At the meeting the author(s) may give an overview, followed by a detailed discussion of the reviewed materials. Notes from the meeting usually go into the project file, although the author(s) may choose to issue a report of the meeting’s findings. Formal follow up is not generally required.
  • Technical Reviews are more formal than Walkthroughs. In Technical Reviews (as defined in the IEEE standards), a review leader is assigned to plan, organize and run the review. This starts with a determination of review focus, scope and objectives. The leader must confirm with the author(s) of the review material that the material is, in fact, ready to be reviewed. A review team must then be selected, and the material must be distributed to them for review. The leader may, optionally, conduct a review team orientation to familiarize the team with the material to be reviewed. The team reviews the material independently of one another prior to the review meeting. The leader conducts the review meeting, and may use a scribe to record the proceedings. After the meeting, the leader will prepare and distribute a report of the findings of the meeting, including recommendations and required corrective action. The leader may determine that a follow up review will be needed, based on the scope and magnitude of the corrective action.
  • Inspections are the most formal type of review, and the most effective at detecting defects. However, they are also the most time consuming and expensive type of review to perform. For that reason, inspections should be very selectively used. You will note that more documentation is reviewed and produced by inspections. The inspection moderator is usually formally trained in inspection facilitation skills, and serves a similar role to that of the leader in Technical Reviews. However, note that the moderator is responsible for ensuring that rework has been completed. Note also some fairly subtle nuances in the nature of the meeting itself – there is no discussion of issues or suggestions for corrective action. The inspection focuses on the detection of defects – no more, no less. Finally, note the emphasis on metrics associated with inspections.
  • Dont Be On Time

    1. 1. SM CMM Integration, IDEAL, and SCAMPI are service marks of Carnegie Mellon University. ® Capability Maturity Model, Capability Maturity Modeling, CMM, and CMMI are registered in the U.S. Patent and Trademark Office by Carnegie Mellon University. When Good Enough Makes a Project Early Enough or: Don’t Try to Be On Time! Jorge Boria, VP at Liveware Inc. SEI Visiting Scientist High Maturity Lead Appraiser Intro to CMMI Instructor
    2. 2. What is this all about? <ul><li>In 1993 a highly reputed oilfield services company hired me into their Austin offices to work in their Process Group. </li></ul><ul><li>We were four people in charge over software development process in fourteen engineering centers (Austin, Corte Madera, Sugarland, Houston, Tulsa, Stavanger, Oslo, Hannover, Montrouge, Clámart, Machida, Gatwick, Cambridge and Delft), in seven countries (USA, Japan, The Netherlands, Germany, UK, France, and Norway) </li></ul>
    3. 3. Peter’s Story <ul><li>One other addition to the team was Pete B. He came preceded by his record as a PM: 12 projects in the last two years, all on time, all under budget. </li></ul><ul><li>He was the Master of the Universe of Project Managers. </li></ul><ul><li>He held the key to all things good. </li></ul><ul><li>Pete and I worked together for over two years. He got tired of Process Improvement and “went back to work” as a PM, in California. </li></ul>
    4. 4. In The Two Years… <ul><li>I worked with Pete in many things. </li></ul><ul><li>Until one day, already friends, I challenged the notion that he had really made the 12 projects come in on time. I accused him of sandbagging his estimates. </li></ul><ul><li>Pete is too great a person to let such pettiness get to him. Instead… </li></ul><ul><li>…He revealed his secrets! </li></ul>
    5. 5. Will Pete’s Secrets Work for You? <ul><li>Well, they might. </li></ul><ul><ul><li>your managers really manage your teams not just gather administrative data </li></ul></ul><ul><ul><li>job assignments can be made flexible and </li></ul></ul><ul><ul><li>individuals can be made accountable </li></ul></ul>IF
    6. 6. Agenda <ul><li>Mike’s Pool Party (a parable of sorts) </li></ul><ul><li>Task Assignment and the On Time Fallacy </li></ul><ul><li>To Be On Time You Have to End Early </li></ul><ul><li>Walkthroughs, Technical Reviews, Inspections and Grandfalloons </li></ul><ul><li>Connecting the Dots </li></ul>
    7. 7. Mike’s Pool Party <ul><li>Every year, in June, Mike Grandstand III holds a grandiose, larger than life pool party. </li></ul><ul><li>It is the party that makes all parties small. </li></ul><ul><li>Door prizes include Porsches, Norwegian cruises and Polar Bear Sight Seeing Tours. </li></ul><ul><li>Wanda has never been invited to Mike’s Annual Pool Party… </li></ul>Until Today!!!!
    8. 8. Mike’s Pool Party is Exclusive <ul><li>Mike only invites friends, or friends of his friends. (Although lately a malicious rumor states that he is inviting ALMOST everyone in town, but Wanda pays no attention to that. Why, he is inviting her!) </li></ul><ul><li>So he sends an invitation over mail, with an address and a date, but no directions. </li></ul><ul><li>Heck, if YOU are invited, you probably should know how to get there, anyway! </li></ul>
    9. 9. Invitation Hi Friend! You are invited to my pool party, on Saturday July 3 rd , at Great Hills and River Lane by the swimming pool! Party starts prompt at 2:00PM If you are late, you’ll miss the door prizes! Bring a bathing suit and a towel.
    10. 10. Where the Heck is That Pool??? <ul><li>Wanda might have gotten the invitation by mistake, BUT SHE IS RESOLVED to go. </li></ul><ul><li>So, she googles up the address. </li></ul>
    11. 11. It’s Only Twenty Five Miles Away <ul><li>Now that’s a relief, because her car will not make it too far </li></ul>
    12. 12. She Gets Turn x Turn Instructions <ul><li>She asks for time of arrival at each intersection. </li></ul><ul><li>She factors in an 18% increase due to the status of her car. </li></ul><ul><li>Now, she is ready! </li></ul><ul><li>Well, as soon as the day comes. </li></ul>
    13. 13. The Day is Here <ul><li>Wanda starts her car, resolved to follow the plan to perfection. </li></ul><ul><li>She reads the map instructions step by step and gets on the road. </li></ul><ul><li>Traffic is almost nonexistent! </li></ul><ul><li>She arrives at the first check point EARLY! </li></ul>
    14. 14. She Is Early <ul><li>Now what? </li></ul><ul><ul><li>Should she speed towards the next intermediate point, stretching the slack to make up for probable delays down the road? or </li></ul></ul><ul><ul><li>Should she stop by the roadside until she catches up to her schedule? </li></ul></ul>What Would YOU do?
    15. 15. Pete’s Projects <ul><li>In Pete’s Projects the planning phase included all stakeholders. </li></ul><ul><ul><li>mainly the team </li></ul></ul><ul><ul><li>but also all others involved as suppliers of services or customers </li></ul></ul><ul><li>Pete would present his plan in detail and request input from everyone AFTER letting them review the requirements and the plan overnight. </li></ul><ul><li>He then added corrections to this input from his own history data (no process database then). </li></ul>
    16. 16. Pete Pill of Wisdom I <ul><li>Pete did not blindly follow the estimates of his teams, nor did he fail to consider the noise that particular individuals introduced into his estimates. </li></ul><ul><li>Pete considered what was possible UNDER the circumstances BASED ON HISTORY that took into account idiosyncratic behavior </li></ul><ul><ul><li>Mark never works on Friday afternoons, although he thinks he does </li></ul></ul><ul><ul><li>Mary fails to consider rework in her estimates, typically a factor of 30% </li></ul></ul><ul><ul><li>Jules’ wife is about to have a baby, so the risk of him having to take time off is very large. </li></ul></ul>
    17. 17. Pete the Risk Manager <ul><li>Pete did a great job of understanding his teams as persons and professionals </li></ul><ul><li>His first secret is that: Do not follow history blindly, do not trust people’s estimates blindly. </li></ul><ul><li>Treat the initial estimates as a risky proposition! </li></ul>Use Your Common Sense
    18. 18. Probability: Laws of Randomness <ul><li>Pete took the estimates as the mean value of a distribution. From his data he had reached an understanding: </li></ul><ul><ul><ul><li>the estimates point to “the best expectation” </li></ul></ul></ul><ul><ul><ul><li>perhaps there is a chance of making that date </li></ul></ul></ul><ul><ul><ul><li>but there is a chance of being late. </li></ul></ul></ul><ul><li>Giving people a break, Pete thought of their estimates as the mean of a normal distribution: </li></ul>Why is this shape giving them a break? What are they estimating?
    19. 19. Progress Through the Tasks <ul><li>The Project can be considered as a network of tasks. If we never return to a previous task once we execute it, the graph has no loops. </li></ul><ul><li>This is a basic tenet of PERT and GANTT charts: no loops. </li></ul><ul><li>Slack time for a task is the amount of time (duration) it can be delayed without affecting the project end date </li></ul><ul><li>Critical path is set of tasks with longest duration, no slack, hence drives end date. </li></ul>
    20. 20. Critical Path <ul><li>Given three tasks A, B, C, where </li></ul><ul><ul><li>A and B both must precede C, but can be done in parallel </li></ul></ul><ul><ul><li>A takes 3 days </li></ul></ul><ul><ul><li>B takes 7 days </li></ul></ul><ul><ul><li>B is critical </li></ul></ul><ul><ul><li>A has slack of 4 days </li></ul></ul>Critical Path BCFG Any delay here delays the end date It could start 4 days late without delaying C A 3 B 7 C 2 E 5 F 9 G 2 D 3
    21. 21. How Pete Sees Critical Path Critical Path is still BCFG but the worst case scenario is ugly! 3 7 11 2 3 4 1.7 2 2.4 2.5 3 3.5 1.7 2 2.4 2 5 7 5 9 13 A B C 2 E 5 F 9 G 2 D 3
    22. 22. More on Randomness <ul><li>If we believe that the estimate points to the mean value, then there is equal chance of being late (ending the task after the estimate) as of being early (ending the task before the estimate). There is pretty much zero chance of ending on time. </li></ul> 50% of the time, we are late 50% of the time, we are early yeah, right
    23. 23. Why Are we Never Early? <ul><li>Because we never let this happen. </li></ul><ul><li>When was the last time a programmer came to see you and said: </li></ul><ul><ul><ul><li>“I could work on it some more days, but it is good enough already. Let’s inspect it!” </li></ul></ul></ul><ul><ul><li>Or do you actually hear this more often? </li></ul></ul><ul><ul><ul><li>“I need more time, with a few more days I can guarantee it is going to be a real beauty!” </li></ul></ul></ul>
    24. 24. What Happens If Early is Impossible There is only chance of being late! 3 7 11 2 3 4 1.7 2 2.4 2.5 3 3.5 1.7 2 2.4 2 5 7 5 9 13 A B C 2 E 5 F 9 G 2 D 3
    25. 25. How Pete Manages Early Delivery <ul><li>Pete has three tools of the same kind to ensure some of the tasks (if not all) come in early. </li></ul><ul><ul><li>Help the person in charge of the task with the tough decisions early enough </li></ul></ul><ul><ul><li>Force an objective decision of the readiness of the product at about 80% of the time to completion </li></ul></ul><ul><ul><li>If it can’t be early, make it as good as it can be, so that there is no later rework. </li></ul></ul>
    26. 26. Pete’s Tools <ul><li>Walkthroughs </li></ul><ul><ul><ul><li>with walkthroughs Pete forces “good” insight and feedback of senior people into every task </li></ul></ul></ul><ul><li>Formal Technical Reviews </li></ul><ul><ul><ul><li>with FTR’s Pete forces objective judgment of the degree of readiness of a work product </li></ul></ul></ul><ul><li>Inspections </li></ul><ul><ul><ul><li>with inspections Pete guarantees there is the minimum amount of rework in the future for a work product. </li></ul></ul></ul>
    27. 27. How Pete Uses Peer Reviews total task duration 0% 40% 80% 100% Walkthrough FTR Inspection Insert good feedback and straighten the course Decide, based on quality as judged by performer’s peers, on readiness Make sure the product is as good as it gets and no rework is going to be needed
    28. 28. Walkthroughs <ul><li>Author may distribute materials to review before review meeting </li></ul><ul><li>Reviewers examine materials on their own </li></ul><ul><li>At meeting, author gives overview of work product </li></ul><ul><li>Team attending has general discussion, during which author walks through work product in detail </li></ul><ul><li>Reviewers </li></ul><ul><ul><li>note errors </li></ul></ul><ul><ul><li>suggest changes </li></ul></ul><ul><ul><li>suggest improvements </li></ul></ul><ul><li>Notes are consolidated into a report for project file </li></ul><ul><li>Author may issue walkthrough report </li></ul>
    29. 29. Technical Reviews <ul><li>Review leader identifies primary focus of review </li></ul><ul><ul><li>specific questions to address </li></ul></ul><ul><ul><li>particular standards or specifications to meet </li></ul></ul><ul><ul><li>interfaces or dependencies to examine </li></ul></ul><ul><li>Leader invites team </li></ul><ul><ul><li>provides specific questions or issues to consider </li></ul></ul><ul><ul><li>distributes materials </li></ul></ul><ul><li>Reviewers examine materials on their own </li></ul><ul><li>Leader conducts review </li></ul><ul><li>Leader issues report of findings </li></ul><ul><ul><li>list of issues and/or deficiencies </li></ul></ul><ul><ul><li>recommendations of team on how to handle open issues </li></ul></ul><ul><ul><li>action items </li></ul></ul>
    30. 30. Inspections <ul><li>Author and moderator distribute materials for review </li></ul><ul><ul><li>work product under inspection </li></ul></ul><ul><ul><li>reference documents </li></ul></ul><ul><ul><li>inspection checklists </li></ul></ul><ul><ul><li>item logging forms </li></ul></ul><ul><li>Reviewers examine material and log defects </li></ul><ul><li>Reviewers attend logging meeting </li></ul><ul><ul><li>create common list from individual input </li></ul></ul><ul><ul><li>add items detected during meeting </li></ul></ul><ul><ul><li>no discussion of issues, suggestions </li></ul></ul><ul><li>Item list used by author to conduct rework </li></ul><ul><li>Moderator verifies completion of rework </li></ul><ul><li>All participants record data on defects found, time spent </li></ul>
    31. 31. Conclusions <ul><li>Aiming to be on time is a guarantee of lateness if you do not allow for early completions on certain tasks </li></ul><ul><li>Early completion is not part of the repertoire of skills that academics have taught us </li></ul><ul><li>To achieve some early completion you have to force it to happen </li></ul><ul><li>You cannot sacrifice quality in the altar of timeliness or it will come back and bite you. </li></ul>
    32. 32. Use Peer Reviews
    33. 33. Any Questions? thank you for your time [email_address] www.liveware.com

    ×