Effective Metrics

  • Uploaded by: Neville Amaral
  • 0
  • 0
  • November 2019
  • PDF

This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA


Overview

Download & View Effective Metrics as PDF for free.

More details

  • Words: 2,354
  • Pages: 14
Effective Metrics for Pragmatic Project Managers

Effective Metrics for Pragmatic Project Managers Johanna Rothman Rothman Consulting Group, Inc. 781-641-4046 Author of Hiring Technical People (look for it in December)

[email protected] www.jrothman.com

What Do You Want to Know About Your Projects? • Where are we? – – – – – –

Is the project on time? On budget? Are we managing the critical path? (Do we know the critical path?) Do we need more or fewer resources (people and other resources)? Are we making progress? Are we tracking to the plan? (Do we have a plan?) How good are our work products?

• How well are we working? – Do people know what they have to do? – How is morale? – Has the team reached the state of team flow?

• How effective are we? – Productivity measures (read Putnam and Myers’ Five Core Metrics: The Intelligence Behind Successful Software Management) © 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

2

1

Effective Metrics for Pragmatic Project Managers

– Cost to market – People and their capabilities – Work environment

• What do your customers care about the most? – Time to market – Feature set – Defect levels

Lo

w

fe De

cts

s itie bil apa ir c the nd le a op Pe

• Every project has constraints and requirements • Internal constraints: Your customers don’t care about these. You do.

Wo rk env iro nm ent Feature set

Project Constraints and Requirements

Ti me to Re lea se

Cost to release © 2003 Johanna Rothman

www.jrothman.com

[email protected]

3

Pragmatic Measurements • View all six sides of the pyramid – – – – – –

Time to release Feature set Defect levels People and their contribution to the project Work environment Cost

• You (your management) care differently about each of these, depending on your context

© 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

4

2

Effective Metrics for Pragmatic Project Managers

To Measure Project Completion … • Project completion is a function of: – How accurate your initial estimation was – How much progress you’ve made (how many features completed and how good they are) – Schedule progress

© 2003 Johanna Rothman

www.jrothman.com

[email protected]

5

Project Completion Measurements • Earned value (Measuring what’s been accomplished and comparing the accomplishments to the time used, rather than only measuring the time used for the project) • Progress towards release criteria • Estimates vs. actuals for major and appropriate minor milestones – The dates the project achieved the milestones vs. the planned dates

• Estimation Quality Factor – The histogram of the completion date over time: how good your estimation is

© 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

6

3

Effective Metrics for Pragmatic Project Managers

About Earned Value • Earned value makes sense for products that can be created in self-contained pieces – You can create a piece, and measure how long and how much it cost you to create it

• Earned value makes no sense for products whose parts are interdependent – If each part is interdependent, how can you tell when a part is done? – If you can’t tell when a part is complete, you can’t measure earned value

• Software projects are composed of interdependent parts – Projects that have a software component are composed of interdependent parts

• Projects with tangible deliverables are great candidates for measuring earned value © 2003 Johanna Rothman

www.jrothman.com

[email protected]

7

Schedule Metrics: Estimation • How do you estimate now? – Range of dates (1/15-3/1) – “about” (About Mar 1) – Spiral in on a date (Q1, February, 2/14-2/21, 2/18)

• If you choose a date, you can use DeMarco’s Estimation Quality Factor, a measurement of how good your estimation was during the project • EQF is good for: – An early warning sign to see if events outside your project are consuming people when they should be on your project – A check against the initial estimations for your next project – If you have a chance of completing this project sometime in the next millennium © 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

8

4

Effective Metrics for Pragmatic Project Managers

Calculating Where You Are vs. Where the Schedule Says You Are • Simple earned value discussion: – – – –

How much time have you spent? How much money have you spent? How much of the self-contained, tangible deliverables do you have? If you’ve spent more time and money than you have deliverables to show than you will overrun the budget and/or schedule

• Inch-pebbles for software projects – Software projects suffer from the “90% complete” problem ! It takes the first 90% of the time to complete 90% of the project. It takes the next 90% of the project to complete the other 90%. – If you can freeze or baseline requirements, you can avoid the 90% problem – Inch-pebbles help you create good estimates © 2003 Johanna Rothman

www.jrothman.com

[email protected]

9

When You’re Not on Track • Ask your staff why • Ask your staff to log their time for a week and let you know • Are you holding things up? • You have numerous possible actions when your project is not on track. However, you need to collect some data first, and then determine which actions to take.

© 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

10

5

Effective Metrics for Pragmatic Project Managers

Examples of EQF From 3 Projects Project 1 EQF 11/22

Estimated End Date

11/2 10/13 9/23 9/3 8/14 7/25 7/5 6/15 1/1

2/1

3/1

4/1

5/1

6/1

7/1

8/1

9/1

10/1

11/1

Date of Estimate

© 2003 Johanna Rothman

www.jrothman.com

[email protected]

11

Project 2 EQF Project 2 EQF 12/12

Estimated End Date

11/22 11/2 10/13 9/23 9/3 8/14 7/25 7/5 1/1

2/1

3/1

4/1

5/1

6/1

7/1

8/1

9/1

10/1

11/1

Date of Estimate

© 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

12

6

Effective Metrics for Pragmatic Project Managers

Context Switching Is a Huge Potential Drain • Multiple parallel projects waste time • If you want to get the work done faster, assign people full time without any other interruptions

Source: Quality Software Management, vol.1, Systems Thinking, Gerald M. Weinberg, Dorset House, New York, 1992.

© 2003 Johanna Rothman

Number of tasks

% Time on each task

1

100

2

40

3

20

4

10

5

5

More than 5

Random

www.jrothman.com

[email protected]

13

Project 3 EQF EQF, Project 3 10/7

Estimated End Date

9/6 8/6 7/6 6/5 5/5 4/4 3/4 2/1 1/1 1/1

2/1

3/1

4/1

5/1

6/1

7/1

8/1

9/1

Date of Estimate © 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

14

7

Effective Metrics for Pragmatic Project Managers

Measurements You Can Start Early in the Project • Do the requirements freeze? Ever? – How many requirements change over time? – Measure the number of requirements you have, the number of major/minor changes per week over the course of the project

• Do you have the people you need to complete the project? – Are they dedicated to the project, or are they working on other things, such as fixing problems from previous releases or other projects? – Measure the number of people assigned week by week. Show when you needed the people

© 2003 Johanna Rothman

www.jrothman.com

[email protected]

15

Example of Requirements Changes Chart Requirements changes

35 25

Major Reqts changed

15

Minor reqts changed

34

28

31

25

19

22

16

10

13

4

-5

7

5 1

Number of changes

45

Week

© 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

16

8

Effective Metrics for Pragmatic Project Managers

Example of Staff-Days Chart 16

14

12

10

People planned people actual

8

6

4

2

© 2003 Johanna Rothman

6

6

/9 /1 1

2

1

/1

/9

/9

www.jrothman.com

1

/1 0 1

9

/1

/9

6

6

6 /9

6 8

7

/1

/1

/9

/9 /1 6

5

/1

/9

6

6

6 /9

6 /1 4

/1

/9

/9 3

/1 2

1

/1

/9

6

6

0

[email protected]

17

Middle of the Project Trends • Defect trends – Defect find and close rates: Are you fixing defects as quickly as you find them? – Cost to fix a defect: How much does it cost you (at which time in the project) to fix a defect? – Fault Feedback Ratio: How many defects reappear?

• What is product performance, and is it good enough? – Do you know how to measure product performance?

• If reliability or performance are important to you, start measuring them now – Compare scenarios by build

© 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

18

9

Effective Metrics for Pragmatic Project Managers

Defect Trends, Early in the Project Defect Trends 70 60

Defects

50 New defects found

40

Defects closed 30

Defects open

20 10 0 1

2

3

4

5

6

7

8

9

10

Week

© 2003 Johanna Rothman

www.jrothman.com

[email protected]

19

Defect Trends: Found/Closed, Full Graph Defect Trends 140 120

Defects

100 New defects found

80

Defects closed 60

Defects open

40 20 0 1

3

5

7

9

11

13

15

17

19

21

23

25

27

Week

© 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

20

10

Effective Metrics for Pragmatic Project Managers

About Cost to Fix a Defect • The more proactive you are, the higher your costs will be later in the project – The fewer defects you find later, the higher the cost of each defect – The lower the total overall cost

• Make sure you know what you’re measuring • NEVER measure cost to fix a defect by individual • Cost to fix a defect is excellent for any project where you provide a product at the end. If you’re providing a service, you may have trouble using this metric. © 2003 Johanna Rothman

www.jrothman.com

[email protected]

21

Cost to (Find and) Fix a Defect, Initial Company (for a specific release)

# of Cost # of people per days person -day

Dan Avery

# of system test fixes

persondays

Average days to fix during system test

System test cost to fix a defect

5

$500

40

125

200

1.6

$800

10

$500

20

30

200

6.7

$3,333

• Avery’s project is different, so update calculation Company (for a specific release)

# of Cost # of people per days person -day

Dan Avery © 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

# of system test fixes

persondays

Average days to fix during system test

System test cost to fix a defect

5

$500

40

125

200

1.6

$800

10

$500

20

30

200

1.0

$200

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

22

11

Effective Metrics for Pragmatic Project Managers

Cost to (Find and) Fix a Defect, Total Picture Proj ect

System test cost to fix a defect

# of Syste m test fixes

Syst em test total cost

prerelease total cost

Average time to fix postrelease

post release cost to fix

postreleas e # of fixes

post release total cost

Total pre and post release defect fix cost

Dan

$800

125

$10 0,00 0

$100,00 0

15 persondays

$7,500

23

$172,500

$272,500

Ave ry

$200

30

$6,0 00

$37,250

5 persondays

$2,500

2

$5,000

$42,250

© 2003 Johanna Rothman

www.jrothman.com

[email protected]

23

Fault Feedback Ratio • What’s the total picture about productivity? – What do the creators create? How much good stuff, and how much to redo? – Best to measure this weekly

• FFR tells you how good the work products are initially, and how much wheel-spinning they are doing Date

Size in LOC FFR (this week only)

Average pre-release cost to fix a defect

1/1

425,000

14%

1 person-days

4/1

800,000

16%

3 person-days

7/1

1,500,000

12%

2 person-days

10/1

2,000,000

18%

3 person-days

© 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

24

12

Effective Metrics for Pragmatic Project Managers

Late in the Project Trends • Defect trends – How many problems are open? ! The more open problems, the longer it takes to handle each one – Do people choose the easiest problems to fix first? ! Then the remaining problems take longer to fix – How long does it take to fix a problem? ! If problems take longer and longer to fix, the system becomes unmaintainable – How many new problems are introduced when fixing problems? ! The higher the Fault Feedback Ratio, the longer the developers take to fix problems

© 2003 Johanna Rothman

www.jrothman.com

[email protected]

25

My Minimum Metrics • • • • • •

Schedule estimates, actuals, and EQF Staffing Requirements changes FFR throughout the project Cost to fix a defect throughout the project Defect find/close trends throughout the project

© 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

26

13

Effective Metrics for Pragmatic Project Managers

Summary • Choose what you’ll need to measure – Start at the beginning of the project

• Act based on your measurements – Projects don’t improve if you leave them alone

• No matter what, learn from your projects…

© 2003 Johanna Rothman

www.jrothman.com

[email protected]

27

Resources and References • Austin, Robert. Measuring and Managing Performance in Organizations , Dorset House, New York, 1996. • DeMarco, Tom. Slack. Broadway Business Books, New York, 2001. • DeMarco, Tom and Tim Lister. Peopleware, Productive Projects and Teams. Dorset House, New York, 1999. • DeMarco, Tom. Why Does Software Cost So Much. Dorset House, New York, 1995. • Weinberg, Gerald M. Quality Software Management, volumes 1-4. Dorset House, New York, 1992-1997 • Wysocki, Robert et al. Effective Project Management. Wiley, New York, 2000. • Hal Macomber’s weblog, weblog.halmacomber.com • I have a number of articles www.jrothman.com, and my Managing Product Development blog: www.jrothman.com/weblog/blogger.html © 2003 Johanna Rothman

© 2003 Johanna Rothman 781-641-4046

www.jrothman.com

www.jrothman.com [email protected]

[email protected]

28

14

Related Documents

Effective Metrics
November 2019 29
Metrics
November 2019 35
Metrics
November 2019 32
Metrics
November 2019 35
Metrics
November 2019 38
Effective
November 2019 38

More Documents from ""

Integrating 6s And Cmmi
November 2019 21
Effective Metrics
November 2019 29
Mind Meeting Metrics Sei
November 2019 19
Bite The Whole Apple
November 2019 18
5 Questions In Pi
November 2019 17
B'marking Roi From
November 2019 18