Going Beyond Methodology to Maximize Performance

A performance framework for both software and non-software development that has been used for approximately a decade, the Team Software Process (TSP) is a proven method to help teams to plan, evaluate, manage, and control their work.

The TSP Symposium 2014 technical program will go beyond the core methodology of TSP to encompass a broader range of complementary practices that contribute to peak performance on system and software projects. The unifying theme of the conference is quality. Ultimately, a quality product and service must be delivered on time and within budget, be secure, be sustainable, and provide value to end users.

Teams using TSP have achieved unprecedented quality and productivity improvements. However, we as a community have also learned that many benefits accrue from incorporating other aspects of engineering, leadership, and culture change into the disciplined practices of TSP.

Symposium Blog

Darryl Davis Previews His Presentation "Tales from the Quality Journey"

Yoshi Akiyama: Why I Travel from Japan to Attend the TSP Symposium

Noopur Davis on Her TSP Symposium Tutorial: Business Model Canvas and Product Canvas

New Keynote Speaker Added to TSP Symposium Technical Program

Architecture-Related Sessions at the TSP Symposium 2014

Bill’s First Three Laws of Software Maintenance

Expert-Led Tutorials Planned for TSP Symposium

Dan Wall on His TSP Symposium Participatory Session: Four Ways to Kill a Good Idea

Message from the TSP Symposium 2014 Technical Chair

About TSP Symposium 2014 Keynotes

Video Highlights

Tutorial Sessions

Make the most of your time in Pittsburgh by registering for one or more of the half- or full-day tutorials on Monday, November 3, or Thursday, November 6, 2014. Attendees who register for two or more tutorials will receive a 10% discount off all tutorials purchased.

The tutorial fee includes participation in the specified tutorial and morning beverages, lunch, and breaks on the day of the tutorial.

Monday, Nov. 3, 2014

Software Lifecycle Recipes: Making a Process Do What You Want It to Do—Building a Great Product, Meeting Customer Needs, Managing Time, Money, Risk, and Quality
(Full-Day Tutorial)

Neil Potter, The Process Group

Empowering Teams with Great Data: Using the Process Dashboard
(Full-Day Tutorial)

David Tuma, Tuma Solutions

The NAVAIR TPI Story: From Software Process Improvement to Process Improvement for All
(Half-Day Tutorial)

Jeff Schwalb, Brad Hodgins, Mark Stockmeyer, and David Saint-Amand, NAVAIR

Strategic Management of Technical Debt
(Half-Day Tutorial)

Neil Ernst, Ipek Ozkaya, and Robert Nord, Carnegie Mellon Software Engineering Institute

Thursday, Nov. 6, 2014

Eliciting Unstated Requirements
(Full-Day Tutorial)

Mary Beth Chrissis, Mike Konrad, Robert Stoddard, and Nancy Mead, Carnegie Mellon Software Engineering Institute

Business Model Canvas and Product Canvas
(Half-Day Tutorial)

Noopur Davis, McAfee, an Intel Company

View Full Schedule

Keynote Speakers

Philip Koopman
Carnegie Mellon University

Holly Ridgeway
PNC Financial Services Group

Jesse Schell
Schell Games

Richard Pethia
Carnegie Mellon
Software Engineering Institute

Participatory Sessions

Common System and Software Testing Pitfalls
Donald Firesmith, Carnegie Mellon Software Engineering Institute

The Moneyball Post Mortem: Earning the Hidden Value
William Nichols, Carnegie Mellon Software Engineering Institute

Four Ways to Kill a Good Idea
Dan Wall, The Wall Group

The Internet of Things and Insecure Design
Art Manion, Carnegie Mellon Software Engineering Institute

View Full Schedule

Stay Connected

Get the latest TSP Symposium news, important dates, and announcements.

Sign Me Up!

Pittsburgh, PA


Sheraton Pittsburgh Hotel at Station Square

300 W. Station Square Drive
Pittsburgh, Pennsylvania 15219

Telephone: 800-325-3535

Hotel Overview
Sheraton Pittsburgh Hotel at Station Square is located in downtown Pittsburgh, in the heart of a 52- acre riverfront complex that includes Station Square restaurants and entertainment. Relax in Pittsburgh's only riverfront hotel, adjacent to 20 restaurants within the Shops at Station Square, Bessemer Court, and the Gateway Clipper Riverboat fleet. Experience exceptional service while having easy access to downtown Pittsburgh attractions such as Heinz Field, PNC Park, the Consol Energy Center, the Rivers Casino, and more!

The TSP group rate of $135.00 is available until Monday, October 20, 2014, or until all rooms at that price have been reserved, whichever occurs first. A limited number of government rate ($125.00) rooms are also available.

Reservations may be canceled up to 6:00 p.m. on the day prior to arrival without penalty.

To make reservations by phone, contact the hotel at 800-325-3535 and ask for the TSP rate.

Update October 14, 2014: The Sheraton Pittsburgh Hotel is fully booked on the night of Sunday, November 2. You can still get the TSP group rate at this hotel for Monday through Thursday, November 3–6.

Parking is available at a self-parking facility for $23 per day.

Local Attractions

Once an industrial hub, Pittsburgh has shed its image as a smoky city and has earned the title of America's "most livable city" by Places Rated Almanac, Forbes, and The Economist while inspiring National Geographic and Today to name the city a top world destination.

Located at the confluence of the Allegheny and Monongahela Rivers and the head of the Ohio River, Pittsburgh was referred to as the "Gateway to the West" from its early days as a frontier fort and came to be known as both "the Steel City" for its more than 300 steel-related businesses and "the City of Bridges" for its 446 bridges.

With 90 charming and unique neighborhoods, Pittsburgh guarantees the perfect place for you to shop, dine, hang out, and explore.


Pittsburgh offers a wide array of restaurants to choose from in and around the city. Throughout the city, award-winning chefs, a slow-food movement, a popular food truck scene, and restaurants specializing in buying local are popping up everywhere.

You can choose from casual family spots, upscale and romantic locations, late-night bites, or convenient take-out. Pittsburgh offers Ethiopian and Mediterranean cuisine, BBQ and steak, fondue and seafood, and much more.


Pittsburgh is within 500 miles of more than half the U.S. population and less than a 90-minute flight from 50% of North America's population. It's under 6 hours by car or train to 9 states, D.C., and Canada.

The Pittsburgh International Airport continues to rank among the world's elite airports according to Condé Nast Traveler, J.D. Power & Associates, OAG Worldwide, and others. The airport offers more than 155 non-stop flights per day to 37 destinations, domestic and abroad.

The airport is approximately 19 miles from the hotel.

The Amtrak station is approximately 1.5 miles from the hotel.

Public Transit
Port Authority of Allegheny County is the public transit agency for the Greater Pittsburgh area, providing bus, light rail, incline, and paratransit service to thousands of riders daily.

Pittsburgh is featured on Google Transit. This partnership with the Port Authority of Allegheny County allows visitors to search for public transportation routes using interactive Google Maps technology.

Hours: 24 hours
Contact: (800) 258-3826

Hours: 24 hours
Contact: (412) 321-8100


The conference registration fee includes two full days of TSP Symposium presentations and working sessions; morning beverages, lunch, and breaks on November 4–5, 2014; and the evening reception at the Sheraton Station Square on Tuesday, November 4, 2014. Tutorial registration is separate.


Sponsor the TSP Symposium 2014

Many TSP Symposium attendees arrive in search of real solutions in new tools, training, and technology.
By sponsoring the TSP Symposium, you can make sure that the solutions they leave with are yours.

Download the TSP Symposium 2014 Sponsorship Guide


Regular after August 15

TSP Symposium only ... $1000
Half-Day Tutorial... $250
Full-Day Tutorial... $400

Available Discounts

Organizations with 4 conference registrants get the 5th conference registration for FREE. This does not apply to tutorial registration. All registrations must be completed at the same time. Simply contact Mandy Mann for detailed instructions.

Some individuals may be eligible to receive the discounted registration rate.

Speakers receive a 40% discount applied to their registration within the discount period in which they register.

TSP Partners may use the code TSP14PART to receive a 10% discount on conference and tutorial registration.

IEEE members may use the code TSP14IEEE to receive a 10% discount on conference registration.

Government and academic employees may use the code TSP14GA to receive a 10% discount on conference and tutorial registration. Eligibility is confirmed by the use of an email address ending in .gov, .mil, or .edu.

Pennsylvania residents will automatically receive a 10% discount upon entering a valid Pennsylvania address during the registration process.

Students at accredited academic institutions receive a 50% discount on conference registration and, subject to availability, registration for tutorials. You will be required to provide proof of current enrollment with full-time status at an accredited institution, and you will need to present your student ID and a government-issued photo ID at check-in to receive your registration materials.

These discount codes do not apply to tutorial registration fees unless otherwise stated. Discount codes cannot be combined.

Additional Info

Endorsement Disclaimer
This forum has been arranged for the purpose of scientific and technical information exchange. The SEI does not endorse the products or services of any participating organization.

Cancellation Policy
Refund requests received in writing by October 3, 2014, will be processed minus a $75 administrative fee. No refunds will be given after October 3, 2014. If you do not cancel and do not attend, you will be charged the full registration fee. Substitute attendees are welcome at no extra charge; however, we request written notification prior to the conference for preparation of registration materials. For refunds, please allow two to four weeks for processing after the conference. Registration fee payments are not transferable to other SEI events.

Permission to Film, Photograph, etc.
By registering, you grant Carnegie Mellon University and/or anyone acting on its behalf ("Carnegie Mellon") permission to photograph, film, or otherwise record and use your name, likeness, image, voice, and comments and to publish, reproduce, exhibit, distribute, broadcast, edit, and/or digitize the resulting images and materials in publications, advertising materials, or in any other form and for any purpose without compensation. You also agree to not photograph or videotape any portion of this event without the prior approval of Carnegie Mellon. Additionally, you grant the Software Engineering Institute ("SEI") permission to communicate with you regarding this Symposium, and other SEI-related services, via the contact information you have provided.

Copyright Permission
Please note that each author must sign off on the copyright permission, not just the primary speaker or author.

Alternative Methods

If you prefer to complete your registration by fax or mail, download the PDF registration form and return the completed form with a credit card or check payment via fax or mail.


TSP Symposium 2014
c/o Registration Systems Lab
779 East Chapman Road
Oviedo, FL 32765 USA

Download PDF Form

Need help with registration?

If you require any technical assistance during the registration process, please call 407-971-4451 or email registration@regmaster.com.

Blog & Social Media

Darryl Davis Previews His Presentation "Tales from the Quality Journey"

Yoshi Akiyama: Why I Travel from Japan to Attend the TSP Symposium

Noopur Davis on Her TSP Symposium Tutorial: Business Model Canvas and Product Canvas

New Keynote Speaker Added to TSP Symposium Technical Program

Architecture-Related Sessions at the TSP Symposium 2014

Bill’s First Three Laws of Software Maintenance

Expert-Led Tutorials Planned for TSP Symposium

Dan Wall on His TSP Symposium Participatory Session: Four Ways to Kill a Good Idea

Message from the TSP Symposium 2014 Technical Chair

About TSP Symposium 2014 Keynotes

Darryl Davis Previews His Presentation "Tales from the Quality Journey"

I've crossed the milestone of 30 years in my career in product development, manufacturing, management, and process improvement. My talk will be an experience-report presentation from my "quality journey," a term that I am borrowing from Watts Humphrey. I will share quality-management experiences and resulting insights from a journey that took me from engineering school to being an employee within corporations and finally to serving as a consultant. Here is a sample.

My engineering classmates at Auburn University and I were fortunate to receive a very practical education. As an undergrad, I was especially excited about the newly formed computer engineering program and a new professor brought in to help get it started. Dr. Charles R. Vick came to Auburn after leading challenging real-world projects. He was a strong promoter of software development as an engineering discipline. I took every one of his classes that I could. He explained that coding would be just one small part of what we would need to do in the real world. He taught project management. He emphasized the importance of quality. He stressed the importance of understanding, analyzing, and reviewing customer and user needs and requirements. He showed us the importance of design and design analysis as well as several techniques for design and code analysis and verification. For example, he required that we embed assertions throughout our programs, and I was amazed at how much that helped.

After graduating from Auburn in 1982, I went to work at Chrysler Corporation. That was right when Lee Iacocca was rebuilding the company. One reason they hired me was because they had hired one of my classmates several months before and he was performing extremely well. My first job there was to design and construct the software for new equipment to test electronic parts at various stages of assembly as the parts were being manufactured. Just before then, about the only electronics in a car had been the radio. But the microprocessor revolution was underway, and suddenly Chrysler was manufacturing things like spark control computers and throttle body injection computers for the engine, and electronic instrument panels and navigation units for the driver. Our parts would go into every Chrysler vehicle made, and we also supplied Volvo.

The task was extremely challenging. We had to run dozens of tests on each part in less than two seconds. It had to be written in assembly language. It required custom hardware that would be developed in parallel with the software. All we had to start with was an in-house-developed, real-time operating system kernel with a primitive input–output system. There was an absolute deadline for the next model year, and it was just months away. Since this was all new to the company, I had little guidance and there was no established process to follow. My managers were highly supportive, however, and trusted me for this important job. They asked if I had any engineering classmates from Auburn whom I would like to work with, and I quickly recruited the first of several friends who would join my team. The only thing we knew to do was apply what we had been taught at Auburn. We met with the test engineers to discuss their needs and requirements. We worked with the hardware designers to determine the hardware capabilities and constraints. I designed a testing controller that would become the core software of every test machine. I called for a software design review with the test engineers, hardware designers, and several members of the software department. I remember that we had an excellent discussion that led not only to refinements in the software design but also in the testing approach and hardware design. At the end of the session, I was surprised to hear participants comment that it was their first time at a software design review. I had assumed that what Dr. Vick had taught us at Auburn was the way most software was being developed in the field, but I was beginning to learn otherwise.

There was only one machine shared by our whole software department for typing in the assembly language code and cross-assembling it. The code could not be executed on that machine. Instead it had to be put on removable media, carried to another building about a mile away, burned onto a programmable ROM chip, and plugged into a circuit board on the target hardware. All this meant that we had extra motivation to strive for no defects from the beginning. As I waited my turn on the development machine, I would carefully review my hand-written code. I wanted it to assemble correctly the first time so that the next person could have a turn at the machine. And I wanted it to run correctly the first time so that I would not have to repeat the long process between coding and testing. This led to very few defects. What was interesting about this—and this reinforces the ideas that Watts ended up promoting in TSP—is that we were forced by these constraints to strive for code that was so good that it would assemble and run perfectly the first time. This I how I started working on my very first job, and it became a disciplined habit that would serve me well for the rest of my career.

The results were very successful. We had put together a great team and did what we had been taught, and the first couple of years were probably the most enjoyable years of my career, with the technical challenges, supportive managers, working closely with friends, the things we were able to create, the successes we had, and the recognition we received.

I eventually left Chrysler for Intergraph Corporation, which specialized in high-end computer graphics. One of the things that attracted me there was that all the software developers had powerful workstations of their own. It was a great tool environment, way beyond what we had at Chrysler. You could quickly type in code, compile it, and test it, all on a super-fast machine with a large graphical display. To my surprise, however, I discovered that in that environment quality was much lower than what we had experienced at Chrysler. While the tool environment made some things faster, it did not force the same discipline, and the behaviors that it led to were part of the problem. One lesson I learned is that tools and automation alone are not enough; the process has to come first. You have to figure out what is the right way to do the work, and then use tools to support and automate that. A sophisticated tool can be a double-edged sword; if the process is bad, automation will just allow you to do bad things more quickly, and the result is poor quality and rework that takes more time in the end. Based on this and other lessons, I would eventually lead process improvement efforts at Intergraph.

In my talk, I will share more stories that relate to things like the importance of design, defect reduction, and making things simpler and better. I'll tell about losing things—where many quality problems arose from a lack of simple version control and configuration management. I'll tell about producing otherwise great products that turned out to be not what the customer really needed, as well as my experiences with filtering out and preventing defects and making reviews effective. My overriding message will be that we have to be good in all of these areas if we want to consistently achieve high-quality results.

—Darryl Davis, Davis Systems

Yoshi Akiyama: Why I Travel from Japan to Attend the TSP Symposium

I have attended all the TSP symposia since the first one, in September of 2006 in San Diego, and have recognized that the TSP is the simplest but most operationally pragmatic approach for a team to conduct projects in a self-managed way and to improve their skills and capabilities as needed in the real world.

What sets TSP symposia apart from other conferences? I have found that teams' management could discover at TSP symposia new ways of improving their teams to meet business goals. TSP has proven to be synergistic with and capable of absorbing new demands such as upstream process, multi-team, and agile methods. And the SEI has six variations of the TSP for different types of teams to help you during your planning in figuring out the steps you need to take.

The people I interact with at the TSP symposia are great. In addition to sharing a common cultural foundation with other TSP practitioners, I met enthusiastic early adopters who are practically applying TSP to, for example, providing software-improvement services and developing medical devices and embedded systems. Learning about how people are successfully attacking software issues of today and the future helps me to plan next steps for my clients.

This year's TSP Symposium program is exciting. Addressing requirements and agility topics has become increasingly popular; TSP can easily absorb both to meet business demands. Cyber security and technical debt are also highly relevant topics to the TSP, and I am looking forward to participating in discussions among the TSP community on these topics and others in the program this year.

—Yoshihiro Akiyama, Next Process Institute, Ltd.

Noopur Davis on Her TSP Symposium Tutorial: Business Model Canvas and Product Canvas

My half-day tutorial is about the Business Model Canvas and Product Canvas. It is more workshop style than lecture, with lots of hands-on learning opportunities.

The problem we face is that most current processes such as Scrum and TSP begin with the assumption that requirements represent a winning product. We can think of this in terms of the three questions for enabling success:

  1. Are we building the right product?
  2. Are we building the product right?
  3. Are we building it at the right speed?

Many quality practices address the second question—including TSP, which does so especially well—leading to very high-quality products. Building at the right speed, the third question, involves other considerations such as automation. Then, there's the first question: are we building the right product? That is where the Business Model Canvas comes in. It provides the link to building the right product, leading to a solid backlog of work that engineers execute.

The Business Model Canvas looks at multiple aspects of creating a product. It starts with a value proposition, then looks at product differentiators, making our customers stick, distribution channels, key partners, and other considerations, covering all of these systematically. The Canvas helps create hypotheses so that when engineering teams start executing, it is possible to validate or invalidate those hypotheses, thereby providing a feedback loop that ensures that sound business decisions are being made.

From the Business Model Canvas, we then go to a Product Canvas with a focus on the product across multiple releases. The Product Canvas allows you to determine if you have the Minimum Viable Product, the smallest possible product that will have value in the marketplace so that you can get that out as quickly as possible and validate the business hypotheses.

In the workshop, we will present this methodology and give you the opportunity to practice it in a controlled setting. We will talk about what the Canvas is and break attendees into teams to work on well-known products or products relevant to their own work. Then we will assess the business models that they came up with. This will be a different way of thinking for attendees because we will look not at engineering goodness or quality goodness but at business goodness.

We anticipate a lively and enlightening session, and we hope that you will join us!

—Noopur Davis, McAfee, an Intel Company

New Keynote Speaker Added to TSP Symposium Technical Program

The Team Software Process (TSP) Symposium 2014 will be held November 3–6, 2014, at the Sheraton Pittsburgh Hotel at Station Square in Pittsburgh, Pa. The program committee is excited to announce the addition of a fourth keynote address to the technical program.

Dr. Philip Koopman, associate professor of electrical and computer engineering at Carnegie Mellon University, will present a Case Study of Toyota Unintended Acceleration and Software Safety. Investigations into potential causes of unintended acceleration for Toyota vehicles have made news several times in the past few years. Some blame has been placed on floor mats and sticky throttle pedals. But a jury trial verdict found that defects in Toyota's Electronic Throttle Control System (ETCS) software and safety architecture caused a fatal mishap. This verdict was based in part on a wide variety of computer hardware and software issues. Dr. Koopman's talk will outline key events in the still-ongoing Toyota UA story and pull together the technical issues that have been discovered by NASA and other experts. The results paint a picture that should inform not only future designers of safety-critical software for automobiles but also all computer-based system designers.

Dr. Koopman has served as a testifying expert witness for automotive unintended acceleration cases, including the 2013 Bookout/Schwarz trial. Additional consulting and research activities include advising the U.S. Department of Transportation on automotive safety standards, leading a team that identifies safety issues with autonomous vehicles for the U.S. Dept. of Defense, and investigating fly-by-wire data network integrity techniques for the Federal Aviation Administration. He was a keynote speaker for the 2014 International Conference on Computer Safety, Reliability, and Security (SafeComp). He has also served as both program chair and general chair for the International Conference on Dependable Systems and Networks (DSN).

Previously announced keynote speakers at the symposium are

  • Holly Ridgeway, chief information security officer for the PNC Financial Services Group and former deputy chief information security officer for the Department of Justice, where she was responsible for leading the activities, detection, monitoring, incident response, reporting, and security services of the Department of Justice Security Operations Center and the FBI's intelligence community
  • Jesse Schell, CEO of Schell Games, author of the critically acclaimed book The Art of Game Design: A Book of Lenses, and former creative director of the Disney Imagineering Virtual Reality Studio
  • Rich Pethia, director of the SEI CERT Division, which conducts research and development activities to produce technology and systems-management practices that help organizations recognize, resist, and recover from attacks on networked systems

We hope you will make plans now to join us in Pittsburgh on November 3–6!

Architecture-Related Sessions at the TSP Symposium 2014

The Team Software Process (TSP) has been proven to help teams to plan, evaluate, manage, and control their work. Architecture-centric engineering has similarly been employed for many years to help develop products that meet their business and mission goals. Part of our strategy in expanding the scope of this year's TSP Symposium has been to bring more architectural thinking to those who have already adopted and are using TSP. We at the SEI have seen how successful combining these techniques can be in the project we undertook with Bursatec, the technology subsidiary of the Mexican stock exchange.

Our experience has demonstrated repeatedly how important it is for a project to get the architecture right in order to develop a quality product at the end. Even when projects don't pay attention to architecture, they are making architectural decisions; however, the failure to consider the long-term consequences of such decisions can result in an escalating and debilitating accrual of technical debt. The tutorial that we will offer at this year's symposium, Strategic Management of Technical Debt (Neil Ernst, Ipek Ozkaya, and Robert Nord) will introduce and explore a way of thinking about the consequences of not being deliberate about architectural decisions. It will also arm attendees with strategies for managing technical debt so that they can take advantage of time-sensitive opportunities, fulfill market needs, and acquire stakeholder feedback.

Three SEI staff members who have led architecture and TSP work, including the leads for the Bursatec project, will present Architecture Best Practices for Project and Technical Leaders, sharing the key concepts that managers need to know to be sure their teams are using architecture effectively. They will share insights and behind-the-scenes tips about not just execution, but also making sure that management is positioned to make the project team a success.

Another architecture-related talk at this year's symposium will cover legacy-system modernization. We know that there are many existing systems that, while almost good enough for today's needs, are beginning to age, perhaps as a result of increasing technical debt. There are options in such cases—throw it away, incrementally improve it, shift to a new platform, and then modernize. The SEI team delivering Architectural Insights into Planning a Legacy System Migration—Phil Bianco, Michael Gagliardi, and William Wood—will share experiences and insights from having participated on projects modernizing large legacy systems.

Pat Donohoe of the SEI will offer an introduction to software product lines. Product lines are a great solution for dealing with families of related products. They provide a better way to manage common software assets than "clone and own," a too-common practice that leads to a variety of defects and unanticipated consequences. In addition to improving quality, product lines provide economic advantage through common ownership of important assets. As with legacy-system migration, there is a range of ways to adopt product lines, and the SEI has a done a lot of good work in the field over the years.

Peter Feiler of the SEI will also give a talk related to architecture, titled An Incremental Life-Cycle Assurance Strategy for Critical System Certification. Feiler's work addresses the early introduction of defects that are detected late in the life cycle. In safety-critical domains such as avionics, we have seen studies from organizations that make critical errors during the requirements or design phases and don't discover them until integration. Formal architectural modeling through languages such as the Society of Automotive Engineers (SAE) standard Architecture Analysis and Design Language (AADL) can enable detection of such defects much earlier in the life cycle. For example, we can capture architecture models for multiple suppliers, integrate the models, and detect problems in some cases more than a year earlier than would have been possible with a traditional approach.

We hope that you will consider attending one or more of these sessions at the symposium and learn something new about applying architectural thinking to your software-development projects.

—James Ivers, SEI

Bill’s First Three Laws of Software Maintenance

  1. If a software product is to survive, you must remove at least as many defects as you put in, on average, in each release.
  2. If you fail to satisfy Rule 1 and your product survives long enough, effort for fixing defects will grow with each release until rework crowds out other maintenance and enhancements.
  3. The longer you wait to achieve equilibrium, the greater the portion of rework that will be factored into each release.

How do I know? It’s simple math.

Let’s try some estimates, based on typical numbers I have seen in industry.

Typical developers produce 6–12 KLOC per year (500–1000 LOC per month). Industry typically releases 3–7 defects per KLOC.

Average find-and-fix cost for released defects is 1.6 per developer week.

Assume an average product and better-than-average defect-injection rate:

  • 750 LOC/month * 4 defects/KLOC = 3 defects released
  • 3 defects/1.6/week = ~2 weeks find-and-fix cost

So a month of development leads to two weeks of rework.

In other words, maintaining steady state implies dedicating about 1/3 of total effort to fixing previously released bugs. This estimate is consistent with industry anecdotes that I have heard. If you cut released-defect density from 4/KLOC to 1/KLOC, the ratio shifts to 1/12.

There are two possible ways to deal with these unalterable facts:

  1. Accept the steady state, and accept that you cannot produce as many products as you’d like. And make sure that the products you do produce are of acceptable quality by fixing the defects you’ve introduced (which will cut into available time by a large factor).
  2. Find ways to reduce defects, thereby shrinking the portion of each development iteration needed to fix legacy bugs. This gives you more time to develop new products. Over the course of a year, if you increase quality sufficiently, you might be able to produce, say, three new products rather than two.

Agile developers refer to maximizing the "work not done." Don’t waste your effort on unproductive activity.

Corollary 1)
Defects maximize the work not done (because you didn’t have the time to get to it).

Corollary 2)
If you want to maximize the wasted work not done, reducing defects is a pretty good place to start.

The first step toward improvement is to make a realistic plan, given current process capabilities, of the time that can be spent developing new products versus the time needed to fix bugs in old products. In my experience report at the TSP Symposium, SEMPRE: The TSP Software Engineering Measured Performance Repository, I’ll talk about mining our rich body of TSP data to help you determine how many defects you inject. This information helps you make better, more accurate, and more predictable plans and provides powerful motivation for achieving a lower defect rate.

Defects are perhaps the most predictable part of the software-development process. We’re all human, and we create more or less the same types of defects over and over again at more or less predictable rates. If you have data such as the data I will be presenting, you can estimate how many defects you are likely to inject and what you have to do to remove them—using data as the basis for your estimation, not just a guess.

During development, injecting dejects is inevitable, but taking them out is optional. To remove defects takes deliberate planning and a commitment. If you don’t remove them before they’re released, the defects will manage you: they will come back as customer defects in your next release. Do you want your customers to tell you want they want in the product, or what defects they don’t want? It’s your choice.

–Bill Nichols, SEI

Expert-Led Tutorials Planned for TSP Symposium

The dates for the 2014 TSP Symposium—November 3-6, 2014, in Pittsburgh, Pa.—are fast approaching, and I want to call your attention to the exciting array of tutorials that we have assembled this year as part of our technical program.

In this message, I am highlighting the tutorials on Monday, November 3, or Thursday, November 6. I encourage you to look at these because each one is expertly led and covers topics that we specifically selected because they hit the most important opportunities for improvement that organizations have. I find it personally important to keep adding more knowledge and methods to my personal toolkit. I believe if you study what each of these tutorials has to offer, you will have a hard time picking among them to choose the best one for you. Here are one-line summaries for each of the tutorials:

In Software Lifecycle Recipes: Making a Process Do What You Want It to Do—Building a Great Product, Meeting Customer Needs, Managing Time, Money, Risk, and Quality (Full-Day), Neil Potter of The Process Group will show you how to evolve implementations of Scrum to meet the needs of your customers, team, and organization.

David Tuma of Tuma Solutions, in Empowering Teams with Great Data: Using the Process Dashboard (Full-Day), will demonstrate how to use an advanced tool to power plans that are connected from personal to team to overall organizational levels. Participants will understand the behind the curtain architecture and power of the tool, learn how to get started quickly, and build toward mastery of empowering teams with great data.

In The NAVAIR TPI Story: From Software Process Improvement to Process Improvement for All (Half-Day), Jeff Schwalb and his NAVAIR colleagues will demonstrate NAVAIRís top-level approach to the relationship among product processes, the data collected and used during development, and how the data is used to analyze the quality of the product. This tutorial will also include a discussion of the process for future improvement.

SEI technical staff members Ipek Ozkaya, Rod Nord, and Neil Ernst will present Strategic Management of Technical Debt (Half-Day) and explore key concepts of technical debt. They will discuss an approach for communicating the tradeoffs of technical debt to colleagues and managers, including practical tools and techniques that can address part of the problem today as well as provide a foundation for managing tradeoffs based on models of economic impacts.

In Eliciting Unstated Requirements (Full-Day), Mike Konrad, Mary Beth Chrissis, and Bob Stoddard of the SEI will discuss the KJ+ method, which can be scalable to address the needs of multiple categories of stakeholders; be usable by a diverse, non-collocated team of requirements analysts; and result in a more complete set of requirements for subsequent system design, implementation, and sustainment.

In Business Model Canvas and Product Canvas (Half-Day), Noopur Davis, vice president of Global Quality at McAfee, will lead tutorial attendees in creating a Vision Statement to a Business Model Canvas, a Product Canvas, and a Minimal Viable Product backlog. The Product Canvas is the nexus between strategy captured in the Business Model Canvas and the concrete Epics and Stories needed by teams.

When you register for two or more tutorials, you will receive a 10% discount on all tutorials. Don't miss this opportunity to add even more value to your TSP Symposium attendance. If you have already registered for the TSP Symposium and wish to add one or more tutorials, you can update your registration or register for tutorials on site.

Good luck on picking the best tutorials for you.

–Alan Willett
2014 TSP Symposium Technical Chair

Dan Wall on His TSP Symposium Participatory Session: Four Ways to Kill a Good Idea

“Are you implying that I’m failing?”

“We’ve been successful; why change now?”

“You’re abandoning our core values.”

“It’s too easy; nobody else does this.”

“This requires a lot more study and discussion. We can’t rush into a thing like this.”

Sound familiar? If so, read on.

My participatory session at the TSP Symposium on Tuesday, November 4, from 3:30 to 5:00 is for anybody and everybody. There are times when all of us–in our work lives, our home lives, and our community lives–need to persuade people to do things that we advocate. Regardless of the value of the changes that we propose, whenever we try to motivate others to change behavior, resistance is natural and inevitable. In this participatory session, I will discuss and demonstrate the kinds of resistance for which you as an advocate for positive change need to be prepared and give you some tactics for combating it so that your idea survives and thrives.

There are four key behaviors that people are likely to employ to derail your idea. The more important the idea, the more important it is to understand what these behaviors are and how to counter them:

  • Fear-mongering: scaring others into believing that a good idea is too risky to pursue
  • Delay: stalling an idea with never-ending questions or requests for more meetings
  • Confusion: throwing irrelevant numbers, facts, and questions into the discussion until support wavers
  • Ridicule: attacking you, the creator of the idea, thereby engendering indirect doubts about the idea itself

The methodology I will present for defending against these behaviors is counter-intuitive. Our natural inclination is to stonewall the person who is trying to derail our idea and try to get them out of the discussion. Instead, I advocate encouraging these behaviors in order to put the people who use them on the defensive. This gets people’s attention and causes them to listen more closely to what you actually have to say.

To get people to make changes in a business setting, you are going to need more than just 51% support. As an agent of change, you have to anticipate objections that will come your way and be prepared to defend against them. You have to do your homework–know your audience, anticipate who is likely to attack and how they are likely to do so, welcome the attacks, and be prepared to diffuse them.

In this session, you’ll learn strategies for countering these four behaviors and the 24 most common variants. Then we will have breakout sessions that will give you an opportunity to try out some of these tactics. You will not only get to understand the theory but also get to practice it in a setting where you can see these tactics in action.

As a business consultant, I’ve seen and experienced all 24 of these. Sometimes through trial and error, I’ve come up with ways to counter them. As a change agent, I’ve found having these tools in my toolkit to be valuable. I hope that you will join us for this session so that I can share them with you!

Message from the TSP Symposium 2014 Technical Chair

Last November, I was asked and agreed to volunteer to be the technical chair for the 2014 TSP Symposium. As with many volunteer jobs I have taken on, I knew this one would be a lot of work! 

So why did I say yes? Mainly because I know that the conferences I've attended over the past 20 years have been a powerful shaping force in my career. They have enriched the community of forward thinking experts with whom I correspond and collaborate. Every conference I have been involved in has added to my toolkit of methods and ideas that benefit me and to the clients and organizations I work with.

For example, I remember a presentation where an executive talked about applying defect prevention techniques to his little league team with great results. He simplified the decisions the kids had to make, and gave them complete authority to make those decisions. The story was funny, energizing and was immediately applicable to work.

Another example, I remember sitting with Watts Humphrey at a conference in Boston in 1995. I argued with him about many of his premises and I wasn’t quite convinced of his methods. But this exchange of ideas changed my path and I worked alongside Watts at the SEI for a decade. Our debates were so entertaining, we just had to keep them going!

Each conference I have attended has stories like these and more. I expect the TSP Symposium to be the conference event of this year. It will be held in Pittsburgh, PA on November 3rd through 6th.
Perhaps I am especially excited because I am the technical chair. I worked hard with my program committee to recruit content and speakers who will entertain, educate and push me to even higher levels of performance and fun. 

I encourage you to check out the detailed program agenda. Even more, I encourage you to register and join us! Early-bird registrants get a significant discount and the first 50 to sign up will also get a choice of one of 12 premium books from the SEI series.

I have learned a great lesson from this experience. In spite of the hard work, this opportunity has paid me back greatly. Each challenge, each conversation, each step in putting together a world-class event has helped me grow in my quest for exceptional leadership.

Yours in the calm pursuit of excellence,

–Alan Willett, Oxseeker, Inc.
TSP Symposium 2014 Technical Chair

About TSP Symposium 2014 Keynotes

As TSP Symposium 2014 Technical Chair Alan Willett says, "There is one constant in the world of software development and that is the pace of change—it's fast. I encourage you to join this conference, where we don't just keep up with the pace of change, we set it."

We are excited about this year's technical program and look forward to telling you a lot more about it.

The conference will explore a broad range of practices complementary to TSP that contribute to peak performance on system and software projects. The unifying theme of the conference is quality. Ultimately, a quality product and service must be delivered on time and within budget, be secure, be sustainable, and provide value to users.

The TSP Symposium 2014 keynote line-up includes

  • Holly Ridgeway, chief information security officer for the PNC Financial Services Group and former deputy chief information security officer for the Department of Justice, where she was responsible for leading the activities, detection, monitoring, incident response, reporting, and security services of the Department of Justice Security Operations Center and the FBI's intelligence community
  • Jesse Schell, CEO of Schell Games, author of the critically acclaimed book, The Art of Game Design: A Book of Lenses, and former creative director of the Disney Imagineering Virtual Reality Studio
  • Rich Pethia, director of the SEI CERT Division, which conducts research and development activities to produce technology and systems-management practices that help organizations recognize, resist, and recover from attacks on networked systems

In addition to the keynote speakers, substantial technical program, and organized networking events, the TSP Symposium 2014 also offers practitioners an in-depth learning opportunity with full-day and half-day tutorials on software-engineering practices that support the development of high-quality products.

Watch this space for lots more detailed information about the exciting program we have assembled!

–Bill Pollak, TSP Symposium 2014 General Chair


Past Presentations


Official Proceedings


The ACE (Accurate Confident Estimating) Process
Carl Wyrwa, Beckman Coulter

Advanced Modeling of Teaming Data to Enable Superior Team Performance
Robert W. Stoddard, Dan Bennett, David Webb, Rushby Craig, Lance Moore - Hill Air Force Base

An Extension of the PSP PROBE Process to Help Students Make More Reliable Estimates in Early Stages of PSP Training
Yoshihiro Akiyama, Next Process, Inc.

See More


Official Proceedings

Using TSP to Develop and Maintain Mission Critical IT Systems
Alex Obradovic, Beckman Coulter

Success with the TSP: Improve Your Project Estimations with Statistical Analysis Tools
Michael Mowle and Kathy Krauskopf, Urban Science

The Hidden Secrets of the TSP Checkpoint
Liliana Cazangiu

See More


Official Proceedings

Keynote: High Maturity Practices: The Way Forward
Girish Seshagiri, Advanced Information Services, Inc.

Leading a Development Team Pilot for IT Project Management Master's Degree Students
Valentina Ivanova, New Bulgarian University

Why Does Agile Software Development Not Require the TSP Disciplines?
Yoshi Akiyama, Next Process Institute Ltd.

See More


Official Proceedings

Keynote: Evaluating and Ranking Software Methods and Practices
Capers Jones, Capers Jones & Associates LLC

TSP is Coming!
Bill Nichols, Software Engineering Institute
for Alok Goswami, Nedbank Sandton

First TSP Results at Ecuador and Colombia: A Shared Successful Effort
Pablo Henriquez, Procesix

See More


Official Proceedings

Keynote Address: To Boldly Go
David Webb, U.S. Air Force

Introducing PSP/TSP Massively in Mexico
Héctor J. González Santos, Kernel Technologies Group

Ten Years with TSP – A Retrospective and a Path Forward
Darryl Davis, Davis Systems

See More


Official Proceedings
Sample Presentations

Keynote Address
Bill Curtis, Senior Vice President and Chief Scientist, CAST

Using TSP at the MSG Level
Ed Battle, NAVO – Leading and Learning

Implementation of the TSP in Small and Medium Size Software Enterprises
Roberto Ramos, Kernel Technologies

See More


Official Proceedings
Sample Presentations

TSP national initiative
Ivette Garcia – Deputy Director General of Digital Economy at the Ministry of the Economy, Mexico

TOWA's TSP Initiative: The Ambition to Succeed
Yuri Ontibon & Miguel Serrano – TOWA

Accelerating CMMI Adoption with PSP/TSP - TCAIM
Gene Miluk, Jim McHale, & Tim Chick – SEI

See More


There were no official proceedings in 2007.

Sample Presentations

Game On! An Industry's Journey,
Karthik Bala, Guha Bala – Vicarious Visions, Inc.

Taking Ownership and Adapting TSP Successfully Over Time
Intuit Engineering Team

PSP Training for Everyone
Dan Wall – Vicarious Visions, Inc.

See More


There were no official proceedings in 2006.

Sample Presentations

MISC PMT: Combining TSP and CMMI Should Mean More Than Just an Anagram
Mike Konrad, SEI, Keynote Address

Process Improvement at NAVAIR Using TSP and CMM
David Saint-Amand, Synthetic Solutions, Inc.

A PSP Commercial Project
Rob Tonneberger, Northern Horizons, Inc.

See More