LEARN FROM NEW PRODUCT FAILURES

Postmortems conducted in teaching hospitals can help companies

profitably gain from the pain of their new product failures.

Jim Hlavacek, Craig Maxwell and Jimmy Williams, Jr.

OVERVIEW: Many companies do not learn from their called Morbidity and Mortality Conferences (or simply
new product mistakes. Consequently, the mistakes are M&Ms), have been the foundation on which the medical
repeated, wasting millions, even billions, of R&D and world learns from its mistakes. Physicians regularly
marketing dollars every year. Product development suc- gather behind closed doors to review unexpected patient
cess would be higher and far less risky if companies outcomes and deaths that occurred on their watch, deter-
would regularly analyze, widely share, and learn from mine what went wrong, and figure out what to do differ-
each unsuccessful new venture. The positive effect of ently next time. This is the one place where physicians
formal post-launch reviews on the top and bottom lines can talk candidly about their mistakes in a legally pro-
is often impressive. Post-launch reviews in companies tected environment.
we studied often improved the new product success ratio
from 30 to 50 percent. This article describes the needed The century-old practice of conducting hospital M&M
culture, process and steps to conduct post-launch re- conferences began in the early 1900s by Dr. Ernest
views from new product failures that closely parallels Codman at Massachusetts General Hospital, and is now
the methodology teaching hospitals use to learn from required for teaching hospital accreditation in the United
their mistakes. States. It is a systematic approach that utilizes the pa-
tient’s history, symptoms, diagnostic images, pathology
KEY CONCEPTS: post-launch reviews, new product tests, autopsy results, prescribed treatments, and other
failures, learning organizations, best practices. information to search for the root cause(s) results of an
Deliver me from the person who never makes a mistake adverse patient outcome. It is like a detective story in
and also from the person who makes the same mistake which a medical resident reports out in a non-threatening,
twice.—William J. Mayo, M.D. co-founder, Mayo Clinic. closed hospital forum what occurred and what should
have been done differently or better for the patient. By
Twice a month in teaching hospitals around the world, reviewing complications and deaths, M&Ms help the
100–200 people stream into auditoriums with the objec- entire multidisciplinary team learn from their mistakes
tive of extracting success from failure. These meetings, and supply needed corrective actions.

Jim Hlavacek is chairman and CEO of The Corporate and more than 130 divisions. He joined Parker in 1996
Development Institute, Inc.® and managing director of and was engineering manager of the Racor division for
Market Driven Management®, both management devel- five years. Before assuming his current position in
opment and consulting firms headquartered in Charlotte, 2003, he was the business unit manager for Parker’s
North Carolina, and specializing in innovation and Fuel Cell Systems business unit. He is an engineer by
growth strategies. He has authored more than 40 arti- training and is the named inventor on over 40 patents.
cles and five books on the front end of innovation and [email protected].
focused growth plans. He has been the principal investi-
gator for post-launch reviews at many Fortune 500 Jimmy Williams, Jr. is director of defense research
corporations, and serves on the board of directors at and development at Alcoa, Pittsburgh, Pennsylvania. At
Nucor, Xtek and Mueller. His Ph.D. is in new product Alcoa, he has responsibility for military and defense de-
and market development from the University of Illinois. velopment programs. Prior to joining Alcoa in 2002, he
[email protected]; www.corpdevinst.com. was employed at McDonnell Douglas Aircraft, which
later was merged into The Boeing Company. In 2000, he
Craig Maxwell is vice president–technology and inno- was the recipient of the Black Engineer of the Year Award.
vation at Parker Hannifin Corporation. He is responsi- He is a mechanical engineer and received his Ph.D. from
ble for innovation initiatives in all of Parker’s nine groups Washington University. [email protected].

July—August 2009 31
0895-6308/09/$5.00 © 2009 Industrial Research Institute, Inc.
M&M conferences model the professional leadership,
safe culture and processes needed for manufacturers to
routinely learn from failed new product launches. Why
Objective post-
don’t bright, highly educated people in R&D, engineer-
ing organizations and business units regularly learn from launch review
their new product development mistakes as teaching
hospitals routinely do? Unfortunately, there are too many
companies that lack a safe culture, objective analysis
replaces hunches
and a proven methodology to analyze and discuss the
lessons learned and translate them into better new prod- and finger-pointing
uct decisions (1).
with open discussion
Don’t Learn from Pain
Most of the roughly 50 CEOs, CTOs, general managers, of what to do
engineering, and marketing managers whom we talked
with said they should, but never do, objectively learn
from their new product failures. Some of the reasons for
differently next time.
not conducting postmortems that we heard frequently
from managers included:
All these reasons reminded us of the saying, “If we only
1. “We have time pressures and oppressive workloads.” looked back through hindsight, we could avoid a lot of
The need to get on with the next project as soon as one future big mistakes.”
project is over dominated the reasons we heard. Learning
organizations, however, will find or make the time to Learn from Your Product Failures
learn from their mistakes.
Successful new products typically have many parents
2. “What’s the value of looking backward?” Some while failures are often orphans. Discussing new product
companies believe little is learned from 20/20 hindsight success is easier than openly talking about mistakes. Typ-
by objectively looking at their mistakes. Many claim they ically, considerable euphoria surrounds a successful new
know the causes from informal discussions and Monday- product because victory-sharing is far less painful than
morning quarterbacking. These are often the same com- admitting and learning from errors. Sensitive topics and
panies where finger-pointing and blame occur. systematic deficiencies are unfortunately not always ad-
dressed at new product reviews or after the launch. James
3. “We don’t have a budget, people or resources.” The Dyson, the founder and chairman of the Dyson vacuum
daily pressure for improved bottom-line results with cleaner company, firmly believes design engineers learn
the same or fewer people is often behind this excuse. But more from failure than success. Dyson stated:
when you consider the costs when a challenging project
An engineer’s life is 99 percent failure. You don’t learn much from
fails after the launch, an objective analysis would be an
success and your successes are few and far between. An innovative
insignificant additional investment, not a cost. Learning engineer goes to work each day excited because you know there are
organizations will allocate the funds to analyze their new hundreds of problems that you have not solved. Failure is actually
product failures. like a drug that keeps you living on the edge and trying to come up
with the right solution.
4. “The root problems are too complex.” Products and
customers’ requirements are becoming more complex, Intuit’s co-founder Scott Cook told us:
but this is still an excuse. Most R&D project failures
The root causes of success are usually not raised or known. I don’t
are rarely caused by a single error. Errors are often a
think there is as much deep reflective thinking, soul searching, and
combination of individual, team and systemic defi- rigor around successes as there is with an objective postmortem.
ciencies. Most medical M&M cases are certainly as Even though Toyota has had many successes, we learned that they
technical and complex as any unsuccessful new product spend much more time determining what they could have done better.
project. When visiting Toyota, you will realize there is never a perfect new
product launch at the company.
5. “We don’t know how to conduct a postmortem review.”
This is probably the most understandable reason. This An objective post-launch review or postmortem replaces
article describes the needed culture, leadership and hunches and finger-pointing with an open discussion of
processes to select and objectively review failed new what happened and what to do differently next time.
products, identify the root causes, and make recom- Even when reviewing a successful new product, people
mendations for going forward. should regularly ask what went well, what went poorly,

32 Research . Technology Management

what needs improvement, and how they might create an
even greater success next time. An objective
Buried Answers from Autopsies
In today’s high-tech medical world, the autopsy, like a investigative team of
post-launch review, still provides a uniquely effective
means of evaluating decisions and exposing poor judg-
ments and bad habits (2). Physicians and design engi-
2–3 people must be
neers can miss important factors. Without autopsies,
physicians and product development people don’t know
selected for each
when they have missed something fatal and so are likely
to repeat the mistake. In short, they miss the chance
to learn from their and others’ professional mistakes.
project to be
Instead, they bury answers to honest and sometimes
careless mistakes. As Dr. Michael Gillette, a senior anes-
thesiologist at Carolinas Medical Center, told us:
reviewed.
If you want to reduce errors or improve your success rate, start by
evaluating the cases where you’re not sure why you lost them with an 4. Interviewing skills—sharp face-to-face listening skills
autopsy performed by an objective person who was not part of the and diplomatically raising questions about sensitive
attending team. negative outcomes are needed; observing body language,
tone of voice, and asking relevant, probing questions in
The hospital autopsy report is always part of the evi- a non-threatening way are critical skills gained from
dence presented at an M&M conference that involves a years of new product development experience.
fatal outcome. A team of qualified resident physicians
not involved with the adverse outcome conducts the 5. Persistence—in scheduling interviews and getting
M&M analysis, presentation and recommendations. In people to share documents and speak up about what
companies, this must be done by a highly competent and actually happened in each project, it is necessary to
objective post-launch review team that was not in any repeatedly ask what happened, why it happened, and
way associated with the failed project. collect evidence in the form of emails, past presentations,
technical data, and cost information.
Post-Launch Review Teams 6. Root-cause analysis—like a detective, a chronological
After choosing failed new products for review, an objec- timeline of events and decisions must be developed and
tive investigative team of two to three people must be root-cause analysis used to build a factual cause-and-effect
selected for each project. Members of each investigative case; this structure also helps concisely present each
team include the principal investigator and the appropri- case for discussion.
ate technical expert(s). The seven essential requirements 7. Organizational and communication skills—the princi-
of the technical experts and especially the independent pal investigator must orchestrate the process of scheduling
principal investigator conducting a new product post- and facilitating the interviews, collecting all necessary
mortem review are: information, putting evidence into an acceptable format,
1. Relevant experience—many years of hands-on expe- and effectively presenting the case and recommendations
rience studying and participating in new product suc- to a wide audience.
cesses and failures help to ask the right questions, identify The investigative post-launch review team is no place for
the symptoms, root causes, and then make practical rookies or those with little or no new product and com-
recommendations. mercial development experience. There should also be
2. Objectivity—the investigative team should have nothing professional respect and good people-chemistry among
to lose or gain by what was done wrong; the principal the qualified principal investigator, the technical expert(s),
investigator must be an outsider neutral to the division, and the CTO, CEO or senior management who commis-
business group or the corporation—never a member of the sioned the study.
off-track venture being studied.
3. Customer-centric—OEMs and end-use customers who Corporate M&M Learning Conferences
tried samples or prototypes but were unhappy or did not There are many factors to consider when conducting an
buy again are always interviewed; experience working M&M-type conference for a company’s new product
with customers helps to ask the right questions and create postmortems (3). The key success factors include the se-
more of an outside-in view to the analysis and can lection of cases, an objective analysis of what happened,
sometimes re-establish a tarnished customer relationship. an open forum to discuss the outcome, recommending

July—August 2009 33
 Performing M&Ms at Intuit, 3M and Toyota
We found excellent examples of regularly learning from failures that formally and informally followed the M&M process
at three notable organic-growth corporations: Intuit, 3M and Toyota.
Intuit, the software company that developed Quicken, QuickBooks and TurboTax, has grown mostly by internal ventures
or organic growth. Scott Cook, the company’s co-founder, regularly encourages everyone to speak up about new product
glitches during and after each project.
Intuit’s software engineers spend a lot of front-end time observing and interviewing customers before freezing a design.
Engineers are urged to voice their views firmly, especially about the customers’ unstated needs, without fear of offending
their colleagues, superiors or the CEO. Scott Cook encourages the design people to challenge anyone in the company by
providing constructive innovation inputs, but to avoid personal attacks. He flattens an already flat organization by wan-
dering around and telling everyone that it is okay to make new product development mistakes at Intuit if we all learn from
them and do not repeat them.
All three of us interviewed Cook and others at Intuit and found regularly scheduled formal meetings and daily informal
discussions of new product mistakes and near misses. Cook told us:
We have 8,000 pathologists who openly and safely speak about new product mistakes. They are failures only if we don’t openly share
and learn from the experience with no punitive actions and then do not repeat them.
Rick Jensen, Intuit’s vice president of product development, added to Cook’s statement:
Through an in-depth postmortem process, we recently documented numerous mistakes we made in launching a new product on a web-
site that was a dismal failure. By debriefing the few customers that bought the financial planning software package and interviewing
many that did not buy, we learned how shoddy our front-end work was. We’re now trying again with the new knowledge from the
mistake—a common process of how failure and near misses breed success in new product development at Intuit.
3M technologists openly discuss their problems and mistakes across departmental, divisional and group lines. The com-
pany has fostered a strong sense of attachment to the company as a whole at its annual Technical Forum. Every technical
person is invited to this event, including people from all 3M’s application and prototyping labs. Many divisions are in-
vited to present their most recent technical or commercial challenges, with the hope that colleagues will help them find
answers. The forum fosters trust, open dialogue, sharing, solving problems, and talking about failures.
Successful and unsuccessful projects, many resembling nightmares, are a major part of the discussions at each annual
Technical Forum. The late Robert Adams, senior VP of R&D at 3M, sponsored the first Technical Forum and visited the
nearby Mayo Clinic to adopt their M&M approach to learning from new product failures (4). At one of our new product
workshops, Adams stated:
We learned how Mayo Clinic in nearby Rochester, Minnesota, regularly and safely reviews mistakes in their M&M meetings and shares
the knowledge throughout the organization. The CEO, CTO and all senior executives must provide the internal protection for technologists

corrective approaches, and a leader or moderator who If you’re not failing you’re not stretching yourself far enough. We
maintains a safe and constructive learning environment owe a lot of our successes at Nucor to learning from project failures
throughout the presentation and discussion of each case. where we really stretched ourselves.
Let’s further describe the success factors for M&M con-
ferences that have been effectively applied to learning Dan DiMicco added,
from new product development failures in our own com- Challenging failed projects that required a technical and commercial
panies and others we interviewed. stretch and took people out of their comfort zone are prime candi-
dates for postmortem reviews.
1. Selection of cases to review
2. Regular reviews and recent cases
In the medical and corporate worlds, more reflection and
deeper learning comes from the more challenging cases. Most teaching hospitals have bimonthly two-hour M&M
These projects require significant risks, frequently fail meetings where each adverse outcome case is discussed
before succeeding, and demand learning from mistakes. for 30–40 minutes. Two to four cases are discussed at each
Dan DiMicco, a research metallurgist by training and M&M conference. Intuit, a key participant in our study,
the current chairman and CEO at innovative Nucor Steel, regularly reviews failed or off-track projects at every oper-
frequently quotes founder Ken Iverson: ating division’s monthly meeting. New product challenges

34 Research . Technology Management

 that exists for physicians at M&M meetings. A lot of our new product ideas get shelved, but people shouldn’t ever fear for their jobs when
that happens. The important thing is not to punish the people involved. Mistakes will always be made in developing new-to-the-world
products. But if a person is essentially right, the mistakes he or she makes are not as serious as the mistakes management might make if it
is so critical that it kills people’s initiative. Long before the buzzword boundaryless organization was used, 3M never had structural bound-
aries and divisional charters like many other companies. It’s perfectly okay to call someone anywhere in the company and offer help or
ask for help on a technical or commercial problem. We have a long-time saying at 3M that “products belong to divisions, but sharing
technology and application know-how belong to the entire company.”
Toyota consistently develops reliable vehicles at lower cost and at greater profit than its competitors, while sustaining a
steady flow of new products. Its new car and truck pipeline has resulted in consistent market share gains. Toyota, where
acquisitions are a rarity, has been an organic growth machine for decades. Most senior executives are engineers who
know how to build high-quality products. The foundation for Toyota’s success is to continually learn from itself and oth-
ers. The company believes in piloting everything first and assuming it won’t get it right the first time. There is no failure,
just deeper learning at Toyota, where learning is not separate from innovation.
Toyota’s intense interest in being successful long into the future stems from its embrace of hansei (hahn-say), the Japa-
nese word for reflection. The hansei method of leaning from new product mistakes was explained to us by several Toyo-
ta executives:
Hansei is the rigorous review conducted after product development action has been taken and is a sobering reality check, regardless
of a project’s outcome. It’s a stern and serious meeting, even after a success, to identify what went wrong throughout and at the end of
every project. Its true meaning is closer to introspection than reflection, because it fosters real insight and learning. In engineering at
Toyota, there can be no kaizen (continual improvement) without hansei. At Toyota, a hansei reflection meeting occurs at major mile-
stone events (often to avoid near misses) and after each automobile is designed, made and launched. The reflection period is a “what
went right and what went wrong” lessons learned event—with an emphasis on things that did not go well and then, as a team, develop-
ing an action plan to address the shortcomings.

Another Toyota executive told us that hansei is about reflecting on what did not go well, where, “You publicly feel bad
and promise never to make the same mistake again.” Hansei or self-reflection is a basic philosophy in the Japanese cul-
ture. Its meaning is to acknowledge your own mistakes and to pledge improvement.
Unlike in the Western world, where admitting your mistakes is often taken as a justification for penalty and serves as
punitive evidence, Toyota managers told us that hansei could be translated to mean that acknowledgement is the first step
to improvement. Furthermore, hansei new product meetings help avoid becoming complacent and arrogant by instilling
a culture of humility, learning and improvement as an everyday activity. To stop hansei means to stop learning. With
hansei one never becomes so convinced of one’s superiority that there is no more room or need for further improvement.
At Toyota, where everyone is seen as a knowledge worker, inventing and sharing new knowledge is not a special event,
it is an expected way of behaving.—J. H., C. M., J. W.

and near misses are also a regular topic, high on the 3. Standard case review format
agenda, at Intuit’s monthly senior leadership meeting.
There should always be a standard format to collect,
In large multi-business corporations, a quarterly formal analyze, discuss, and make recommendations for each
postmortem of one or two cases per session has helped failed product selected for review. All M&M conferences
institutionalize the learning process. In smaller or medi- we visited use a straightforward and simple one-page
um-size companies, this is typically a semi-annual or an- format. The analysis for companies should also be brief
nual forum. The older the case, the more apt people are and not resemble a lengthy report. The Table, next
to forget what happened. We have found that new prod- page, is a format for new product postmortems that was
uct postmortems should be conducted within 6 to 12 adapted from Mayo Clinic’s M&M conference format and
months from when they were killed or shelved. has been employed in many manufacturing companies.
In today’s workplace of shorter employee tenure with
4. Objective case analysis
the same organization, it is often necessary to interview
people from the project who are no longer at the com- Based upon our experience conducting many failed post-
pany. For every failed postmortem case study, customers launch reviews, the project manager, like the attending
who were involved with the new development must be physician, should never analyze or present his/her failed
interviewed to maintain an outside-in approach of “cus- product or case. Competent, independent and objective
tomers first” to the entire post-launch review process. investigators should gather chronological data about

July—August 2009 35
critical events and decisions for the failed project. Most
medical and company investigators use root-cause anal-
ysis to analyze and present each case. The principal
Intuit regularly
investigator should review the facts of the case with the
former project team before making the formal presenta- reviews failed or
tion. Every case is presented with the backing of relevant
theories, science and related published articles. The in-
dependent investigators must be objective when devel-
off-track projects at
oping and presenting the case because the audience of
technical or engineering professionals will sense imme- every operating
diately when the truth or objectivity is deserted.

5. Safe learning environment

division’s monthly
Every physician and company engineering executive we
interviewed stressed the utmost importance of creating
meeting.
and maintaining a safe learning environment before,
during and after each M&M meeting. Dr. Harry Rubash, The M&M conference must occur in a legally protected and closed-
chief of the Department of Orthopedic Surgery at Harvard door environment. If not, problem outcomes won’t be raised or the
Medical School, stated: discussion will be so vague and brief that no learning or corrective
approaches will occur. It must not be a punitive situation for the at-
In the earlier years, M&M conferences were confrontational and not
tending physician or any member of the attending medical team.
the safest learning environments. [We found the same in some compa-
nies new to conducting postmortems.] You must first have a non-puni- The atmosphere throughout each postmortem presentation
tive culture for M&M conferences to be effective learning experiences. must be factual and serious. We observed a high degree of
It is typically the role of the respective department chairman or CEO spontaneous discussion at the well-conducted new product
to create and maintain the safe, but constructive, climate of critique. postmortems. The discussion after each presentation is of-
He or she must have that mind-set embedded in the entire department
ten heated and blunt, but constructive. If tough and delicate
before each M&M case study and conference begins.
issues are avoided in the closed-door meeting, the entire
Dr. Daniel J. Berry, professor and chairman, Department session will be of questionable learning value.
of Orthopedic Medicine at Mayo Clinic, added to the All project team members admitted having a sense of
need for a safe learning environment: public shame from being wrong, but not a sense of guilt
when it was an understandable mistake. After one post-
mortem presentation, the project manager stood up and
Format for a New Product Postmortem said to us, “I had no idea what I didn’t know when we
jumped into the project.” The same reflective thinking
1. Background and feeling of shame was found at Intuit. Jana Eggers,
• Name of the failed venture. who heads Intuit’s innovation lab, told us:
• Dates project began and was terminated or shelved.
• New venture leader and cross-functional team members. When I conducted a “When Learning Hurts” session for a software
• Objective and qualified principal investigators. application that failed in its initial market but is now finding many
2. Inputs customers in a different market segment or application, team mem-
• Face-to-face interviews with people who were bers spoke of feeling the pain as their case was discussed. The feeling
participants in the project. of pain occurs at most Intuit postmortem review sessions.
• Face-to-face interviews with OEM and end-use
customers who were involved.
• Face-to-face interviews with distributors/dealers and/or 6. Conference leadership and the presentation
key suppliers.
• Obtain all e-mails, business plans, documents, trials, and The moderator or leader of the M&M conference sits near
project presentations. the podium where each case is presented. The CTO or VP
3. Methodology of engineering typically occupies this role in companies
• Develop timelines and milestones of critical events or and must have the appropriate technical experience and re-
decisions.
• Document the unfavorable outcomes with data. spect, and be neither passive nor weak when encouraging
• Develop fishbone diagrams for the project and processes. constructive debate about sensitive issues or questionable
• Develop a root-cause analysis on the fishbone diagrams. decisions. Furthermore, the moderator of a company post-
4. Recommendations mortem or M&M conference cannot be a social cheerlead-
• What went well for the project. er when seeking closure about what should have been done
• What went wrong for the project. differently. As Dr. Edward N. Hanley, chief of orthopaedic
• Lessons learned and corrective actions. surgery at Carolinas Medical Center, firmly stated to us:

36 Research . Technology Management

The leadership of the M&M meeting is the most crucial aspect of the Recommendations from Postmortems
entire M&M process. A good leader sets the learning tone, has the
reporting resident follow a timeline of critical incidents, presents the Most companies listed two to five major prescriptive
facts, keeps it non-personal, and enforces the ground rules. The M&M actions from each postmortem study. Some of the cor-
leader must also encourage open and often heated discussion, and rective actions from CTOs, commercial managers, and
then demand closure in terms of what the attending physician and our companies included:
team would do differently the next time—to say they’d do nothing
different is unacceptable. The M&M process is only as good as the 1. Model the front-end business case after the venture
person in charge of the M&M meeting. capitalists’ more rigorous approach to concept and fea-
sibility studies that are not biased to incrementalism.
From our experience at many company postmortem meet-
ings, the moderator should also help keep the case presen- 2. Have some engineering or design people dedicated
tation and discussion period on the topic and on time. 100 percent to new-to-the-world projects and separate
However, a strict adherence to time limits should not ham- them from technical people who support current customers
per the opportunity for people to speak up or learn from and existing product lines.
each case and develop better practice recommendations. 3. Have technologists conduct voice-of-the customer
The meeting leader should at times inject humor if a dis- interviews and relentlessly look for unmet user job needs
cussion gets too tense or someone in the audience be- at OEMs and end-users or the customer’s customer.
comes long-winded.
4. Require certification in program and project manage-
ment skills for all new product teams and stop using
7. Recommendations from each case bureaucratic new product tracking approaches.
Every well-led M&M meeting and company postmor- 5. To develop more robust solutions and systems, involve
tem review requires the resident physician or compa- strategic suppliers and co-suppliers as collaborators
ny’s principal investigator to state what they would have much earlier.
done differently. If it is a systemic problem, the changes
required in the current system are discussed, outlined to 6. Require a strong intellectual property case as part of
the audience and corrective follow-up occurs. Dr. Berry the front-end business strategy.
at the Mayo Clinic stated to us: 7. Develop value propositions and pricing based upon
Each case must be a constructive analysis, presentation, open dis- both laboratory and field trial data.
cussion and then a deep-rooted search for improved outcomes or
8. Shelve, kill or refocus projects earlier on different
recommendations that everyone at the meeting can benefit from. This
last step, Corrective Recommendations, is where the real individual,
market applications.
team and organization learning takes place. 9. Establish recognition and reward systems that encour-
Every M&M conference and new product postmortem age more company-wide collaboration and projects that
must be an environment where anyone can speak up with significantly grow sales and increase margins.
recommendations, regardless of rank, title, age, specialty, 10. Make process development a parallel activity to the
or experience. Younger physicians in hospitals and young development of new-to-the-world products.
engineers in companies were quite vocal that no rank or
In companies where postmortems were a routine activity
hierarchical positions should affect who makes recommen-
for learning from failed ventures, executives stated that
dations and what is said regardless of who is in the audi-
their new product success rate improved significantly. The
ence. In large companies, 100 or more people might attend
CTO a t a major specialty chemical company told us:
a post-launch review meeting where one or more cases
are presented. A number of companies widely distribute Since we began postmortems five years ago, we’re now getting a lot
the “company confidential” postmortem report within the more bang from our development and engineering resources. We
company. The names of project team members should nev- are also seeing fewer tweaks and more proprietary next-generation
er be listed or mentioned in the report or discussions. solutions.

The atmosphere at product postmortem meetings is

Common Threads at Company Postmortems
meant to discourage attitudes of denial. The postmortem
learning process, if well presented and correctly led, en- After attending M&M conferences and many postmor-
gages new product development people in a “corrective” tems in our companies and corporations that regularly
view of their mistakes. All managers reported to us that conduct them, we asked ourselves, “What common
the open forum resulted in learning and positively affect- threads were observed across these organizations?” With
ed their future performance due to the specific corrective the proven M&M methodology for learning from mis-
actions they later practiced. A number of executives shared takes as our reference point, we found three common
with us their corrective actions that resulted from con- themes at every M&M conferences, at company post-
ducting postmortems on failed projects. mortems, and especially at Intuit, 3M and Toyota.

July—August 2009 37
1. Put customers first
The Hippocratic Oath or a modern version of it is part of Strong CEO
the rite of passage for all physicians completing medical
school. The oath stresses keeping the good of the patient
(or customer) as their highest priority. Every M&M con-
leadership is
ference seeks to add to the body of knowledge to con-
tinually provide improved care to patients by asking,
“What can we do better for the patient the next time?”
mandatory to begin
Even though the employees at 3M, Toyota and Intuit
never took a Hippocratic Oath, we found a similar em- to create a learning
phasis on the customer as the primary reason they pur-
sue innovation. Like the M&M conferences, the patient culture from
or customer was the center of all activity when these
companies conducted new product postmortems. The
following examples show how “the customer first” is at
reviewing mistakes.
the core of these three companies’ innovation processes.
Toyota senior executives and engineers said they spent an learn from your actions more than you ever believe. You can move
inordinate amount of time with the dealers, sales people, values and create a learning culture if you so wish to start the cul-
and especially the technicians and service department tural change. Openly and in company-wide forums state how you
employees who work on many different cars beyond the screwed up with stories about how you were wrong and what you
warranty period. Toyota’s executives are obsessed with should have done differently. Be straight, tell the truth about the many
building better cars for unmet customer needs. mistakes you’ve made, and you’ll build a bridge to your innovators.
Have your direct reports do the same. Lead with humbleness and open-
3M has always encouraged technologists to visit their ness about mistakes—not with just hero stories and fraudulent behav-
customers’ factories, laboratories and workplaces where ior where mistakes are not safely discussed and learned from. Driving
people have job outcomes to perform—watch them, ask out fear must start at the top. Your people want and deserve a safe
place to work. There is no learning without error. When we make
questions and then bring them new prototype solutions.
mistakes, I often state, “How fascinating!” and then I ask, “What can
This deeply-ingrained, customer-driven philosophy of we learn from that situation?” When you bootstrapped any new ven-
technologists goes back to 3M’s founder, William F. ture, you faced set-backs, mistakes and errors almost daily. That gives
McKnight, who said: you a reference point for risk, fear, innovation, and growth the rest of
your life.
Get well beyond the purchasing agents because they know the unit
price of everything, but not the value-in-use of our new solutions.

Intuit’s Scott Cook told us he often states to their software 3. Leave egos at home
engineers, After we attended M&M conferences and similar new
No matter what your business problem is, go observe and talk in product reviews at Intuit, 3M and Toyota, and at our own
depth with your customers and prospects. Listen and observe intently. and other companies, we would later comment on each
Major wins only come after bathing and swimming with present and session with similar statements that included,
potential customers.
• “What a humbling experience that was.”

2. Senior leadership is committed to postmortems • “You could feel the peer-group pressure in the audi-
torium.”
When we speak of senior leadership, we first mean all
the Cs—chairman, CEO, COO, CTO, CFO—and then • “The discussion period painfully helped everyone
the group or senior vice presidents, general managers, learn how to avoid repeating the mistake.”
and finally the department managers. Many senior ex- Dr. Kevin Raskin at Massachusetts General Hospital
ecutives spend little or no time on organic growth issues summarized the always-humbling public learning expe-
and creating the necessary “fast and smart failing” and rience from mistakes at M&M conferences as follows:
learning culture that demanding projects require. In short,
strong CEO leadership is mandatory to begin to create a Physicians, especially surgeons, often are egocentric when their case
learning culture from safely reviewing new product mis- is selected. There is no place for physicians with large egos at any
takes. Scott Cook told us: M&M conference. In order to create a safe and constructive learning
environment, egos must be left outside the M&M conference room.
I believe if the CEO doesn’t start the process to create a new product Everyone must come into every meeting with the mind set of learning
learning organization, it won’t really start. You have to realize that as and sharing in a humble, non-defensive and non-threatening manner
a CEO, you’re a role model and an example for everyone. People when their case is selected.

38 Research . Technology Management

Storytelling Sustains the Learning Process ing new product success rates. Face-to-face learning
from mistakes, like the M&M conferences, is sometimes
A corporate culture of safely learning from new product
painful, but vitally necessary if product development
failures must be actively and continually maintained with
teams are to become more effective.
factual storytelling. Companies, like tribes, keep the good
parts of their culture alive with oral histories. David Pack- We stress the word teams over individuals because new
ard and Bill Hewlett, the engineers who founded Hewlett- product learning is always a cross-functional team process
Packard, were known for telling stories about failures, that involves multiple disciplines collaborating within
learning and successes as they wandered around HP’s and across an organization and with customers, suppliers
labs. Concrete, vivid and vicarious new product stories and often outside third parties. The time-honored way in
exert extraordinary influence because they transport peo- which medical teams learn from multiple disciplines,
ple (especially skeptical technologists) from the role of collaborate, and help each other safely learn from their
critic into the role of participant. New product failure and negative experiences is lacking in many R&D, engineer-
success stories keep pessimistic listeners at bay. Storytell- ing, marketing, and business organizations. Hallmark
ing bypasses the normal defense mechanisms including learning companies embrace failure for its information
denial, “we’re different,” and “it won’t work here.” value and thus minimize or reduce risk in future new
product development decisions.
The transfer of learning from mistakes through story-
telling is a far more compelling way to help change Against the huge cost of making the same mistakes a
mindsets and behaviors than pep talks or lectures. Sto- second, third or fourth time, the benefits of hindsight can
rytelling doesn’t replace objective root cause-and-effect contribute immensely to innovation success and share-
analysis for failed projects. However, it is ideally suited holder value. High-performing new product companies
to communicate change, stimulate, learn, and sustain realize that the ability to learn faster than their competi-
trial-and-error experimentation throughout the organi- tors is the only truly sustainable competitive advantage.
zation. Because new product storytelling and M&M con- These leading new product companies and the medical
ferences are non-adversarial and non-hierarchical, they profession provide proven and practical guidelines for
can be effectively used by people at all levels in the companies that have not yet learned how to profitably
organization to change, maintain or institutionalize the gain from the pain of their new product failures.
culture of learning from mistakes.
The favorite form of sustaining new product learning at Acknowledgements
3M is telling stories about new product development.
The many 3M stories, including the famous one about Edward N. Hanely, M.D., chairman, Dept. of Orthopaedic
Post-It Notes, help people connect with product devel- Surgery, at Carolinas Medical Center, provided a compre-
opment passion, setbacks, persistence, and success. hensive understanding of the M&M process and introduc-
3M’s penchant for new product development storytell- tions to attend M&M conferences at Mayo Clinic and
ing started with the company’s first innovation heroes. Harvard’s Massachusetts General Hospital. Scott Cook, co-
The former CEO, Desi DeSimone, said at one of our founder of Intuit, Inc., helped the authors focus on key con-
market-driven product development workshops: cepts in the research design and provided many insightful
observations. Brian Woolf and Lewis Kasper served as
New product stories are our alchemy. They are like a medicine that challenging reviewers prior to submitting the manuscript.
helps us learn, heal and try another experiment. New product stories at
3M show people—product champions, teams and executive sponsors—
how to transform dreams into innovation success. Stories help us to References and Notes
aim high, be persistent and learn from each other about technical and 1. Garvin, David A. 1993. Building a Learning Organization. Harvard
commercial challenges. The story-intensive culture at 3M isn’t just an Business Review, July-August, p. 85.
accident or act of folklore, it is a central way we encourage curiosity, 2. Dobbs, David. 2005. Why Autopsies Are Good for Us: Buried
creativity, experimentation, so-called crazy ideation, and a persistence Answers. New York Times Magazine, April 24, p. 14.
3. The three authors attended M&M meetings at Carolinas Medical
to discover new opportunities in spite of ever-present obstacles. Hun- Center, Cleveland Clinic, Harvard’s Massachusetts General Hospital
dreds of 3Mers sprinkle stories of success and failure wherever they go and Mayo Clinic, and interviewed physicians before and after each
in the company to perpetuate the fickle culture of innovation. The sto- M&M meeting. A total of 15 physicians were interviewed at these
ries of failure at 3M help reduce failures and also avoid near misses. four leading medical organizations. The authors also interviewed
executives at 3M, Intuit, Nucor and Toyota to observe and ask how
they routinely use the M&M approach to learn from their new product
and new process failures. More than 35 technologists, engineering
Learning Is a Team Process managers, commercial managers, senior executives, and CEOs were
interviewed at these four companies.
Knowledge creation, learning, and sharing information 4. The late Robert Adams, senior VP of R&D at 3M, planted the idea
are probably nowhere more important than in technical many years ago, with one of the authors, Jim Hlavacek, to research
organizations where information quickly becomes out- the medical M&M approach to learn from mistakes or failures. This
article is dedicated to Adams, who, for more than four decades, shared
dated. Product development projects that go wrong are many invaluable insights about managing technology, creativity,
the raw material for organizational learning and improv- innovation, and growth.

July—August 2009 39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.