4568
MAY 19, 2011
STEVEN C. WHEELWRIGHT
WILLIAM SCHMIDT
Baria Planning Solutions, Inc.:
Fixing the Sales Process
Christy Connor looked up from her computer and gave a weary sigh. As the recently hired
director of North America Sales Support for Baria Planning Solutions (BPS), a consulting firm serving
manufacturers, Connor knew that she would be very busy trying to improve the disappointing
performance of her team. The Sales Support group’s difficulties were having a direct negative impact
on the performance of the entire company. The importance of her mission was highlighted by the
constant stream of calls, meetings, and emails that she had been fielding since she joined the
company last month. The urgency and critical nature of the problem was summarized by a recent
email from her boss, Brandon Ali, the President of North American Sales:
From:
To:
Subject:
Sent:
Brandon Ali
Christy Connor
2010 Performance
Mon 2/21/2011 4:23 A.M.
I have again looked at the win-loss report for 2010 and I am very concerned. Of the 271 qualified new
sales opportunities that arose in calendar year 2010, our win rate is now projected to be a dismal 15.5%.
We had originally forecast a win rate of 17.5%. Since our customers typically sign 3-year contracts, we
will be struggling for the next couple of years because we’ve missed our forecast sales productivity
target. Even more worrisome is the drop in our renewal rates. We had consistently achieved a renewal
rate of around 91% to 92% over the last few years, but in 2010 it is 84%.
The Sales group directors are complaining that salespeople don’t get the timely assistance that they
need from the Sales Support group. While your team is partly organized along the same industry
divisions as the Sales group, some of the directors are suggesting that your entire team be organized
into industry-specific divisions in order to provide sharply focused and fully dedicated support to each
Sales division. I know your people work hard, but we have apparently missed proposal deadlines for
several opportunities, and each time it undermined the customer’s impression of us.
I realize that you have inherited a bit of a mess, but it is critical that you get your arms around this
situation and resolve it quickly. We appear to be digging ourselves deeper into a hole with every
passing week. Please have a proposal to me by next week with recommendations on how we can turn
________________________________________________________________________________________________________________
HBS Professor Steven C. Wheelwright and writer William Schmidt prepared this case solely as a basis for class discussion and not as an
endorsement, a source of primary data, or an illustration of effective or ineffective management. This case, though based on real events, is
fictionalized, and any resemblance to actual persons or entities is coincidental. There are occasional references to actual companies in the
narration.
Copyright © 2011 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685,
write Harvard Business Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. This publication may not be digitized,
photocopied, or otherwise reproduced, posted, or transmitted, without the permission of Harvard Business School.
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4568 | Baria Planning Solutions, Inc.: Fixing the Sales Process
this situation around. As part of this effort, I want you to quantify the staffing, utilization and costs
from your proposal based on the sales support requests for 2010 and compare it to our actual staffing
and utilization in 2010.
When Connor was in business school, she had learned about analyzing service operations
processes using production management techniques. Believing she could apply those techniques to
the challenges at hand, she decided to investigate the scheduling and work processes for her team.
Company and Industry Background
BPS, a publicly traded firm with $95 million in annual sales, helped its customers reduce
procurement costs and improve the performance of their suppliers. Through a combination of
software, data analysis, project management, and consulting, BPS scrutinized its customers’ spending
categories, identified sources of potential savings through initiatives such as supplier consolidation
and purchase standardization, and implemented procurement and change-management projects to
realize those savings. Companies offering this type of solution—sometimes referred to as spend
analysis or spend management—had experienced relatively rapid growth over the last few years.
BPS, founded in 1997 to take advantage of the nascent popularity of similar tools and analysis
methods, was considered an early mover in the field.
Spend analysis was once the near-exclusive domain of smaller, specialty solution providers like
BPS. During the past several years, however, several large software and consulting firms, including
SAP and Accenture, have entered the market, leading to consolidation among smaller firms. BPS,
which originally served the U.S. energy sector exclusively, had survived by expanding its industry
coverage and capabilities through the acquisition of other narrowly focused industry-niche
providers. In 2007 BPS acquired a firm that primarily served the government sector, including
federal and state agencies and some large municipalities and nonprofit organizations. In 2008 BPS
acquired two more firms, one serving a range of manufacturing companies and another that
primarily served the retail sector.
While BPS initially allowed the acquired firms to continue operating semi-autonomously, the
company had since worked hard to integrate parts of each operation where synergies could be
achieved. The amount of effort required had not been trivial. While all four firms provided solutions
that addressed the same types of customer needs, the similarities ended there. In addition to serving
different industries, the firms had entirely different technology platforms, service delivery processes,
and even different fiscal year end-dates. Most of the pressing integration work was complete by the
end of 2009, although some technology migration projects were still in the works. By 2010, BPS
commanded a respectable 18% share of the market, but in a highly competitive market BPS could not
afford to lose any momentum in sales growth.
Recent Performance
The company’s performance in 2010 was extremely disappointing in both new client attainment
and existing client renewals. Exhibit 1 identifies the original and latest forecast for the sales
opportunities that were qualified in 2010. BPS categorized all sales opportunities into one of four
types:
1.
2
New Sale: a new opportunity with a company that is currently not under contract with BPS.
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Baria Planning Solutions, Inc.: Fixing the Sales Process | 4568
2.
Renewal: a renewal of a contract with an existing customer. The terms of the contract can be
materially different from the original contract, but there is often a significant amount of
overlapping scope between the two contracts.
3.
Expansion: an expansion of the services delivered to an existing customer even though the
customer’s current contract is not near its renewal date.
4.
Pilot: a special type of New Sale in which a company engages BPS in a short-duration pilot in
order to get more exposure to the BPS solution in the context of an actual project. Pilots often
enter the New Sales pipeline at the completion of the pilot project.
The most recent update to the sales forecast in Exhibit 1 was mid-February, 2011. By this time BPS
knew with certainty the outcome of all of the Renewal opportunities from 2010 and could make a
very good forecast of the outcome of the New Sale, Expansion, and Pilot opportunities. The total
number of opportunities in 2010 was higher than originally estimated, but the win rate was now
projected to be much lower than anticipated in most opportunity categories. Although the Sales
Support group was not entirely to blame for the lower win rates, it was widely recognized that the
team had difficulties meeting proposal deadlines and this had prevented BPS from putting its best
foot forward in several selling situations.
Organization of North American Sales
Although BPS served customers on a global scale, its direct sales force focused on companies in
North America. BPS had several sales partnerships that allowed it to penetrate foreign markets, and
many of BPS’s North American customers utilized the BPS solution outside of North America. BPS’s
sales partners typically required very little support during their sales cycles, and when they did it
was delivered by the product management team, leaving the North American Sales organization to
focus on its primary objective: drive new sales and renewals in North America.
The North American Sales organization was structured into three groups—Sales, Sales
Operations, and Sales Support (see Exhibit 2). The Sales group, led by Chuck Dee, had four units
organized by the four major industry sectors that BPS served:
1.
Energy, including oil and gas companies and utilities.
2.
Government, including all federal, state, and local government agencies, as well as a handful
of large nonprofit organizations.
3.
Manufacturing, covering companies in chemicals, construction, diversified manufacturing,
electronics, high technology, and transportation.
4.
Retail and Other. While this sales division consisted mostly of large retail and wholesale
operations, it was also used as a catch-all for any sales opportunity that BPS pursued which
did not fit neatly into the other sales divisions
The industry alignment of the Sales group was a legacy of BPS’s acquisition strategy, but the sales
leadership believed that it was extremely useful to maintain this industry orientation.
The Sales Operations group, led by Jane Albright, was responsible for competitive intelligence,
win-loss analysis, sales training, staffing, and managing the sales force compensation plan. While this
group was not directly involved in any sales opportunities, it served a pivotal role in building the
infrastructure for a successful sales force. Albright and Dee were sometimes at odds over how the
sales force should be structured. Dee felt that all of the teams in the Sales Support group should be
HARVARD BUSINESS SCHOOL | BRIEFCASES
3
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4568 | Baria Planning Solutions, Inc.: Fixing the Sales Process
aligned by industry sector, effectively providing fully dedicated support for the Sales group. Albright
believed that such a structure would require a large increase in staff. She and the former director of
Sales Support—Christy Connor’s predecessor—had successfully lobbied for the current hybrid
organization of the Sales Support group.
Organization of the Sales Support Group
During BPS’s acquisition spree it initially maintained the structure of each firm’s sales team and
sales support functions. Since these predecessor firms were industry-focused, the sales force and
support staff were originally organized by industry as well. In 2008 BPS hired Albright to serve as the
director of Sales Operations and provide a consistent administrative and organizational
infrastructure that each sales team could leverage. One of Albright’s first contributions was to help
streamline the Sales Support group, consolidating some sales support functions into teams that
served all industries and leaving one function, Proposal Support, with staff dedicated to each
industry. When Connor asked Albright about the rationale for this structure, she replied:
After BPS completed its string of the acquisitions the firm was under tremendous pressure
to get costs under control. Of course the market downturn that began in 2008 added another
level of urgency to the situation. We felt that it was imperative, however, to maintain our
industry-centric, consultative sales process. The only way we could hope to legitimately
achieve this was to keep critical elements of our sales organization organized by industry so
they would continuously be immersed in industry issues. After looking hard at the sales
support function we decided that we could streamline the team and maintain an industry
focus with the hybrid model that exists today. Data Engineering, Data Analysis, and Pricing do
not require any meaningful industry-specific expertise, but there are some real benefits to
having Proposal Support staffed with experts in the customer’s industry. There may still be
issues that need to be ironed out, but I remain convinced that the hybrid structure is critical for
us to effectively balance staffing constraints against the requirements for industry expertise.
Exhibit 3 provides a diagram of the process flow for the Sales Support group. The four teams in
the group have the following responsibilities:
1.
Data Engineering: collect data from prospects, transform data into a standard format,
examine data for consistency and quality, and resolve any data irregularities.
2.
Data Analysis: process data through BPS spend analysis software and identify
opportunities for cost reductions and supplier performance improvement.
3.
Proposal Support: validate, quantify, and prioritize cost/performance opportunities;
develop configuration of software, project management, and consulting to address the
identified opportunities; and assist the salesperson in developing and pitching proposal.
4.
Pricing: develop a pricing proposal in accordance with sales strategy, including options for
risk-sharing pricing, which tied the payments that BPS received from the customer to the
results BPS delivered to the customer.
This type of hybrid team structure did appear to be unique in the market. Most companies
roughly the same size as BPS in the spend-analysis space had organized their sales support teams
entirely by geography. Having an industry focus allowed BPS to distinguish itself from other midsized competitors through a deeper understanding of industry-centric issues, and put the firm at
parity with small single-industry and large diversified competitors.
4
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Baria Planning Solutions, Inc.: Fixing the Sales Process | 4568
Dee took a different view on the situation. He felt that a hybrid structure didn’t go far enough and
was adamant that his team was losing opportunities because the Sales Support team wasn’t able to
meet its deadlines. He made his position clear when Connor met with him, saying “Since some of the
Sales Support teams are shared across sales divisions they don’t feel the urgency or the ownership
that is necessary to push proposals across the goal line.”
Solution Selling Process
BPS had adopted a solution-selling process several years ago as customers—who had recently
come to enjoy unprecedented competitive choice in buying business solutions—began demanding
greater assurances that any solution that BPS proposed would achieve its promised cost reductions
and performance improvements. The new selling process was supposed to meet this demand and
thereby lead to significant increases in sales and renewal rates by providing tangible performance
guarantees to customers in a timely fashion.
Solution-selling was essentially a consultative sales process. It shifted a considerable amount of
effort from the post-contractual to the pre-contractual stage. Clients provided BPS with detailed data
for analysis, and BPS invested significant resources to gather and properly analyze the data. Typically
both parties signed nondisclosure agreements prior to data collection and often a letter of intent. The
end result was a highly detailed and execution-ready proposal that explained the services BPS would
deliver, the precise projects that BPS would implement, the timing of the project milestones, the
resources required from the customer, and any risk-sharing pricing or performance guarantees BPS
would provide.
If the solution-selling process proceeded as intended, customers gained by having a clear
understanding of the benefits of the BPS solution and guaranteed performance targets. BPS gained by
having clearly communicated expectations, which increased the customer’s confidence in the BPS
solution and dramatically improved the likelihood of attaining the guaranteed performance targets
(which are costly to BPS if not achieved). Assuming the customer agreed to move forward to a
contract, the process also benefited both parties in all but eliminating the need for further extensive
contract negotiations.
The solution-selling process required the Sales Support team to devote substantial upfront effort
to develop a proposal. Team members tracked the time they spent on each sales opportunity.
Exhibit 4 provides a summary of the time required for each Sales Support function across the
different types of opportunities. Regardless of the large workload associated with the process,
everything had to be completed in a timely fashion. Customers and prospects were unwilling to
significantly delay their solution selection choice due to the vagaries in the selling process of a single
vendor. The competition in the market was stiff and vendors had to be responsive with a high quality
proposal or risk early elimination.
Feedback on the process from BPS customers and prospects varied widely. While some companies
greatly appreciated the in-depth analysis and the greater confidence that it engendered in the BPS
solution, others complained that BPS “didn’t have their ducks in a row” by the time of the sales
presentation. In some cases, the BPS sales team has had to delay the presentation by days or even
weeks. Dee shared a recent experience with Connor:
I was on a call just last month with one of my salespeople trying to convince a prospect to
give us more time on a proposal review session. The prospect was polite but made it clear that
they were on a strict schedule and we had to either present on time or drop out of the running.
He was willing to give us a day or two, but we needed two more weeks. Unfortunately the
HARVARD BUSINESS SCHOOL | BRIEFCASES
5
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4568 | Baria Planning Solutions, Inc.: Fixing the Sales Process
Sales Support team just couldn’t pull all the moving parts together in time. We presented what
we had, but it was not well received. This was not a last minute project either…we had a full
four weeks of advance notice.
Members of the Sales Support group were all too aware of the criticisms that had been leveled
against them. While they believed that the solution selling process could be quite effective in helping
the company win sales opportunities, they also believed that the expectations placed upon the team
could be unrealistic. Connor had one-on-one discussions with each member of her team in order to
better understand the pressures that they faced. As one team member put it:
Our workload is crazy. Some weeks I have a reasonable amount of assignments and I can
readily meet all of my obligations. In other weeks I have such a backlog of requests that I find
myself working late every night and over the weekend. I love the fast pace, but when it gets
really busy I am forced to explain that the sales deadlines won’t be met without cutting corners
on the analysis…and sometimes not even then! My husband has given up trying to schedule
family activities during those times and takes to referring to me as his “shadow” wife.
Other team members echoed these comments and blamed the group’s difficulties on issues over
which they felt they had little control, including lack of visibility into the sales pipeline, being
brought in late on contract renewals, and sales support requirements that varied widely from one
sales opportunity to the next.
Demands on the Sales Support Group
With the exception of renewals, sales opportunities arose randomly within a quarter. Some of the
sales divisions tended to be busier than others across quarters, which seemed like an artifact of the
different fiscal year ends of the predecessor companies and the differing selling seasons among
industries.
Renewals were based on the expiration date of the customer’s current contract. The Sales Support
team typically entered the planning process for a renewal about 5 weeks prior to the renewal date,
and the renewal presentation with the customer was usually scheduled to take place 2 to 3 weeks
after that; this left the team 2 to 3 weeks to conduct the analysis for the renewal. By conducting the
analysis as close to the renewal date as possible the team could utilize the very latest information
on the current engagement with the customer. The four steps in the sales support process for
renewals had average times to completion of 8 hours, 12 hours, 22 hours, and 4 hours respectively
(see Exhibit 4), which seemingly could be completed in ample time before the customer presentation.
Initially Connor suspected that her team may simply be under-staffed, but she set aside this
suspicion as she dug into data on the frequency and composition of the support requests that the
team received in 2010. To gauge staff utilization in roles that had variable effort levels (exactly the
situation that the Sales Support group faced) the company used a rule of thumb: if a function utilized
less than 70% of its capacity, it was considered at “low” risk of being caught short of staff due to the
variable nature of the work. Utilized capacity between 70% and 85% represented “medium” risk, and
between 85% and 100% “high” risk. When utilized capacity exceeded 100% the function was
obviously already “over” capacity. Connor summarized the number of requests that her group
received by industry sector in each quarter of 2010 (see Exhibit 5) and created annual summaries of
the requests for 2009 and 2010 (see Exhibit 6). As a first draft of the analysis Ali had requested,
Connor used this data and the information on mean process times to create a rough estimate of her
group’s annual staff utilization levels (see Exhibit 7). She compared these utilization levels to the
company rule of thumb and it appeared at first blush that her team wasn’t being stretched too far.
6
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Baria Planning Solutions, Inc.: Fixing the Sales Process | 4568
Connor considered cross-training her staff so people could pitch in wherever the workload was
the greatest. After looking at the annual utilization levels she had pulled together, it wasn’t clear,
however, that cross-training would provide much benefit. She asked Albright about this, but the
response was deflating:
We explored cross-training in the past and it was a bit of a disaster for two reasons. First,
the skill sets across the four functions on your team are very different, which makes most types
of cross-training ineffective and cost-prohibitive, with a couple of exceptions. We found that
Data Engineers could be cross-trained as Data Analysts, and vice versa, at a per-person cost of
approximately one month of salary and benefits, but neither type of resource could become
proficient at proposal support or pricing. The Proposal Support team members could be crosstrained to support multiple industries at a per-person cost of approximately three weeks of
salary and benefits, but Chuck has a very low tolerance for that idea—he feels it’s moving us
further away from his desired organization structure. As for the Pricing team members…well,
let’s just say it is best to leave them focused on pricing issues.
Second, the cross-trained team members were most effective when they stayed in each role
for at least three months. When we shifted people into roles for shorter time periods, their
productivity went way down.
Next Steps
Connor reviewed her notes. There were a number of actions that BPS could try. Should she hire
more resources? Abandon the industry-focused sales support team structure? Reorganize to a fully
industry-centric organization as Dee requested? Attempt to streamline the solution-selling process?
Give up on solution-selling altogether? Try novel cross-training tactics? Or was there a more effective
way to improve the team’s performance? The challenge was determining what course of action
would actually work. Connor considered the irony of asking Ali to extend the due date for getting
him a proposed solution. She knew he wouldn’t agree—which made her empathize even more with
the people on her team.
HARVARD BUSINESS SCHOOL | BRIEFCASES
7
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4568 | Baria Planning Solutions, Inc.: Fixing the Sales Process
Exhibit 1
Forecast New Revenue from 2010 Opportunities, Excludes Contracts Not Up for Renewal
Opportunities
Win Rate
Annual
Revenue
Average
Duration
(years)
Total Contract
Revenue ($)
Original Forecast for 2010a
New Sale
Renewal
Expansion
Pilot
Total
250
117
60
80
507
17.5%
90.0%
90.0%
25.0%
$250,000
$250,000
$ 75,000
$ 50,000
3
3
3
1
$
$
$
$
$
32,812,500
78,975,000
12,150,000
1,000,000
124,937,500
Latest Forecast for 2010b
New Sale
Renewal
Expansion
Pilot
Total
271
117
53
103
544
15.5%
83.8%
90.0%
22.0%
$225,000
$250,000
$ 70,000
$ 50,000
3
3
3
1
$
$
$
$
$
28,353,375
73,534,500
10,017,000
1,133,000
113,037,875
a Original Forecast was finalized in May 2010.
b Latest Forecast was updated in February 2011.
8
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
-9-
Exhibit 2
North Americcan Sales Organiz
zation Chart, February 2011
4568
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4568 | Bariia Planning Solu
utions, Inc.: Fixin
ng the Sales Process
Exhibit 3
Sales Supp
port Process Flow
F
Exhibit 4
Sales Supp
port Time by Function
F
(hou
urs)
New Sale
Renewal
E
Expansion
Pilot
Sales Support Requests in 2010
271
117
53
103
1-Data En
ngineering
Min
Max
Mean
Standarrd Deviation
Weightted Mean Proceess Time
12
85
24
16
4
21
8
3
9
77
22
15
8
65
20
13
2-Data An
nalysis
Min
Max
Mean
Standarrd Deviation
Weightted Mean Proceess Time
11
80
24
15
3-Proposaal Support
Min
Max
Mean
Standarrd Deviation
Weightted Mean Proceess Time
18
155
44
30
4 - Pricing
g
Min
Max
Mean
Standarrd Deviation
Weightted Mean Proceess Time
4
24
8
3
10
Meean All
20
7
31
12
4
9
55
16
10
3
29
10
4
18
8
61
22
9
11
107
29
20
6
42
12
9
32
3
8
4
1
2
16
6
2
3
17
6
2
7
BRIEFCASES | HA
ARVARD BUSIN
NESS SCHOO
OL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Baria Planning Solutions, Inc.: Fixing the Sales Process | 4568
Exhibit 5
Requests for Sales Support, by Quarter (count)
2010
3rd
4th
Quarter
Quarter
1st
Quarter
2nd
Quarter
Annual
Total
Quarter
Average
New Sale
Energy Sector
Government Sector
Manufacturing Sector
Retail and Other Sectors
Total
19
8
22
12
61
17
18
27
10
72
16
20
23
7
66
20
8
34
10
72
72
54
106
39
271
18.00
13.50
26.50
9.75
67.75
Renewal
Energy Sector
Government Sector
Manufacturing Sector
Retail and Other Sectors
Total
9
2
5
11
27
10
5
5
4
24
5
14
11
2
32
4
3
27
0
34
28
24
48
17
117
7.00
6.00
12.00
4.25
29.25
Expansion
Energy Sector
Government Sector
Manufacturing Sector
Retail and Other Sectors
Total
2
1
4
4
11
4
2
4
2
12
4
5
3
2
14
3
3
10
0
16
13
11
21
8
53
3.25
2.75
5.25
2.00
13.25
Pilot
Energy Sector
Government Sector
Manufacturing Sector
Retail and Other Sectors
Total
5
4
7
12
28
6
4
9
7
26
6
3
8
5
22
5
4
12
6
27
22
15
36
30
103
5.50
3.75
9.00
7.50
25.75
Total Requests
Energy Sector
Government Sector
Manufacturing Sector
Retail and Other Sectors
Total
35
15
38
39
127
37
29
45
23
134
31
42
45
16
134
32
18
83
16
149
135
104
211
94
544
33.75
26.00
52.75
23.50
136.00
Renewals Won
Energy Sector
Government Sector
Manufacturing Sector
Retail and Other Sectors
Total
8
2
4
8
22
10
4
5
3
22
4
11
10
2
27
4
3
20
0
27
26
20
39
13
98
HARVARD BUSINESS SCHOOL | BRIEFCASES
11
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4568 | Baria Planning Solutions, Inc.: Fixing the Sales Process
Exhibit 6
Year Over Year Requests for Sales Support (count)
2009
2010
Growth
Energy Sector
Requests processed
New Sale
Renewal
Expansion
Pilot
Total
Renewals Won
Renewal Win Rate
New Sales Win Rate
65
27
11
15
118
25
92.6%
18.5%
72
28
13
22
135
26
92.9%
16.7%
10.8%
3.7%
18.2%
46.7%
14.4%
Government Sector
Requests processed
New Sale
Renewal
Expansion
Pilot
Total
Renewals Won
Renewal Win Rate
New Sales Win Rate
50
23
10
14
97
21
91.3%
24.0%
54
24
11
15
104
20
83.3%
22.2%
8.0%
4.3%
10.0%
7.1%
7.2%
Manufacturing Sector
Requests processed
New Sale
Renewal
Expansion
Pilot
Total
Renewals Won
Renewal Win Rate
New Sales Win Rate
94
44
18
32
188
40
90.9%
12.8%
106
48
21
36
211
39
81.3%
11.3%
12.8%
9.1%
16.7%
12.5%
12.2%
Retail and Other Sectors
Requests processed
New Sale
Renewal
Expansion
Pilot
Total
Renewals Won
Renewal Win Rate
New Sales Win Rate
37
16
9
24
86
14
87.5%
32.4%
39
17
8
30
94
13
76.5%
15.4%
5.4%
6.3%
-11.1%
25.0%
9.3%
12
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Baria Planning Solutions, Inc.: Fixing the Sales Process | 4568
Exhibit 7
Utilization Across Sales Support Functions
Workload
(requests)
Full Year 2010
1–Data Engineering
2–Data Analysis
3–Proposal Support
Energy Sector
Government Sector
Manufacturing Sector
Retail and Other Sectors
4–Pricing
Mean
Process
Time
(hours)
Resources
(headcount)
FullyLoaded
Cost
($/year)
Available
Work
Hoursa
Utilization
Utilization
Rule of
Thumb
544
544
20
18
8
7
$70,000
$75,000
14,080
12,320
75.8%
79.4%
Medium
Medium
135
104
211
94
544
32
32
32
32
7
3
3
5
2
3
$105,000
$105,000
$105,000
$105,000
$60,000
5,280
5,280
8,800
3,520
5,280
81.2%
62.5%
76.1%
84.8%
67.7%
Medium
Low
Medium
Medium
Low
a Hours per year assumption: 1,760
HARVARD BUSINESS SCHOOL | BRIEFCASES
13
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4420
APRIL 13, 2012
ROY D. SHAPIRO
PAUL E. MORRISON
Bayonne Packaging, Inc.
Cold grey light came through the window of John Milliken's cubicle in the Production office of
Bayonne Packaging at 6:30 AM on Monday, January 2, 2012. The new VP of Operations, Milliken had
arrived a half-hour before the first shift started on his first day of work at Bayonne to review reports
that had been prepared for him, and to begin his tour of the factory and interviews with key
Manufacturing and other personnel. When he had been hired mid-December, the president, Dave
Rand, had asked him to analyze Bayonne's operations swiftly and present his recommendations by
the end of the week.
Company and Industry Background
Bayonne Packaging, Inc., was a $43 million company located in Bayonne, N.J., a sub-chapter S
corporation founded 48 years earlier by Rand’s father. The board was composed of family members,
a local banker, and outside counsel. Bayonne was a "specialty packaging" paper converter that
produced customized, complex-design packaging that was used by industrial customers for
promotional materials, software, luxury beverages, and gift food and candy. Except for a few lowvolume operations such as laminating and gold- or silver-foil finishing, Bayonne provided all the
necessary services from design assistance through final delivery of the package. Bayonne's sales force
worked closely with customers to develop the artwork and package design, culminating in a proof
for customer approval. Bayonne then created the printing plates and die, sheeted the paper from roll
stock, printed the artwork on 4- and 6-color presses, die-cut the printed sheets into "blanks,"1 and
folded and glued the blanks into the final product, which was typically finished at this point and
ready to be shipped to the customer or a contents fulfillment house. In some cases Bayonne provided
additional finishing work if needed such as attaching string-and-button fasteners, Velcro dots, or
other attachments.
1 "Blanks" are die-cut paper shapes ready to be folded and glued into the final product. To create blanks from printed sheets,
the die-cutter sliced through the paper to cut out the shape, and made other, more shallow impressions to create the creases for
folding.
________________________________________________________________________________________________________________
HBS Professor Roy D. Shapiro and Boston University Professor Paul E. Morrison prepared this case solely as a basis for class discussion and not
as an endorsement, a source of primary data, or an illustration of effective or ineffective management. This case, though based on real events, is
fictionalized, and any resemblance to actual persons or entities is coincidental. There are occasional references to actual companies in the
narration.
Copyright © 2012 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685,
write Harvard Business Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. This publication may not be digitized,
photocopied, or otherwise reproduced, posted, or transmitted, without the permission of Harvard Business School.
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4420 | Bayonne Packaging, Inc.
The paper packaging industry grew rapidly in the 1980s and early 1990s as consumer goods
companies sought to make a greater impact with their promotional materials or moved their
promotional budget from print media and broadcast forms to the package itself at the point of
purchase. In addition, the explosive growth of software packaging, which featured expensively
printed large "boxes," provided additional customer market segments that were often willing to
spend freely to make a quick impact in a crowded marketplace. Bayonne had grown from just over
$10 million in sales in 1982 to $32 million in 2001. The company then faced new challenges with the
bursting of the dot-com bubble and the subsequent migration of software sales and distribution from
CDs to the Internet. Bayonne survived by diversifying into new markets where the company could
apply its great strength in innovative and difficult package design and the ability to fold and glue the
complex blanks.
President Rand had asked Milliken to focus on three problem areas: cost, quality, and delivery. At
the end of November 2011 Rand had fired the previous long-serving VP Operations. Rand told
Milliken, "Our sales are up—we have to run two shifts now. But we ran a loss for the first time last
year since 2001. [See Exhibit 1 for income statements.] We're getting more and more complaints
about quality, and, what might be even worse for our customers, we're delivering late more often. I
understand we're a job shop and there's usually a tradeoff between keeping your costs down, getting
good quality, and hitting your delivery promises—but lately it seems we can't even hit two out of the
three. What started to go so wrong for us? Your predecessor couldn't explain it to me, and his 'plan'
of 'We'll just have to try harder' told me he had no idea what to do. I hope you can do better."
Touring the Factory in January 2012
John Milliken graduated in 2001 from Rensselaer Polytechnic Institute with a BS in Mechanical
Engineering. For the past five years he had been the Operations Manager of a small packaging firm
serving the northern New Jersey pharmaceuticals industry, so he came to Bayonne familiar with the
general manufacturing processes Bayonne used to design and deliver customized small-unit
packaging. During the hiring process he met most of the management at Bayonne and also the
factory supervisors on both shifts, and had asked for several reports to be prepared for him when he
came to work in January. Digging into the pile, Milliken focused on October 2011 since that was
Bayonne's highest-volume month and, as October 31 closes the fiscal year, it would show him
complete and audited 12-month financial statements for the company. He reviewed the Income
Statement, keeping in mind Bayonne's practice—a common one—of recognizing revenue when it
billed the customer, and it billed when it shipped product. Milliken then turned to a production
report that listed standard setup and run times, as well as scheduled production and standard hours
for October in key work centers (Exhibit 2). A second report showed "good pieces in/out" for the
month (Exhibit 3). The last report presented the daily and cumulative dollar volumes shipped in
October, net of customer returns (Exhibit 4). He also had his own chart showing the usual flow of
orders through the plant's departments (Exhibit 5).
Quality Control
Milliken left his office and crossed the factory floor to the Quality Control office to find QC
Manager Fran Schuler inside. They chatted about the procedures for the start of each shift, then
Milliken asked where the main problems arose.
Schuler told him that quality problems were concentrated in Fold & Glue with either missing
glued lines or excess glue. Schuler showed him a report from October, their worst month of fiscal
2011, indicating that 6% of products were found defective due to glue problems and were scrapped,
with a further 1% of shipped product rejected by the customer due to glue problems. There were also
2
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Bayonne Packaging, Inc. | 4420
some problems in Finishing—a much lower volume department—primarily due to orders shipping
with some or all pieces missing an attachment like a button or zipper. If the customer returned these
they could usually be reworked rather than scrapped.
"What's your staffing, and your procedures for preventing or finding defects?" asked Milliken.
"The supervisor signs off on the first good piece the operator runs," said Schuler. "That goes in the
Work Order Jacket." The Work Order Jacket traveled with the job; it listed the routing, the standard
setup and run times, any special instructions, and ship-to information. It also held the customer's
signed proof along with samples signed by operators, supervisors, and QC at each operation.
Schuler told him that QC had one inspector on each shift covering the Composition, Sheet, Print,
and Die-cut departments, and a second for Fold & Glue, Finishing, and the shipping dock. The
inspectors went from machine to machine checking two pieces every hour during the production run,
and performed a final inspection of material before it shipped out. Schuler said, "In Shipping we
check the product against the proof in the Work Order Jacket, and against any special instructions
like special packing or ship-to. Usually it takes about 15 minutes to do an order. Of course we might
not have the Work Order Jacket at that point if it's been partialed since the jacket gets filed up in the
Production office after the first shipment." If an order was running late it was sometimes possible to
rush a "partial" quantity of the whole order to satisfy part of the customer need, with the rest of the
order completed and shipped later.
Sales Management
Milliken next spoke with VP Sales Alex Wascov and asked him about his biggest concerns. "Ontime delivery," said Wascov. "We're selling a piece of an expensive promotional campaign. The
customer has bought ads, Internet, direct marketing mailings, point-of-purchase display units and
coupons, commission payments to the retail channel so they'll push the product, people handing out
samples on the street—you name it. It's all scheduled to hit, to have impact, on a specific date. If our
product isn't there, the customer goes nuts. We have orders as small as 1,000 pieces, but we sell some
of those for a high price per piece—they're just as important to the customer. In October, since you
ask, we were late more than 20% of the time. Two years ago I hit the roof if it was 5% in a month. The
factory doesn't agree with me on this, they take a bow for being "on time" if they get a partial out. But
that's nowhere near good enough for the customer. You try getting the customer to give us a second
look if we've been late once. I'm telling you, it's getting worse not better, year after year." Milliken
nodded. "The second biggest problem is the boxes popping open with not enough glue or no glue at
all. Sometimes there's too much glue laid on so it bleeds out, it looks bad or you can't open it because
it's stuck so bad. Don't get me wrong—we have a beautiful product, great designs, classy printing.
But the glue problems are real, or sometimes the product is just missing some Finishing piece.
Delivery is the big issue, though."
Milliken asked why customers sometimes wanted to "move up" or expedite a due date to receive
product sooner than they had originally been promised, assuming schedule was part of a coordinated
marketing project. Wascov said that, first, some customers had learned or heard that Bayonne ontime delivery was not to be trusted. Second, the customer may have originally wanted the material
sooner but someone had settled for the standard date, and then later come back to get what they
really wanted. Third, some other component of the marketing project might become available earlier
than anticipated, giving the customer hope that all of the pieces, including Bayonne's, could come
together sooner. Fourth, sometimes the customer was putting together the project for some other
company that wanted more lead-time—for example, it was not uncommon for retail channels to
demand that promotional materials be staged sooner than originally promised.
HARVARD BUSINESS SCHOOL | BRIEFCASES
3
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4420 | Bayonne Packaging, Inc.
Hoping to switch away from this litany of woe, Milliken asked, "Other than the problems, how is
Bayonne doing?" Wascov looked immediately happier. "Great!" he said. "We're grabbing into new
markets, getting aggressive about taking customers we've never had before. This year for the first
time we've had some big, solid hits, like up to two-hundred thousand piece orders in candies, pretty
much a whole new product for us. Same thing with those little corporate gift sets. I told my sales
people to do what it takes, price aggressively to the market, and boy they surely did. Dave Rand
complains about the margins, but he doesn't complain about getting the volume. The designers had
to learn a few new tricks, and the Fold and Glue operators too, plus we got into some FDA
requirements about coatings, adhesives, and liners, but amount of business we're pulling in is great.
If we could just get a track record for delivering on time we can win a lot more customers too,
especially with the kinds of products we haven't touched before, just over the river. You know how
much promotional material money there is in Manhattan?!"
Composition, Sheeting, Printing, and Die-Cut
By mid-afternoon Milliken had worked his way through most of Bayonne's factory work centers.
First he toured the Composition Department, where both printing and package designs were
developed and finalized, printing plates made, and die-cutting dies ordered. From there Milliken
passed through the Sheet Department, where the Jagenburg sheeter turned roll stock into sheets of
paper stacked on skids to be printed. In the Print Department he watched 4- and 6-color Heidelberg
presses print the sheets. He finished this part of his tour in the Die-Cut Department, where Bobst diecutters cut the printed sheets into blanks—the flat cut-out shapes of printed paper ready to be folded
and glued into finished product. He was led through these areas by Sean Quinn, the manager for
these departments. Quinn had a supervisor within each department on both shifts reporting to him.
Watching the assistant sheeter operator expertly maneuver the clamp truck to fetch and stack up
rolls of paper, Milliken asked Quinn how they scheduled the sheeter and the Heidelberg printers,
and how often a lack of stock caused delays.
Quinn told him that he scheduled the sheeter by what needed to be printed in the following day
or two. They rarely stocked out of because they kept enough variety on hand, and their supplier,
International Paper, could restock from their warehouse in Montvale, New Jersey, within a day or
two in almost all cases.
Quinn scheduled printing by due dates, which were either the standard ones originally promised
to the customer (typically three weeks from signed proof, depending on the number of operations
and the size of the order), rush orders sold from the start with shorter-than-standard lead times or,
about twice a day, dates expedited after the order was placed.
In Quinn's office looking out a window at pallets of printed sheets waiting to be die-cut into
blanks, Milliken asked how the Bobst die-cutters were scheduled. "As much as we can by having the
same die," Quinn said. "We sell enough of the same package designs so we can gang orders.2 Jerry,
the cost estimator, knows that when Order Entry routes the orders they give us 30 minutes as a
standard setup time—they know it really takes two or three hours if you have to change over a Bobst
die, but they figure on me ganging. I juggle the orders I have in the department and that generally
allows me to gang six or eight orders on a setup. If I didn't gang, I couldn’t get enough run time to
stay on schedule." Milliken asked how long Quinn usually held an order waiting for more to come in
to the department for ganging. Quinn said it was usually about a week, sometimes as much as two.
2 Orders were "ganged" when they were similar enough in some setup characteristic so that several orders could be run
sequentially after a single setup. In die-cutting, if several orders had the same paper thickness, blank shape, and crease lines,
they could normally all use the same die and be run on a single setup.
4
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Bayonne Packaging, Inc. | 4420
While he talked, Quinn illustrated his scheduling by pointing to a board where small magnetic
clips each held a piece of paper with a Work Order number and its die number. Most were placed
neatly in columns under the two Bobst machine numbers, lined up in the order to be run. Pushed off
to one side of the board was a jumble of clips with orders that had not yet been scheduled to run.
Milliken studied the jumble for a moment. Each order appeared to have a different die number.
Milliken asked Quinn about the schedule that was generated twice weekly by the computerized
scheduling system. "It's useless," Quinn said bluntly. Milliken asked him why. Quinn said, "The
biggest reason being, between rush orders and ganging the orders to keep the machines running, I
couldn't afford time-wise to do what the printout says. Apart from that, what the schedule tells you
is, excuse me, garbage. It says you have an order when you've never seen it, or that you still have an
order that you know you finished and got out of here. That doesn't have to happen too many times
before you just don't worry about the printout."
Fold & Glue
Milliken's next stop was the Fold & Glue Department, where the die-cut blanks were turned into
finished product. The department had four kinds of machines: Two International Queens and one
International Royal, which were high-speed machines but complex to setup; four International
Staudes, which were slow machines but easier to setup, best used for low-volume jobs; and two
International 3A machines which were "window/patch" machines, using rotating cylinders which
attached clear plastic "windows" and also tear-strips on envelope-style products.
Milliken asked department manager Rick Gomes about his problems. Gomes told him, "The
department runs pretty good, but we get a lot of problems that are not our fault. Sean Quinn sends
me a lot of orders that are one or two days from the due date, or even late already. So we have to run
them as soon as possible, but there's already orders ahead of them here lined up. So I can't gang
orders like he can. With those, and the rush orders too, I have to break into runs, get the order done,
and then go back to the first order I broke into—so that makes for partials." "What about the glue
problems?" asked Milliken. "Getting the hot melt glue guns synchronized to the belts and swords3 is
the tough thing about a setup," said Gomes. "If you do the setup quick and dirty, it runs slow—but I
can't run slow with our volume. Usually we get it alright, but sometimes not. Let me show you here
on this Queen." Milliken was familiar with fold & glue machines, but he stood with Gomes looking at
the 35' straight-line machine and nodded while Gomes talked over the noise. "The feed takes the
blank from the stack here we get from Die-cut, lays it down flat, then the fingers grab and move it. At
the first glue gun a line of glue gets laid down, then the swords pick up the tabs, fold them over and
lay them down. They go between the top and bottom belts, which squeeze the fold while the glue
sets, for the first fold. On this order, that happens again at the second guns. Getting it running at
22,000 to 27,000-an-hour like this with the swords, belts and guns synched, not too little, not too
much, and not folding off—it's not easy. The Staudes are quick to setup. They're great for little jobs,
but they run slow, of course."
Milliken looked at a hot melt glue gun closely. The tiny glue orifice in the brass nozzle was open,
with some hardened residue around it. The fitting from the glue canister was tightened on with an
automobile-style hose clamp. He touched the nozzle and it shifted slightly. "How often do you
change the filter?" he asked. Gomes said, "When it's needed - we can tell."
Milliken asked, "How do you decide between the Royal and Queens or the Staudes where to run
orders?" Gomes said, "If we could do what we wanted we'd run orders of 60,000 or more on the
3 "Swords" were long thin smoothly curving metal pieces. When the side or tab on a blank lying flat and traveling down the
length of the machine at high speed touched the sword, the sword guided it up and folded it over along an indented fold line
pressed into the blank by the die-cutting operation.
HARVARD BUSINESS SCHOOL | BRIEFCASES
5
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4420 | Bayonne Packaging, Inc.
Queens or the Royal, and anything less on the Staudes. 'Course, a lot of times we can't do that."
Milliken asked about the shop floor schedule. Gomes said, "Just personally, I don't look at it. By the
time I get orders, I'm just going by what's getting expedited and due dates. Whatever fire is hottest,
I'm putting it out."
They watched a grizzled operator setting up the Royal, banging forcefully on a sword adjustment
with a steel mallet. The part didn't move. The operator swung the mallet harder and the part jumped
about an inch. "What's your maintenance schedule?" asked Milliken. "Well, operators are responsible
for clean-up at the end of shift, and checking oil and grease points at the start. If something breaks
down, Maintenance is good about getting right on it. We've got a pretty good supply of spare parts
there in the cage."
Scheduling
The Finishing Department manager was out sick so Milliken decided to tour there the next day.
Back in the Production office Milliken sat down with Jim Worthen, the plant scheduler, and asked
him about his job.
Worthen told Milliken that his job was to do whatever it took to deliver on time. The plant held a
daily production meeting every morning where the main topic was what had gone wrong the
previous day, the status of large, important orders, and what the department managers would do to
deliver on time any newly expedited orders. Worthen indicated that generally they did not tell Sales
if an order was going to be late. He said, "We're supposed to but a lot of times we don't know for sure
until the end because we're hoping to get a partial out on time. If you tell Sales ahead of time an order
is going to be late, Wascov just starts hammering us and calling up Dave Rand and the rep starts
screaming. It creates a lot of unnecessary work—and then they bump something else, so the ripple
effect is just more jobs late. Of course, they don't blame themselves for that, do they?"
Milliken asked, "When you partial, do you break orders in half or have a little part and a bigger
part usually? And do you break into more than two pieces?" Worthen said, "It's always just the first
and second piece, and they're just about always the same size—it's easier that way to justify doing all
the additional setups if the partials don't have tiny numbers."
Milliken said, "Sean Quinn and Rick Gomes say they can't use the schedule. Why is that?"
Worthen told him that much of the shop floor reporting was missing or wrong, forcing Worthen to
spend several hours each day trying to scrub the data using information on work order pieces
completed in each operation which the operators recorded by hand in the Work Order Jackets.
Milliken asked why the data reported through the shop floor computer terminals was so
inaccurate. Worthen told him that operators recorded pieces completed in the Work Order Jacket and
their start and stop times on setup and run. The Estimator used that data for estimating standard
times, and Payroll got attendance data when the operators swiped their bar-coded ID cards through
the terminals. Milliken knew that operators got a 30-minute meal break in the middle of their 8-hour
shift. Roll stock use at the sheeter, plates in Comp, and ink in Print were recorded by different,
manual systems, and the experienced operators there maintained their own local inventory records to
make sure that they did not run short. As a result, no one felt a real need for accurate terminal-based
shop floor production reporting.
Milliken asked what rules the computer-based scheduling system used. Worthen told him that the
computer recalculated and printed a new schedule twice a week for two "buckets"—the machine
capacity in hours available in each work center in the Monday-Tuesday-Wednesday and the
6
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Bayonne Packaging, Inc. | 4420
Thursday-Friday-Saturday (half-day) buckets. The system added up the standard setup and run
hours for every order released to the plant and "filled the buckets" by scheduling orders by priority.
Priority was set by the "critical ratio". The ratio's numerator was the number of hours between the
calculation date and the due date, and the denominator was the total standard setup and run hours
for all remaining operations plus 48 hours for each remaining operation. The lower the ratio, the
higher the priority. Worthen added, "My rule of thumb is that if the critical ratio when the job starts is
lower than 2, I have to watch that job, it's born in trouble."
While Worthen was explaining this, Milliken sketched his understanding of the rules with an
example (see below). He showed this to Worthen, who nodded and remarked, "Nobody ever got that
before the first time I told him."
Operation
Comp
Sheet
Print
Die-cut
Fold & Glue
Total hours:
Today:
Critical ratio:
Standard setup and run hours
3
2
6
3
8
22 setup and run + (5 ops x 48 hours each) = 262
January 2.
Due date: January 25. Hours: 23 x 24 = 552
552/262 = 2.11
Milliken asked, "Has this extra two days at each work center for scheduling purposes, in addition
to the work time there, got something to do with partials, like the orders really turn into two orders
on their way through?" Worthen said, "No, it's a bunch of things that cause orders to just sit around
and not get worked on; I'm just telling you that in my experience, anything less than a two ratio is
almost always late."
Milliken asked, "Who has authority to expedite orders?" Worthen said, "Well, Dave Rand of
course. Wascov gets on the horn to him, or he would call John McNulty [Milliken’s predecessor]
directly to get orders expedited. Since John's gone, Wascov's been calling me. Quinn and Gomes
complain, and Joe Pensiero in Finishing actually gets it worst of all since he's dead last after any other
delays. But the system actually works pretty good. If an order gets expedited after it's out on the
floor, we give it to Neil Rand. He takes the Work Order right through whatever is left and makes sure
it happens, puts it on a machine right away with no delays, wherever it's routed. Red carpet
treatment." Worthen grinned. "Of course Neil's worked all over the factory since he was a kid—he's
actually a good setup man himself in Fold & Glue. His orders are never late unless we hand it to him
already past due—so of course Wascov is always trying to get expedites authorized." Milliken asked
politely, "Neil is a family member?" "Yeah," said Worthen. "He's Dave's uncle. He wasn't ever really
executive material, but he's a great guy; everybody loves him."
Back in Milliken's office the mid-afternoon January sun was already sinking. Aware that President
Rand wanted his recommendations by the end of the week, Milliken pulled out a fresh pad of paper
and started organizing his thoughts. He had quite a few questions to answer, among them:
1.
Why were delivery, cost, and quality all tanking last October? It seemed like a good month to
analyze because he had fairly complete data for it, and it probably showed Bayonne
operating under its greatest stress.
2.
The plant had seemed relatively busy as he walked through, though he knew that operators
and supervisors tended to look that way while the boss was touring. Still, given Bayonne's
HARVARD BUSINESS SCHOOL | BRIEFCASES
7
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4420 | Bayonne Packaging, Inc.
delivery performance, he wondered if capacity were adequate, or if some or even most of the
work centers would be over-stretched even with better management than he had observed.
How did the quality problems play into this? Were orders being run on the right machines?
Could he quantify the effect of the runs being broken into by the expediting and partialing?
3.
8
What was the practical effect of the informal system Bayonne used to schedule the order in
which jobs were run? The computer system was obviously useless right now, but would it be
worth trying to fix the data and then get adherence to the scheduling system?
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4,595,992
Total COGS
Net Profit Before Tax
(365,694)
469,020
419,383
415,331
834,714
3,285,673
597,671
525,748
116,495
29,884
40,520
COGS:
Shipped Material
(Scrap)
Regular DL
OT DL
IL, supervisory
Other Mfg OH
Gross Profit
Selling Expense
Admin Expense
Total Selling and Admin.
5,140,467
(75,455)
5,065,012
Oct-11
Income Statements ($000)
Gross Sales
Less: Customer Rejects
Net Sales
Exhibit 1
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
-7.2%
9.3%
8.3%
8.2%
16.5%
90.7%
64.9%
11.8%
10.4%
2.3%
0.6%
0.8%
101.5%
-1.5%
100.0%
203,936
741,305
277,494
259,875
537,369
3,663,360
2,704,465
405,229
396,420
74,879
25,107
57,261
4,438,500
(33,834)
4,404,666
Sep-11
4.6%
16.8%
6.3%
5.9%
12.2%
83.2%
61.4%
9.2%
9.0%
1.7%
0.6%
1.3%
100.8%
-0.8%
100.0%
259,618
613,920
167,988
186,314
354,302
2,440,409
1,820,380
268,781
271,835
12,217
15,272
51,924
3,075,860
(21,531)
3,054,329
Aug-11
8.5%
20.1%
5.5%
6.1%
11.6%
79.9%
59.6%
8.8%
8.9%
0.4%
0.5%
1.7%
100.7%
-0.7%
100.0%
3,319,472
6,824,884
1,493,610
2,011,801
3,505,411
23,656,955
17,770,912
2,529,993
2,499,511
173,746
134,120
548,673
30,758,666
(276,828)
30,481,838
Nov-Aug-11
10.9%
22.4%
4.9%
6.6%
11.5%
77.6%
58.3%
8.3%
8.2%
0.6%
0.4%
1.8%
100.9%
-0.9%
100.0%
5,586,250
9,418,505
1,547,642
2,284,614
3,832,256
27,430,107
20,782,617
2,616,251
2,874,192
294,789
125,285
736,972
37,071,038
(222,426)
36,848,612
2010
15.2%
25.6%
4.2%
6.2%
10.4%
74.4%
56.4%
7.1%
7.8%
0.8%
0.3%
2.0%
100.6%
-0.6%
100.0%
5,802,848
9,749,343
1,817,970
2,128,525
3,946,495
25,144,513
19,296,302
2,163,419
2,547,251
265,193
139,575
732,771
35,069,202
(175,346)
34,893,856
2009
16.6%
27.9%
5.2%
6.1%
11.3%
72.1%
55.3%
6.2%
7.3%
0.8%
0.4%
2.1%
100.5%
-0.5%
100.0%
4420 -9-
4420 | Bayonne Packaging, Inc.
Exhibit 2
October Scheduled Orders, Standard Setup and Run Time (minutes)
Standard Timesa
Machine or Work Center
Composition (plates)b
Jagenburg sheeter (sheets)
Heidelberg press (sheets)
Bobst die-cut (sheets)c
Int. Royal/Queen F&G (pieces)d
Int. Staude F&G (pieces)
Int. 3A window/ patch (pieces)e
Finishing (pieces)f
Setup
Run
Number
Mach.
40
20
50
30
180
40
100
30
2
0.0033
0.0083
0.0075
0.0023
0.015
0.011
0.1
1
1
2
2
3
4
2
N/A
Sheets/Pcs
Scheduled
Orders
Sched.
1,438
3,185,032
3,162,737
3,108,971
6,209,329
2,242,039
782,274
687,601
310
310
310
310
77
233
88
28
Standard Hours
Total
Total
Setup Hrs
Run Hrs
207
103
258
155
231
155
147
14
48
175
438
389
238
561
143
1,146
Total Hrs
Per Mach.
255
279
348
272
156
179
145
N/A
Note: October 2011 had 347 scheduled work hours net of breaks, including half-day Saturdays.
a "Standard Times" were estimates of setup and run time per job, derived from times recorded in Work Order Jackets from
previous, similar jobs.
b In Composition, Setup time is per order while Run time is per plate. Approx. 2/3 of the orders were for 4-color (4 plate)
printing; the rest were for 6-color.
c Changing dies took 2–3 hours but the standards assumed that four to six or more different orders using the same die would
be "ganged" or run together on a single setup. Since Cost Estimating could not know—when routing any individual
order—how many orders it might be ganged with at Die-cutting, Estimating simply routed each order for 30 minutes.
d Blanks could run on a Royal/ Queen or a Staude, but no order ran on both.
e 30 of the 3A orders ran on the Royal/Queens, the remaining 58 on the Staudes.
f Finishing was labor-intensive piece-work. Bayonne was able to vary the labor as needed, and so effectively could apply as
much capacity as it needed.
10
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Bayonne Packaging, Inc. | 4420
Exhibit 3
Material Flow Through Operations, October 2011
Machine/ Work Center
Sheet
Print
Die-cut
Royal/Queen
Staude
3A
Finishing
Pieces Ina
Pieces Out
9,555,097
9,488,211
9,326,912
6,209,329
2,242,039
782,274
687,601
9,488,211
9,326,912
9,233,643
5,588,396
2,085,096
768,193
675,335
Total pieces shipped:
8,441,686
a Material scheduled at Sheet, Print, Die-cut has been converted from sheets to pieces, sheets averaged 3 pieces
Exhibit 4
Value of Actual Shipments in October (dollars)
Date
Daily
Cumulative
Orders Shippeda
Late
Partialed
1-Oct
3-Oct
4-Oct
5-Oct
6-Oct
7-Oct
8-Oct
10-Oct
11-Oct
12-Oct
13-Oct
14-Oct
15-Oct
17-Oct
18-Oct
19-Oct
20-Oct
21-Oct
22-Oct
24-Oct
25-Oct
26-Oct
27-Oct
28-Oct
29-Oct
31-Oct
189,380
(3,405)
140,208
3,674
76,914
366,641
7,711
(23,639)
163,777
243,332
56,747
208,154
113,677
134,586
204,803
(5,080)
211,883
456,738
184,201
192,185
191,585
284,708
338,036
222,682
419,952
685,559
189,380
185,976
326,184
329,858
406,772
773,413
781,124
757,485
921,262
1,164,594
1,221,341
1,429,495
1,543,172
1,677,758
1,882,562
1,877,482
2,089,365
2,546,103
2,730,304
2,922,489
3,114,074
3,398,782
3,736,819
3,959,501
4,379,453
5,065,012
15
6
5
7
9
12
11
9
12
9
7
15
12
8
6
15
12
18
14
16
15
19
17
22
28
34
4
2
0
1
1
2
0
1
2
0
0
1
1
0
0
1
0
2
1
3
3
4
5
8
12
19
3
1
1
2
2
3
1
2
1
2
2
4
2
1
1
3
2
4
3
3
4
7
6
12
15
20
353
73
a Counts each partial separately
Total pieces shipped:
107
8,441,686
HARVARD BUSINESS SCHOOL | BRIEFCASES
11
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
4420 | Bayonne Packaging, Inc.
Exhibit 5
Flow of Orders Through Production Department
Quote
Large
To Customer
Composition
Dept.
No
Order
Size?
Royal/
Queen
Staudes
Design
OK?
Yes
3A?
Sheet
Dept.
Print
Dept.
Die Cut
Dept.
12
Small
Yes
3A
No
Finish
Dept.
BRIEFCASES | HARVARD BUSINESS SCHOOL
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Bayonne Packaging, Inc. | 4420
Exhibit 6a Bobst Die-Cut Machine
Exhibit 6b International Queen Fold and Glue Machine
HARVARD BUSINESS SCHOOL | BRIEFCASES
13
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
Understanding How to Use The 5-Whys for Root Cause Analysis
Abstract
Understanding how to use the 5-Whys for Root Cause Analysis. The 5-Why method of root cause
analysis requires you to question how the sequential causes of a failure event arose and identify the
cause-effect failure path. ‘Why’ is asked to find each preceding trigger until we supposedly arrive
at the root cause of the incident. Unfortunately it is easy to arrive at the wrong conclusion. A Why
question can be answered with multiple answers, and unless there is evidence that indicates which
answer is right, you will most likely have the wrong failure path. You can improve your odds of
using the 5-Why method correctly if you adopt some simple rules and practices.
Keywords:
Five Whys, Root Cause Failure Analysis, RCFA, cause-effect tree, Fault Tree
Analysis, FTA
The Five Whys approach to root cause analysis is often used for investigations into equipment
failure events and workplace safety incidents. The apparent simplicity of the 5-Whys leads people
to use it, but its simplicity hides the intricacy in the methodology and people can unwittingly apply
it wrongly. They end up fixing problems that did not cause the failure incident and miss the
problems that led to it. They work on the wrong things, thinking that because they used the 5-Whys
and the questions were answered, they must have found the real root cause.
Description of the 5-Why RCFA Method
The 5-Why method helps to determine the cause-effect relationships in a problem or a failure event.
It can be used whenever the real cause of a problem or situation is not clear. Using the 5-Whys is a
simple way to try solving a stated problem without a large detailed investigation requiring many
resources1. When problems involve human factors this method is the least stressful on participants.
It is one of the simplest investigation tools easily completed without statistical analysis. Also
known as a Why Tree, it is supposedly a simple form of root cause analysis. By repeatedly asking
the question, ‘Why?’ you peel away layers of issues and symptoms that can lead to the root cause.
Most obvious explanations have yet more underlying problems. But it is never certain that you
have found the root cause unless there is real evidence to confirm it.
You start with a statement of the situation and ask why it occurred. You then turn the answer to the
first question into a second Why question. The next answer becomes the third Why question and so
on. By refusing to be satisfied with each answer you increase the odds of finding the underlying
root cause of the event. Though this technique is called ‘5-Whys’, five is a rule of thumb. You
may ask more or less Whys before finding the root of a problem (there is a school of thought that 7
‘whys’ is better; that 5 ‘whys’ is not enough to uncover the real latent truth that initiated the event).
Implied in the Five Whys root cause analysis tool, though not often stated openly, is the use of a
cause and effect tree—known as a Why Tree. The method is also called Fault Tree Analysis. It is
best to build the Why Tree first so that the interactions of causes can be seen. Sometimes only one
cause sets off an event, other times multiple causes are necessary to produce an effect. The Why
Tree for even a simple problem can grow huge, with numerous cause-effect branches.
The 5-Why method uses a Why Table to sequential list the questions and their answers. Table 1 is
an example of a completed 5 Why Table for a late delivery that lost a company an important Client.
1
Some contents for this topic are from the website http://www.isixsigma.com/library
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
1
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
Note how each answer becomes the next question. It is vital that each Why question uses the
previous answer because that creates a clear and irrefutable link between them. Only if questions
and answers are linked is there certainty that an effect was due to the stated cause and thus the
failure path from the event to its root is sure.
5 Why Question Table
Team Members:
Date:
Problem Statement: On your way home from work your car stopped in the middle of the road.
Estimated Total Business-Wide Cost: Taxi fare x 2 = $50, Lost 2 hours pay = $100, Order was late to Customer
because Storemen did not get to work in time to despatch delivery and Customer imposed contract penalty of $25,000,
Lost Customer and all future income from them, estimated to be $2Million in the next 10 years.
Recommended Solution: Carry a credit card to access money when needed.
Latent Issues: Putting all the money into gambling shows lack of personal control and responsibility over money.
Why Questions
1. Why did the car stop?
2. Why did gas run?
3. Why didn't you buy
gas this morning?
4. Why didn't you have
any money?
5. Why did you lose
your money in last
night's poker game?
3W2H Answers
Evidence
(with what, when, where, how, how much)
Because it ran out of gas in a back
street on the way home
Because I didn't put any gas into
the car on my way to work this
morning.
Because I didn't have any money
on me to buy petrol.
Because last night I lost it in a
poker game I played with friends at
my buddy’s house.
Because I am not good at
‘bluffing’ when I don't have a good
poker hand and the other players
jack-up the bets.
Solution
Car stopped at side of
road
Fuel gauge showed
empty
Contact work and get
someone to pick you up
Wallet was empty of
money
Keep a credit card in the
wallet
Poker game is held
every Tuesday night
Stop going to the game
Has lost money in many
other poker games
Go to poker School and
become better at ‘bluffing’
6. Why
Table 1 A 5-Why Analysis Question Table
Build the Why Tree One Cause Level at a Time
Many people start into a 5-Why analysis by using the 5-Why Table. With each Why question they
put in an answer and then ask the next Why question. This question-and-answer tic-tac-toe
continues until everyone agrees the root cause is found. Forgotten is the fact that an event can be
produced by multiple causes and multiple combinations of causes. Using the Why Table alone is
permitted if there is only one cause of every effect listed on the table.
The logical connectivity between events and all their causes can be seen with a Why Tree. Building
a Why Tree gives you a good chance of spotting all the issues that could have been in play prior to a
failure event. By only asking Why questions without the Why Tree to guide you, you may never
find all the real root causes. Questions can always be answered, but that does not mean that the
answer is right, or that all necessary causes of the problem are identified. It is unrealistic to do a 5Why analysis by only completing a Five-Why Table of questions and thereby expect to arrive at the
real root cause just because the questions were answered. First you must draw-up the Why Tree
one level at a time and ask the 5-Why question for each level to find the real failure path through
that level of causes.
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
2
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
Car stalled at intersection
and would not restart
Engine stopped
Fuel supply failure
Ran out
of fuel
Lost fuel
flow
Tank
empty
Failed fuel
pump
Holed fuel
line
Ignition failure
Contamin
ated fuel
Fuel/air
over-rich
Water in
fuel
Air supply
restricted
Excess
fuel in
mixture
Fuel supply
restricted
Air filter
blocked
Excess fuel
injected
Fuel filter
blocked
Fuel line
blockage
Lost compression
No spark in
cylinder
Spark plugs
not arcing
Fuel setting
changed
Settings
altered by
someone
Distributor
leads
damaged
Points set
wrong
Pistons
holed
Distributor
not working
Distributor
damaged
Leads
loose
Electrical
supply fault
Water in
distributor
Distributor
not sealing
Distributor
cover open
Water over
distributor
Distributor
body
cracked
Hosed
engine
down
Figure 1 Partial Why Tree of Passenger Car Engine Failure
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
Valve
timing error
3
Drove thru
deep water
puddle
Cylinder
head loose
Piston rings
broken
Spark plugs
unscrewed
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
Figure 1 shows a partial Why Tree for a stalled car (the complete Why Tree would be a monster).
The analysis team uses their collective experience and knowledge of the causes of a stalled car to
logically develop the first level of the Why Tree. Once we identify all the possible first level causes
of the problem we then ask the Why question to find the real first level cause.
A passenger car can stall from electrical system failure, OR from fuel system failure, OR by loss of
engine compression. Each cause is presumed independent of the other and so the connections to the
top failure event are known as OR gates. For each of the three causes we can set off and develop
the next layer of causes. After the second level of causes we can image third level causes for each
one of them. After the third level we build the fourth level, and on and on and on we can rush until
there are many branches, with dozens of boxes in our Why Tree of the causes and effects. If you
did that you will have wasted a lot of people’s time, thrown away your business’ money, and almost
certainly you have the wrong root cause.
It is wasteful of time and people to build an entire Why Tree of a failure incident in the first meeting
unless all the evidence is known down to the true root cause. Figure 1 shows the top of a Why Tree
with three branches going towards possible root causes. Two of those branches will prove to be
unnecessary and their development is pointless. At each level the true failure path should be
identified by the evidence and the other possibilities eliminated. There is no value spending time in
a 5-Why failure analysis developing branches that did not cause the top failure event. (If you were
conducting a risk analysis, and not a failure analysis, you would develop all the branches.)
The approach to take with a 5-Why root cause analysis is to start the Why Tree with the top failure
event and identify all first level causes. Use the evidence and logic to prove which one(s) brought
about the incident. Once the first level cause(s) are confirmed you tackle level two causes and
confirm which of them produced the level one effects, and so on. In Figure 2 the first Why
question to ask is, ‘Why did the car stall?’ The answer is the engine stopped working. The
believable evidence is that the engine would not restart. The level two question becomes, ‘Why did
the engine stop working?’ A critical component failure in any of the three systems that allow an
internal combustion engine to work—electrical, fuel and mechanical—will stop the engine.
Fuel supply failure
Car stalled at intersection
and would not restart
Failure
Event
Engine stopped
First Level
Causes
Ignition failure
Lost compression
Second Level
Causes
Figure 2 First and Second Level Failure Causes
At level two there are three reasons as to why the engine did not work. You must not ask a third
level Why question until you know the right answer to the level two question. Your 5-Why
analysis must stop here pending sure evidence as to which path the root cause belongs to. Most
people using the 5-Why method will expect the team conducting the Five Why analysis to
collectively select the cause of the engine stoppage. But there are three possible paths to take, only
one of which is the right one (presuming that there was no interaction between systems in causing
the failure event). If you accept one of the level two causes as an answer without having sure
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
4
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
evidence you are guessing. If you guess, I have no sympathy for what later happens to you and the
analysis. It is just plain wrong to accept any answer to the Why question—only with real evidence
or impeccable logic can you know the true cause. If you have sure evidence then you and the
analysis team will know the right answer to the level two question. If you do not have true evidence
you will direct people on the team to go and investigate the event and come to the next meeting
with hard facts so that the true failure path can be identified.
Car stalled at intersection
and would not restart
Failure
Event
Engine stopped
First Level
Causes
Fuel supply failure
Ignition failure
Second Level
Causes
Lost compression
No spark in
cylinder
Spark plugs
not arcing
Distributor
leads
damaged
Points set
wrong
Third Level
Causes
Distributor
not working
Distributor
damaged
Leads
loose
Electrical
supply fault
Fourth Level
Causes
Fifth Level
Causes
Water in
distributor
Distributor
not sealing
Distributor
cover open
Distributor
body
cracked
Water over
distributor
Hosed
engine
down
Drove thru
deep water
puddle
Sixth Level
Causes
Seventh
Level Causes
Figure 3 Levels of Failure Causes
You use the evidence as the proof test to confirm the cause(s) for every level in the Why Tree. The
evidence alone confirms the path to follow. Impeccable logic that withstands scientific scrutiny can
also be used to identify the failure path. As you work your way down the cause and effect Why
Tree you accept only the answers that are proven by sure, true evidence and/or unquestionable
scientific logic. It is evidence and/or clear logic that decides the path to take, not someone’s
opinion. We fill-in the Why Table as each cause level event(s) is confirmed. In Figure 3 the Why
Tree has gone down to a 7-Why level. There are many questions to be asked and answers to be
proven. If you do not have true answers for each level, immediately stop the analysis and send the
team out to investigate and find the facts. It is only with accurate hard evidence that the real causes
and circumstances are certain. With real evidence and sound logic you know the causes are true.
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
5
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
Select Your Starting Question From Well Up the Why Tree
If you select the top failure event too low down the Why Tree you may not find the true root cause.
Had the top failure event been the engine would not go, and we asked the first Why question as,
‘Why did the engine stop working?’ you would have made an assumption that the car stalled
because of an engine problem. You can also stall a car by running into the back of the car in front
of you and damaging the engine.
It is vital to start high up the Why Tree when you ask the first Why question. Figure 4 shows by
setting the top failure event as the stalled car, and asking the first question as, ‘Why did the car
stall?’ it ensured that all other causes of stalled car engines were eliminated by the evidence. There
was no car accident, so logically the stalled engine had to be due to a problem with the engine itself.
It is better to start well up the Why Tree and ask a few unnecessary questions that are easily
answered, than start too far down and totally miss the real cause and effect path of the incident.
Car accident
Fuel supply failure
Car stalled at intersection
and would not restart
Failure
Event
Engine stopped
First Level
Causes
Ignition failure
Lost compression
Second Level
Causes
Figure 4 Importance of Starting with the Right Why Question
How to Handle AND Gates in a 5-Why Analysis
Many failure incidents require multiple causes to happen together to trigger the next level event. A
fire needs fuel, oxygen and an ignition source, all three must be present simultaneously. On a Why
Tree for a fire you will always have the configuration shown in Figure 5, with Fuel AND Oxygen
AND Ignition present. In the Why Tree the three must be shown acting together to cause the higher
effect—they produce an AND situation gate.
Failure
Event
FIRE
Fuel supply
Ignition source
Oxygen supply
First Level
Causes
Figure 5 AND Gate for a Fire
In Figure 6 there are two necessary joint causes of ‘water in distributor’ which form an AND gate—
both must have happened for the next higher level event to occur. For water to be in a distributor
there must have been an opening of some type and there must have been water flowing over the
distributor body to go into the distributor. Unfortunately, a standard 5-Why question table does not
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
6
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
let you recognise when AND gates exist. Since an answer is used to create the next question they
can link just one cause to each effect. Why Tables used alone will make you follow one causeeffect path even when AND situations arise. For the case in Figure 6 two causes exist—
‘Distributor not sealing’ AND ‘Water over distributor’
Car stalled at intersection
and would not restart
Failure
Event
Engine stopped
First Level
Causes
Fuel supply failure
Ignition failure
Second Level
Causes
Lost compression
No spark in
cylinder
Spark plugs
not arcing
Distributor
leads
damaged
Points set
wrong
Third Level
Causes
Distributor
not working
Distributor
damaged
Leads
loose
Electrical
supply fault
Fourth Level
Causes
Distributor
not sealing
Distributor
cover open
Distributor
body
cracked
Fifth Level
Causes
Water in
distributor
Water over
distributor
Distributor
cap
cracked
Hosed
engine
down
Drove thru
deep water
puddle
Sixth Level
Causes
Seventh
Level Causes
Figure 6 AND Gate in a Why Tree
You can never know if there is only a single cause for an effect until you first develop the whole
sub-cause and sub-effect tree for the effect under consideration. Where there are two or more
causes acting together to produce an effect each individual cause will need its own Why Tree to
identify and prove all sub-causes and effects leading to the higher effect being investigated. People
will claim in AND situations that because the higher effect requires two or more causes to arise, like
a fire needs oxygen, fuel and ignition, you only need to remove one of the causes to stop the effect.
This thus justifies using the Why Table alone since even if it only follows one cause-effect path any
missed AND situations are still prevented from arising because a lower root cause is removed. The
problem is your analysis is illogical, superficial, incomplete, and the proposed solution is unlikely
to be effective in the long term when you only use a 5-Why Table. Having a Why Tree for each
leg in AND scenarios is the correct practice to adopt. It helps you to make the best business
decision because you find the most effective, least cost sub-cause and effect leg to prevent.
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
7
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
It is vital you always build a cause and effect tree of your failure incidents. Only a tree can show
where events branched and if single or multiple causes had to be present to produce an effect. If you
only use the Why Table without a Why Tree to direct you may not even suspect there is an AND
situation. You are very likely to generate a seemingly sensible answer and write it down without
realising an AND situation was present. It is only when the answer to a Why question in the Table
cannot fully explain the occurrence of an event that you will wonder what is going on. If you are
lucky you will become uncomfortable because the logic is clearly incorrect or incomplete. It is then
vital to stop and build the full Why Tree to see if there is an AND situation. You then use a 5-Why
Table for each branch to confirm the evidence proves your logic is right.
If you have an AND gate in your Why Tree each cause-effect path ought to be separated into its
own Why Table (or on the one Why Table you give each cause-effect path its own set of rows and
identify the rows, such as with colour coding or shading, to differentiate them from the other causeeffect paths). For the case in Figure 6 you could start two new Why Tables (or separate sets of table
rows)—one for ‘Distributor not sealing’ and another for ‘Water over distributor’—to fully
understand what happened to expose an unsealed electrical system to water.
Going to the True Root of Failure—Identifying Latent Causes
In Table 1 you saw a Why Table of a delivery failure caused by a storeman not getting to work on
time. It turned out that he had lost all his money in a card game and had none to buy petrol to get to
work. Because he did not turn up a delivery was missed and the company was penalised $25,000
and lost years of future business. Though there are several possible answers listed in the Why Table
to address the top problem caused by not having money, we have stopped short of asking the most
important question of all—Why did he gamble all the money he had? This person had a low paying
job, he had a family to support, he had debts to pay, yet he bet all his money in the game. If we
were to do a Why Tree on this man’s poor betting decision it would look like that of Figure 7.
No fuel in car
No money to buy
fuel
Lost all money in
card game
Wants to be with
friends
Bet all his money
Cannot play poker
well
Has a limited income
‘Easy come easy go’
attitude
Lives from day to
day
Personal Values
and Attitudes
Latent
Causes
Figure 7 Why Tree of Late Delivery Latent Causes
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
8
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
We have not fixed the real root cause of the failure—this storeman’s gambling problem. The
gambling is a latent cause. It remains there waiting to trigger more troubles in future. This
company will have many problems with this storeman because his view of life is ‘easy-come-easygo’ He is his own worst enemy and by default he is his employer’s worst enemy as well.
The values and beliefs of this storeman were what actually caused the failure event that cost his
employer $25,000 penalty and $2M lost future business. If we did not know that the man was a
problem gambler you could not prevent the root cause. In fact, if we did what was suggested in the
Why Table, that he carry a credit card in case he runs out of money, we would put him at high risk
of getting into serious gambling debt. His illness would make him max-out his credit card. Once
we learn of the latent cause we are able to make a more effective choice that reduces the chance of a
repeat problem (have him carry a mobile phone to ring someone to get him to work).
As the manager of this man you are in a dilemma. Clearly he is a huge risk. Should you dismiss
him or put him in a lesser important job. Even in a lesser position this man’s attitudes and value
will always expose the company to trouble. The latent causes are the most critical ones to find, but
they are usually the hardest to deal with and often they are embarrassing for all involved.
Figure 8 The Progression of Failure Incidents and Events
Until you know the latent causes that triggered a failure event, and you stop or prevent them, you
have not addressed the real root cause. Our experience with failure events around the world and
across many decades has highlighted the existence of a causal sequence of circumstances that exist
when failures occur. Figure 8 shows how an incident can first be traced to its scientific Physics of
Failure factors, going back further we can see situational events relating to the circumstances
present at the time. Further back we can spot business process failures that allowed unwanted
situations to arise. Finally we get into the region of latency where individual beliefs, values, lack of
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
9
Phone:
Fax:
Email:
Website:
+61 (0) 402 731 563
+61 (8) 9457 8642
info@lifetime-reliability.com
www.lifetime-reliability.com
The masters of plant and equipment reliability improvement
vital knowledge and personal opinions breach business system protection. A failure event should
be traced back to its latent factors.
Conclusion
The 5-Why root cause analysis method is simple in concept but requires real evidence, sure logic
and great discipline in its use if you want to find the true root cause of a failure event or problem.
You can only complete a 5-Why Table to the true root cause if you first have the correct Why Tree
for the occurrence. There are many incidents and events that can cause the top level failure and
during the investigation you must find all cause and effect branches to the root cause(s). If you go
the wrong direction you will fix the wrong thing and leave the true root causes behind. The missed
causes will sit in your business awaiting the next opportunity to bring you more strife and trouble.
You must keep all the evidence when a failure happens. If you do not have real, honest evidence
then stop the analysis and go find factual proof. It is pointless to go further since all you do will be
speculation, opinion and guesswork. If you get the right root cause it will be entirely due to good
luck—hardly acceptable and adequate for sound decision making. Make it a company policy—an
unbreakable law—to always keep all the evidence safe. With complete, true evidence and facts you
can uncover the whole story down to its real roots and cut them out.
My best regards to you,
Mike Sondalini
www.lifetime-reliability.com
C:\Users\user\Documents\LRS\LRS Methodology\Understanding How to Use The 5Whys for Root Cause Analysis.docx
10
UV0833
OPPORTUNITY CONSULTANTS, INC., 2007
The lackluster reputation of the student club Opportunity Consultants, Inc., (OCI) was
beginning to concern the administration of the Darden Graduate Business School. The poor
quality of the work produced by the club, which provided pro bono consulting services to local
small businesses and nonprofits in Charlottesville, Virginia, was negatively impacting the
school’s brand within the community. Unless rapid and sustainable initiatives to improve client
satisfaction were implemented, the club was in danger of being disbanded. To avoid that
undesirable result, OCI leadership enlisted the support of a select group of second-year students
to help the club improve its performance.
The student “turnaround” team conducted multiple informal interviews with club
members and analyzed the processes driving client acquisition, project selection, member
recruiting, and project staffing, as well as the overall dynamics of project teams. The work of the
turnaround team revealed several factors that it believed limited productivity and interfered with
the quality of the work of the project teams. Based on these findings, the team proposed several
initiatives that it thought would significantly improve client satisfaction and begin to turn OCI’s
reputation around during the next year.
Proposed Initiatives
Client acquisition and project selection
While advertising for clients drew a sizable number of applications, there were several
drawbacks. For one, the year-to-year response rate was unpredictable. In addition, there was a
frequent disconnect between the expectations of a potential client and the resources of the club.
Many applications were rejected outright because of scope and feasibility, while others were
discarded when it was obvious that the potential clients viewed OCI as a source of cheap,
“worker-bee” labor. Increased advertising improved the number but not necessarily the quality of
potential projects. Currently, OCI had accumulated $2,200 in its treasury. Next year’s treasury
would consist of the money not spent this year plus the income from completed projects
multiplied by the standard $300 fee; there was talk of changing the fee.
This case was prepared by R. William Reynolds (MBA ’07), and Robert D. Landel, the Henry E. McWane Professor
of Business Administration. It was written as a basis for class discussion rather than to illustrate effective or
ineffective handling of an administrative situation. Copyright © 2007 by the University of Virginia Darden School
Foundation, Charlottesville, VA. All rights reserved. To order copies, send an e-mail to
sales@dardenbusinesspublishing.com. No part of this publication may be reproduced, stored in a retrieval system,
used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying,
recording, or otherwise—without the permission of the Darden School Foundation.
This document is authorized for use only in Laureate Education, Inc.'s CMBA SP004-Systems Thinking for Organizational Improvement course at Laureate Education - Baltimore, from
September 2017 to November 2018.
-2-
UV0833
The club credited a number of its applications to personal referrals from satisfied clients
in the community and to its reputation. One referral from a satisfied influential client would often
lead to numerous good referrals. Unfortunately, the reverse was also true, and dissatisfied clients
had discouraged their peers from utilizing the club. Compared to the rapid turnaround in the
performance of business clients served by OCI, improvement in the club’s reputation would take
longer by several years as many former business...
Purchase answer to see full
attachment