Article in asq referenced magazine quality progress the human-element-in-a-bpo

Operational excellence and client satisfaction are unsustainable without employee satisfaction. This article details the fundamental building blocks for gaining employee engagement and achieving

employee satisfaction together with operational excellence and customer delight through lessons from the Toyota Production System’s (TPS) cell structure, single-piece flow and stop-correct-proceed concepts woven with lessons from the theory of cognitive dissonance, the theory of systems thinking and the Hawthorne effect applied within a lean Six Sigma framework.

HCL Technologies, a global IT business process outsourcing (BPO) services company headquartered in Noida, Uttar Pradesh, India, worked with a client, an innovative global manufacturing conglomerate. Before partnering with HCL, the client’s global order management process had a turnaround time (TAT) of about 24 hours and a quality score of 95%. HCL accepted a service level agreement (SLA) at 99% quality in order entry (OE) and a significantly reduced TAT of six hours. Though this was initially achieved through best practices, it could not be sustained because employees interpreted the system as being intrusive, and it reduced their job satisfaction.

This posed a challenge in aligning good people with best practices. This case study shows how the systems theory of sociology, theories of

psychology—namely, the Hawthorne effect and cognitive dissonance theory—and TPS production engineering helped us:

• Find the root cause of the actual problem.• Define a strategic way forward.• Implement the holistic solution that resulted in employee engage-

ment.This approach improved employee satisfaction and, at the same time,

reduced defects and TAT in a sustained manner, and paved the way for exceeding client expectations and delighting them during a two-year time frame. The result was a huge improvement in the visibility, velocity and reliability of the client’s global supply chain with bottom-line benefits of $10 million in annual savings. The case study was performed in the lean Six Sigma framework.

Client overview and relationship with HCL

The client is a $23 billion company with local subsidiaries worldwide. It also operates a division called global channel services (GCS) in which global customers can place orders with the division directly or through local subsidiaries. GCS also helps integrate new acquisitions. It has a turnover of $300 million. GCS outsources work to HCL.

HCL has been a BPO partner with GCS since Aug. 27, 2008, providing

C ASE STUDY

The Human Element

APPLYING

PSYCHOLOGICAL

AND PRODUCTION

THEORIES HELPED

STREAMLINE

BACK-OFFICE

OPERATIONS

AND ACHIEVE

OPERATIONAL

EXCELLENCE

By Balasubramanian

Viswanathan,

HCL Technologies

S I X S I G M A F O R U M M A G A Z I N E I A U G U S T 2 0 1 5 I 9

offshore support (24 hours, five days a week) for its end-to-end order management and shipment-tracking process. This case study covers only the order manage-ment aspect—that is, converting a customer’s purchase order into an internal sales order placed in the system to commence delivery (see Figure 1).

The process works as follows:Customers worldwide have contracts with the cli-

ent for purchases. For every customer, the client has policies that decide the business rules for the supply. Contract and policy details are captured in customer notes for reference. For regular customers, pricing also is captured in the master pricing database (MPDB).

Occasionally, customers are given quotes for specific orders in which the pricing and terms of trade may be different. This is stored in a database called link order to quotes (LOQS). One-time changes also are made through mail.

Thus, there are four different data sources: customer notes, MPDB, LOQS and mail.

These customers continually place purchase orders with the client. Purchase orders come in as emails, e-fax

or electronically. They are immediately acknowledged and verified for completeness and accuracy. They are matched with these four data sources to establish con-gruence on all parameters, namely: customer notes (contracts and policies), MPDB, LOQS and mail.

In cases when this verification shows areas that are inconsistent, ambiguous or incomplete, it is escalated to the client and followed up on until the client pro-vides clarification. Next, the orders are converted into sales orders and logged in the client enterprise resource planning (ERP) system. A final acknowledge-ment is sent to the customer with a promised date of delivery or dispatch. The ERP system moves the orders to the proper factory or warehouse to handle the supplies.

Table 1 provides details of the process scope.

Figure 1. Learnings from operational excellence considering the human element in a BPO

99% of orders are error-free92%

05

1015202530

Huge improvement in quality

Wk

1

Wk

29

Wk

19

Wk

9

Wk

51

Wk

41

Wk

31

Wk

21

Wk

11

0

2:24

4:48

7:12

9:36

12:00

14:24

Wk

1

Wk

30

Wk

27

Wk

24

Wk

21

Wk

18

Wk

15

Wk

12

Wk

9

Wk

6

Wk

3

Wk

52

Wk

49

Wk

46

Wk

43

Wk

40

Wk

37

Wk

34

Wk

31

Wk

28

Wk

25

Wk

22

Wk

19

Wk

16

Wk

13

Wk

10

Wk

7

Wk

4

4 hours9 hours

Huge improvement in turnaround time

Jobsatisfaction

Clientsatisfaction

Operationalexcellence

Unsustainable

Jobsatisfaction

Clientsatisfaction

Operationalexcellence

Standard best practices+

Learnings from other disciplines

Standard best practices

Sustainable

Nonvalue-added time (NVAT)Turnaround time (TAT)

Def

ectiv

e or

ders

$10 million annual savings to client

13 times the annual contract value

How

to

mak

e yo

ung

empl

oyee

s ap

prec

iate

bes

t pr

acti

ces

Our innovation—convert best practices into roles

Toyota production

system

Cognitive dissonance

theory

Hawthorne effect

Theory of systems thinking

Guideposts

Clie

nt p

robl

em—

deliv

ery

not

mee

ting

ser

vice

leve

l agr

eem

ents

Best prac-tice

“We do not like best practices”

Best prac-tice

Best prac-tice

Best prac-tice

• Scorecard • Visibility to quality • Pride in quality

• Peer check

• Validation

Standard solutions

• Manage updates • Training

BPO = business process outsourcing

Table 1. Process scope

CustomersProducts

(SKU)Factories

Markets (countries)

OrdersLine

items

2,500 100,000 180 200 12,000 50,000

10 I A U G U S T 2 0 1 5 I W W W . A S Q . O R G

The Human Element

Many customers request purchases just once or twice a year, leaving little room for specialization. Clients manage this with geographically specialized customer account analysts (CAA), who have a huge amount of tacit knowledge to manage this complexity. Transitioning to our BPO here, we lacked this skill and knowledge as a team, and that created a huge challenge to quality and TAT.

Order entry complexities

The outsourced task was the OE transaction. We discovered there were two complexities in managing quality in OE:

1. Verifying the large number of parameters per line item in the order. An unambiguous sales order contains 40 parameters for just a single line item. For instance, a simple order—Office Depot orders for a box of stationery for our client—will still require an entry of 40 parameters: 19 of the 40 are fatal param-eters; that is, an error in any of them can make the order defective.

The 40 parameters are included in Figure 2. Thirteen of them are available in the purchase order,

and the remaining 27 must be painstakingly gleaned from client masters lists and contracts.

A standard order has four line items. Given the split of 40 parameters across the header (one per order) and line (as many lines as there are in the order), an average order will have 65 fatal parameters, and any one of them—if wrongly captured—makes the full order erroneous.

The client expects 99% quality, which translates to one error out of 100 orders or 6,500 parameters. The defects per million opportunities (DPMO) is 1/6,500, or 154. In other words, the short-term sigma expected is 5.11.

2. Capturing the parameters on a legacy system. Legacy systems are those systems without graphic user interface (GUI), drop-down boxes and standard mas-ters as tabulated databases. Data must be typed in or cut and pasted, which can result in frequent typo and oversight errors.

Before offshoring, the scenario was as follows: The client had CAAs dedicated to customers from different parts of the world. They received orders in the form of emails, e-fax or electronically. They entered the orders into the system in real time. CAAs worked only eight

Figure 2. Process—converting a purchase order into a sales order

Customers Orders Line items

2,500 12,000 50,000

100,000 products

200+ factories

160+ countries

Product Customer Logistics Price Date Basic

From

pu

rcha

se o

rder

Product Ship to Price Purchase order datePurchase order number

Buyer code Sold to Income terms Request date

Quantity Charge to Payment terms Cancel date

From

cl

ient

mas

ters

, co

ntra

cts

Seller code Facility Ship complete Received date Account code

Unit of measurement Ship-to code Route Chip exact Entered date Desk code

Conversion factor Sold-to code ViaSpecial instruc-tions

Promise date Sales order number

Units in inner pack Charge-to code Export marks Expedite date Purchase order type

Inner pack in outer packDomestic/delivery agent

Contact number Expedite flag

Pack type Future date

Load slip

Purchase order to sales order

Manufacturing

Sales order

SalesPurchase

order

Customer

40 parameters

in an unambiguous sales order

Legacy system

No graphic user interface


The Human Element

hours each day and placed orders in the system, but customers sent orders around the clock. Thus, placing sales orders for purchase orders that arrived after the CAAs finished their eight-hour shift was postponed to the next day. Thus, the TAT was about 12 hours. The system reportedly had a 95% quality rate.

Off-shoring process goals were a TAT to be managed within six hours from the current level of 12 hours, and improving quality to 99% from the current level of 95%.

Journey: phase one

Ad-hoc process: During phase one, which was imme-diately after the outsourcing transition, we performed the same OE and added a quality check. Quality was just 92% and TAT was seven hours. The client was growing increasingly dissatisfied, shown by two con-secutive client satisfaction surveys from the business process manager.

In February 2008, a survey based on seven weeks of performance had the following results:

• Quality—neutral rating. • TAT—satisfied rating. • Comments—“Quality continues to be a challenge

on OE (primarily due to implementation of the more complex OE system).”

In July 2008, a survey based on 19 weeks of perfor-mance had the following results:

• Quality—dissatisfied rating. • TAT—neutral rating.• Comments—“Continue to see too much vari-

ability from agent to agent and order to order on accuracy. Basic attention-to-detail errors on OE and shipment tracking occur too frequently. Seems to be a lack of knowledge transfer across the BPO team.”

Journey: phase two

Managed solution with best practices: During phase two, the root cause of the problems in meeting the goals were identified and solved with standard best practices, but staying at the level after the goals were

Figure 3. Diagnostics—basic process augmented with best practices

Client

Self

Order entry

system

Managing updates as a process

Validate

Order management process

Visibilityto quality

Pride inquality

Executivescorecard

Training

Peercheck

Client

Incomplete

Type errors

Oversight error

Corrected

Clientmaster

andcontracts

Four defects until now

Update 1Update 2Update 3Update 4

Update

Update

Update

40 parameters

Clari�ed

Perfect


The Human Element

achieved could not be sustained. Quality had improved to 98%, but dropped to 94% again. The TAT increased to 15 hours.

We diagnosed the problem in three categories: pro-cess, people and attitude, and then we identified root causes and found solutions. Figure 3 illustrates the findings, which included:

• Incomplete orders with less than 13 parameters. The young executive would already have pro-cessed hundreds of such orders for a given end customer. Thus, he or she can easily judge the missing information and complete the order. Usually, he or she is right. When the executive becomes more confident, he or she can judge whenever he or she finds an incomplete order. But there’s a problem: While he or she is correct 90% of the time, the process should be correct 99% of the time, and therefore, we will miss the SLAs. The solution lies in following the best prac-tice called validation—a process of checking an order for full information and escalating it to the client if the order is incomplete so a complete order can be formally entered in the process. Validation preempts judgment errors.

• Typos and oversight errors are a problem with legacy non-GUI systems. The solution is to have

a peer-check done after OE, or create peer cou-ples—each one to check the other’s entry.

• Slow knowledge transfer. A problem to be addressed through training.

• “Where am I going wrong?” A problem to be addressed through individual executive scorecards.

• Missing updates. A problem to be addressed through a disciplined process of managing updates.

• No visibility to quality. A problem to be addressed through real-time visibility with whiteboards. Unlike TAT, this is easily visible through the accu-mulated backlog.

• Quality is a mere metric. An attitude problem to be addressed through evangelizing concepts such as zero defects, thereby making quality a matter of pride.

Basically, the solution lies in sandwiching the OE process between two best practices: validation and peer check. Practically, an executive would perform a validation before OE, enter the order and get it peer checked before he or she moves it forward for the mandatory quality check. Quality scores are displayed prominently in the shop floor in real time. Executive performance analyses, followed by training and online testing, have been formalized. Databases were to be

Best practices actually reduced job satisfaction

Figure 4. Voice of employees 2008—“Why wasn’t performance sustained?”

Validate

Peer check

Peer couple

Training and online test

Executive scorecard

Executive error history

Quality error drive

Team building

Consultant update

MPDB update

2008 initiatives

Employee reaction to the initiative— without considering the human element

“Questions my optimism and confidence.”

“Hurts my ego.”

“Too much policing.”

“Why learn what I already know?”

“I am measured.”

“Low-impact knowledge.”

“Good, but not great.”

“Not all are working.”

“Practical, but partial.”

5 whys analysis

MPDB = master pricing database


The Human Element

The Human Element


updated as a separate process. Quality is evangelized as a matter of pride through zero defect training.

Over 20 weeks, these solutions were implemented sequentially as 10 initiative ideas—one additional new initiative launched every two weeks—giving time and space for executives to accept and adapt to change.

As a result of this, performance improved. Errors declined and quality improved to 99%-plus. The client was delighted. In fact, we won the client’s internal operational excellence award in 2008.

But TAT worsened from seven to 15 hours. The internal operational excellence group (OEG) warned us of client dissatisfaction again. “You are sitting on a volcano of missed TAT,” warned the OEG committee during our final OEG presentation while they handed us the internal award. The client advised us to focus on TAT. Defects started increasing as well. Quality scores dipped to 97% and the client became concerned about the dual problem: poor quality and poor TAT. We realized that the solutions were unsustainable.

Journey: phase three

Optimized solution—operational excellence with a human element: During phase three, the sustainability problem was root caused at a deeper level and solved by applying theories drawn from psychology and pro-duction science. In the end, quality was more than 99% and TAT was less than four hours, and both were sustained.

Theories used in the root cause analysis included Elton Mayo’s Hawthorne effect, which explains the reversal phenomenon. The theory emerged from an early 20th century shop floor experiment on team behavior performed at the Western Electric Plant in Hawthorne, IL.

For a production-related study, Mayo performed the experiment to identify the optimal lighting intensity to maximize productivity in electrical assembly work with glass and metal, which required much attention to detail. The hypothesis was that there existed an opti-

Figure 5. The breakthrough thought

Clari�ed

Perfect

Self

Orderentry

system

Validate


Order entry

Peercheck

Client

Incomplete

Planned

Current

Type errors

Oversight error

Corrected

Clientmaster

andcontracts

40 parameters

Clari�ed

Perfect

Role: Order entry2

1

2

3

Role: Peer check3Role:

Validator1

The Human Element


mal lighting intensity beyond which the glare would affect productivity. The lack of light would have a similar negative effect on productivity, as well.

A set of two tables with 20 workers was selected from the shop floor, which contained a total of 100 tables and 1,000 workers. The light intensity was slowly increased for these two tables by adding bulbs over the tables. As a result, pro-ductivity went up. More bulbs were added, and productivity increased. This continued until the glare was so great that it seemed to be an obvious disad-vantage, but productivity contin-ued to rise.

Unable to increase the num-ber of bulbs, Mayo decided to do the reverse: He removed bulbs, one by one, and slowly reduced the lighting. But he observed that pro-ductivity went up further. After he brought the light intensity to such low levels, it was difficult to see. Still, productivity increased.

The results showed continuous improvement in productivity with every unit of increase and decrease in light intensity. Mayo observed that it wasn’t the light making a difference. The 20 workers at the two tables were being closely watched by the other 980 workers, and this attention made these 20 workers give their

best, irrespective of the lighting. “Attention (on the team being experimented on)

improves performance (irrespective of all other factors like light intensity),” Mayo concluded. He also realized that once attention is withdrawn, performance lowers.

While performing root cause analysis of the prob-lems in our own operation excellence process, we real-ized that a similar situation was seen in our operational excellence drive. When the new process is unveiled and everyone—including the client, sales groups, quality groups and their own operational heads—are

Figure 6. Balanced workflow


Now, these roles must be linked

Now, the cell can be balanced

Validate10 minutes

Order entry30 minutes

Peercheck

10 minutesCell

Role: Order entry2 Role:

Peer check3Role: Validator1

Validate Order entry Peercheck

Order entry

Order entry

Balancedcell

Figure 7. Toyota Production System principles

Cell System—From process to rolesAfter

Before

ValidateMailbox Order entry Peercheck

Peercheck

Peercheck

Peercheck

Peercheck

Qualitycheck Monitor

Order entry

Order entry

MailboxValidate

Qualitycheck

Validate

Validate

Validate

Order entry

Order entry

Order entry

Order entry

watching the team perform, there was immense atten-tion placed on the team, and each team member fol-lowed best practices.

After the client confirmed its delight over the per-formance, success was declared and recognition given, attention was withdrawn and the team no longer used the same best practices. This resulted in a reversal of performance back to original levels. In other words, the Hawthorne effect penetrated our process.

What we learned from this was that we did not insti-tutionalize the best practices, and that resulted in a reversal of performance. But why did the Hawthorne effect happen in our situation?

Employee attitudes

By accident, we discovered a key aspect: “Proven good people did not like following the best practices.” During a relaxed, casual conversation with a colleague who was efficient in his job, when we gently prodded him on why he did not follow the best practice of a peer check, he quickly replied: “I do not like to check others and, similarly, I do not want my work to be checked by others.” We root caused this further on workers’ opinions to all 10 initiatives.

We observed the employee dissatisfaction can hamper continuous improvement. We met with all employees, and they explained how they perceived the best practices. Consider one employee’s responses to questions in this exchange:

• Why don’t you validate orders? “I can perfectly judge all missing information. By asking me to escalate and find out from the client, you’re ques-tioning my optimism and my confidence in my abilities. I know I can do it right—even without validation—through my own judgment.”

• What happens if it goes wrong? We make errors 10% of the time, but we cannot afford more than 1% of errors per our contract. “Others may make mistakes, but not me.”

• Why don’t you perform peer checks? “This hurts my ego. I do not want someone to check my work. He or she is no better than me.”

• Have a peer couple. You check his and vice versa. “No—this is too much policing.”

• Why don’t you like the training sessions? “It takes place on an off day. Besides, we already know 95% of what’s being taught.”

• Do you find the executive scorecard and execu-tive error history useful? “Earlier, we were never

Figure 8. Synopsis on diagnostic approach and implementation method

Validate

Peer check

Peer couple

Training and online test

Executive scorecard

Executive error history

Quality error drive

Team building

Consultant update

MPDB update

2008 initiatives

Employee reaction to the initiative— without considering the human element

Innovation Applying TPS and cognitive dissonance

“Questions my optimism and confidence.”

“Hurts my ego.”

“Too much policing.”

“Why learn what I already know?”

“I am measured.”

“Low-impact knowledge.”

“Good, but not great.”

“Not all are working.”

“Practical, but partial.”

Cell

Best practices into roles

Roles linked to cells

Single-piece flow through the cell

Pull (built-in check)

Downstream: Auto quality control at every stage

Upstream: Auto turnaround time check at every stage

Perpetual visibility

Nonvalue-added time

Quality

Monitoring

Perpetual training

Stop, correct and proceed

MPDB = master pricing database TPS = Toyota Production System


The Human Element

Post

of

f-sho

ring

2009

Full

year

Opt

imiz

ed

Toyota Production System and cognitive dissonance theory

2008

2

nd h

alf

Man

aged

2008

1st h

alf

Ad h

oc

Pre

off-s

horin

g

2007

Figure 9. Process—success glide path

Customers Orders Line items

2,500 12,000 50,000

100,000 products

200+ factories

160+ countries

Purchase order to sales order

Manufacturing

Emails, e-fax, EOC

Sales order

Global customer

Sales

CAA

Purchase order

Mailbox

CustomerValidate Peer

checkQualitycheck Monitor CAA MFGOrder

entry

Orderentry

Orderentry

10 minutes 10 minutes10 minutes

30 minutes

Cell system—99%, 4 hours

Validate Peercheck Quality check

Quality check

CAA MFGOrderentry

Best practices into the process—97%-99%, 15 hours

Order entry CAA MFG

Basic process—92%, 9 hours

CAA enters the order themselves

Challengeslegacy system

No GUI

CAA MFG

Pre off-shoring—95%, 24 hours

Customer account analysts (CAA)

EOC = electronic order center GUI = graphic user interface MFG = manufacturing

measured. Now that’s started and the error his-tory tells us nothing new. We already know where we make errors.”

• What about the zero defect drives? “Good, but nothing great there.”

• How about the spirit of teamwork? “You think everyone is working to the same extent? There are those who leave for a 15-minute break and return two hours later. The team leader never questions them. How can we talk about teamwork and team spirit?”

• Master pricing databases and other customer notes? “Low-impact initiatives.”

What we learned from this exchange was that best practices actually reduced employee satisfaction. This is operational excellence without considering the human element.

The results were tabulated (see Figure 4, p. 13). Employees’ attitudes toward best practices were found to be negative and therefore, their behavior quickly moved away from best practices. However, we know that best practices cannot be changed so easily. So the question became, “How do we change employees’ attitudes?”

Changing attitudes

Based on lessons from the social psychologist Leon Festinger’s cognitive dissonance theory and TPS, which stresses the philosophy of single-piece flow, cell structure, stop-correct-proceed and building con-sensus, we achieved sustained improvement in quality and TAT, as well as improved job satisfaction of team members.

In the 1940s, Yale University in New Haven, CT, suffered a rare social event in which its students were manhandled by police. Public opinion was almost totally against the police. Festinger decided to conduct an experiment.

He chose 200 subjects in the community: students, teachers, non-teaching staff and members of the gen-eral public. He initially measured their attitude toward police at 100% negative, which was expected. He tried to change their behavior, not their attitude, by asking each one of them to write a 1,000-word article support-ing the police action.

He put the people into four groups and told three of them that they would be paid to write the articles. The first group would receive $1,000, the second group


The Human Element

$100, and the third group $10. The fourth group would not be paid. This created a gradation across the groups.

Writing a 1,000-word article takes time and involve-ment. Two months later, he called on all of the groups to measure their attitude again. He expected changes, and thought those who had received $1,000 must have changed their attitudes slightly in favor of the police.

But to his surprise, he found contradictory results: Most of them still had negative outlooks toward the police—as much as they did on day No. 1. When he measured those who received no money, however, Festinger found they had changed their attitude sig-nificantly, so much so that some of them were now actually favoring the police. The gradation was in line with the money received.

After much reflection, Festinger explained: • Those groups receiving money felt that they wrote

in support of police only because of the money. No dissonance.

• Those who didn’t receive money had to hold two conflicting thoughts or attitudes at the same time: one small thought in support of the police (created from their own behavior) and the other large one against the police (created from their experience).

• When they held two thoughts together in their conscious mind, this resulted in cognitive disso-nance. Over time, they had to give up one attitude

Table 2. Voice of the employees

Figure 10. Quality—measured as defective orders every week

99.5% of orders are

error-free

Cognitive dissonance theory

Toyota Production System

0

5

10

15

20

25

30

Wk

1

Wk

4

Wk

7

Wk

10

Wk

13

Wk

16

Wk

19

Wk

22

Wk

25

Wk

28

Wk

31

Wk

34

Wk

37

Wk

40

Wk

43

Wk

46

Wk

49

Wk

52

Wk

3

Wk

6

Wk

9

Wk

12

Wk

15

Wk

18

Wk

21

Wk

24

Wk

27

Wk

30

Wk

33

Wk

36

Wk

39

Wk

42

Wk

45

Wk

48

Wk

51

Sustainable results


The Human Element

• “Wonderful concept to make things happen.”

• “Ownership of role.”

• “Roles are defined clearly and make us work with ease.”

• “Team members develop expertise to their role.”

• “People have become more responsible in the role they play because it becomes their duty to make it error free.”

• “New agents can enter the order quickly with less error because of the validation process.”

• “Increase the confidence level of new joiners for processing orders without defects.”

• “Because of the peer check, there is always a cushion for the validator and order entry agents because they can feel much more comfortable while entering orders.”

• “Employee involvement—It emphasizes the team to integrate with work cells for motivation.”

2009 employee survey: Operational excellence taking into account the human element

to be able to write an essay. The attitude that comes from behavior is far more powerful than the one coming from experience, so much so that it pushes the attitude coming from experi-ence out of the mind over a period of time. The attitude in favor of the police remains.

• Based on his experiment, Festinger concluded that “behavior changes attitudes” goes against the commonly accepted notion that “attitudes change behavior.”

Applying cognitive dissonance theory in our own context, we asked the question: What can change the attitudes of employees? The answer was: If best prac-tices become a behavior, and it’s sustained over time, it can change the attitude. However, our first attempt at doing this failed. We forced best practices behavior and it failed. So how would we proceed? We thought to induce the teams to follow best practices, we needed to get them to change their behavior. The question is how to move from enforcing behavior to inducing behavior.

Changing behavior

We found the answer to this predicament in Peter Senge’s theory of systems thinking. In the 1970s, Stanford University in California conducted a novel

experiment in which 26 friendly student volunteers were told to act like thieves and guards in a prison. It is a human tendency that if we ask someone whether he or she would like to be a thief or a career criminal, some would say “Yes” and most others “No.”

The friendly student volunteers were divided into two groups: thieves to be locked in a cell in the base-ment of the Stanford University’s psychology building, and guards armed with sticks. They were told that no manhandling between groups would be allowed, and closed-circuit TV cameras would watch them. Both groups were asked to act their roles convincingly for six weeks. After the experiment, they would receive $45,000. If they abandon the project before six weeks, they would receive nothing.

On day No. 1, each thief was submissive and the guards acted with little authority. By day No. 6, how-ever, forgetting that this was a mock experiment, the thieves started crying uncontrollably. What hap-pened was that the guards joined together, covered up the cameras and took the thieves one by one and started manhandling them. The experiment had to be stopped, and everyone was sent home.

Some would say the century’s greatest discovery in sociology had just taken place: Irrespective of who you are, if you are given a role to play and you are put in

Figure 11. Client satisfaction—TAT

“Very satis�ed” rating on two consecutive client satisfaction surveys—May and August

00:00

02:24

04:48

07:12

09:36

12:00

14:24

Wk

1

Wk

4

Wk

7

Wk

10

Wk

13

Wk

16

Wk

19

Wk

22

Wk

25

Wk

28

Wk

31

Wk

34

Wk

37

Wk

40

Wk

43

Wk

46

Wk

49

Wk

52

Wk

3

Wk

6

Wk

9

Wk

12

Wk

15

Wk

18

Wk

21

Wk

24

Wk

27

Wk

30

Wk

33

Wk

36

Wk

39

Wk

42

Wk

45

Wk

48

Wk

51

Sustainable results

15 hours

3:40 hours7 hours

Nonvalue-added time (NVAT)Turnaround time (TAT)


The Human Element

a system that supports that, whether you like it or not, you will play it. Senge concluded that systems and roles create their own behavior, not the people who are a part of that.

We took a cue from this theory and learned that to create a behavior, we needed a role and a system. This gave us a breakthrough thought: People do not like best practices and, therefore, do not follow them. People play roles, whether they like them or not. Therefore, we should convert best practices into roles and establish the right system to support these roles.

Best practices into roles. To put best practices into roles, we took three best practices in a sequence: vali-dation (V), OE and peer check (PC). Earlier, the sys-tem expected every executive to perform all the three roles. We converted best practices into roles, and each of them was assigned to one individual.

There are three roles now (those of the V execu-tive, OE executive and PC executive), and people are assigned specific roles depending on their skills sets and past performances. As a result, executives found their workloads becoming easier. Figure 5 (p. 14) illustrates this breakthrough thought.

These three roles needed to be linked into a cell to complete the task. Of these three, OE may take 30 minutes per order, while V and PC may take only 10 minutes per order. Therefore, we needed one V executive, followed by three OE executives, and one PC executive in every cell to balance the workflow. This system (Figure 6, p. 15) is TPS, a 20th century system of procedures, values and beliefs that can substantially improve quality and reduce wastes, including TAT. Several principles of TPS were deployed:

• Cell and single-piece flow: To complete the pro-cess (see Figure 7, p. 15), all of the roles were linked into a cell. When a single order flows through the cell, multiple sets of eyes can scruti-nize it, and thereby quality improves. The single-piece flow meant that unless you pass your work downstream, you cannot receive the next unit of work from upstream. Thus, if a particular stage has a bottleneck and is slow, the preceding stages will stop working, which will immediately make the bottleneck visible for correction.

• Pull made to work: The V executive now diligent-ly validates and does not use his or her judgment because his or her role is only to validate. In cases where he or she does not validate by escalating matters—but still fills in the missing information using his or his judgment—and sends the order to the next stage, the OE executive will return the order and demand that he or she properly validates the order. Otherwise, the OE executive will have an error based on the V executive’s judg-ment, and he or she will not allow this.

Best practices start to become a behavior. Similarly, when the PC executive finds an error, he or she alerts the upstream OE executive. If the OE executive con-tinues to make the error, the PC executive can reply: “I just pointed this out to you last time as well. Please be extra careful.” Thus, best practices are enforced once again. Every person downstream expects higher quality

Table 4. Economics of quality

Annualized benefits from the project

By avoiding costs USD $M

Cost of rework 55,562

Business lost through wrong dispatches 400,000

Return cost through wrong dispatches 192,000

Customer loyalty and business erosion through wrong dispatches

1,920,000

Interest cost saved through faster cash cycle 98,630

Total savings by avoiding costs 2,666,192

By increasing revenue

Growth from quicker turnaround time, with one additional day of business

480,000

Growth in business from reduced cancellation of orders

2,4000,000

Growth in business from better servicing 4,800,000

Total earnings by increasing revenue 8,880,000

Total savings plus earnings 10,346,192

Savings and earnings from the project = USD 10.3 million.

At the rate of 46.53 Indian Rupees (Rs) to one U.S. dollar (July 2009) the economic value of the quality initiatives Rs 48 crores.

Table 3. Client satisfaction

Client satisfaction survey (CSAT)—for > 99% quality and < 4 hour turnaround time (TAT) performance

Rating

Quality TAT Comments

April 2009

SatisfiedVery

satisfiedSource of client delight:

Responsiveness to last CSAT to drive improvements in order

entry TAT and quality is a source of our delight.

July 2009

SatisfiedVery

satisfied


The Human Element

from the upstream operator. Downstream to upstream, there are automatic quality checks occurring.

In another example, consider when the entire team leaves for a 15-minute break and returns on time—except for one team member who returns after an hour. Because this structure is a single-piece flow, the rest of the team will be unable to pass on work and therefore, the entire cell will stop functioning.

When the laggard returns, he or she will face a huge embarrassment when he or she realizes the entire team has not been working because of his or her absence. The person will certainly receive negative feedback and pressure from the group: “If you’re leaving for an hour, please tell us so we can also take off for an hour, rather than sit here idly waiting for your arrival.” Thus, every person upstream expects better time manage-ment from the downstream operator.

From upstream to downstream, there is a TAT check. Best practices continue to become behavior. Again, the internal energy of the team controls TAT and quality, as well as reinforces best practices as behaviors. Team leaders and managers do not even come to know about this. They just see the improvement.

• Stop-correct-proceed: When an error is identi-fied by anyone, he or calls everyone for a quick huddle and explains the error. Root cause analy-sis is performed and the problem is corrected. Now, everyone can get back to their work.

This exercise, which can last just a few min-utes in a group setting, acts as a micro training session, taking place exactly when errors occur. This on-the-spot micro training reinforces best practices as behavior and replaces normal week-end training sessions, which are considered boring.

• Sensitivity to nonvalue-added time (NVAT): The work happens in 12 parts. From the time one part is complete to the time the next activity starts, there is waiting, or NVAT. This is usually much higher than the value added time (VAT). For instance, when the TAT increased to 15 hours, just two hours were VAT and 13 hours were NVAT.

The team now realizes that what must be man-aged is NVAT, not VAT. Thus, the team avoids the tendency to work faster, which can affect only VAT, and instead focuses on working smarter through better scheduling, which can affect NVAT. This realization reduces the stress level on the floor when TAT goes up. Attention to this matter quickly brought down the TAT from 15 hours to four hours.

• Best practices become behaviors: TPS strate-gies thus help reinforce best practices and makes them into behaviors.

Figure 8 (p. 16) provides a summary of the initiatives and the various employee reactions and results. Figure 9 (p. 17) outlines the improved process.

Best practices become behavior

A year later in 2009, a survey of the same executives who told us the previous year that they didn’t like to work with best practices revealed a complete change in attitude, as Table 2 (p. 18) shows. Among other results from the project, key concepts addressed included:

• Sustainability: These principles have proved to be sustainable over a long 300-week period. When a special factor causes an imbalance, the team can bring things back into balance quickly within the next few weeks.

• Wider applicability: The principles are applicable to any form of complex, back-office processes involving teamwork as agreed by the eminent jury of the Quality Council of India.

Figure 10-11 (p. 18 and p. 19) show the final results of quality and TAT. Quality—as measured by defective orders per week—decreased during the 51 weeks, and TAT decreased to 3:40 hours.

Table 3 shows the results of client satisfaction, and Table 4 provides a synopsis of the annualized benefits from the project.

Clearly, an understanding of the appropriate psy-chological and production theories, as well as TPS con-cepts, helped us achieve great success in this project. Without realizing and appreciating how the human element affects the processes we were trying to change, no substantial progress would have been possible.

ACKNOWLEDGEMENTS

The author thanks Rangarajan, Rajagopal Swaminathan, S.K. Pillai, Anand Sivaraman, Purushotham Rangachari, John Sugirtharaj and members of HCL’s order management team for their work and support during this project. The author also thanks the eminent jury of the Quality Council of India (Avik Mitra, Anil K. Srivatsava, S.N. Nandi, Chandrasekhara Borate, AKS Sharma and Jain).

BIBLIOGRAPHY

Festinger, Leon, A Theory of Cognitive Dissonance, Stanford University Press, 1957.Gillespie, Richard, Manufacturing Knowledge: A History of the Hawthorne Experiments (Studies in Economic History and Policy: USA in the Twentieth Century), Cambridge University Press, 1993.Liker, Jeffrey K., The Toyota Way, McGraw-Hill Education, 1994.Senge, Peter, The Fifth Discipline, Crown Business, 1994.


The Human Element