IMB 833

EXCEED: IMPROVING PRODUCTIVITY AND RELIABILITY IN

DELIVERY OF SOFTWARE SOLUTIONS

SIVAKUMAR S AND B MAHADEVAN

Sivakumar S, Assistant Professor of Operations Management & Decision Sciences at IIM, Tiruchirappalli and B. Mahadevan,
Professor of Production & Operations Management, prepared this case for classroom discussion. Exceed Communications LLC
cooperated and provided information to the Indian Institute of Management Bangalore in connection with the preparation of
this case and it was reviewed and approved before publication by a Exceed Communications designate. No funding was sought
or received from Exceed Communications LLC for the development of this case. This case was also developed from available
and permitted sources of information. This case is not intended to serve as an endorsement, source of primary data, or to show
effective or inefficient handling of decision or business processes.

Copyright © 2021 by the Indian Institute of Management Bangalore. No part of the publication may be reproduced or
transmitted in any form or by any means – electronic, mechanical, photocopying, recording, or otherwise (including internet) –
without the permission of Indian Institute of Management Bangalore.

This document is authorized for use only in PROF. SAJI GOPINATH's Lean Six Sigma- Section B (EOM-02) at IIM Kozhikode - EPGP Kozhikode Campus from Dec 2024 to Jun 2025.
 Exceed: Improving Productivity and Reliability in Delivery of Software Solutions

It was a chilly Monday morning on February 6, 2017. Mohamed Karamathullah (Karamath henceforth), Managing
Director, Exceed Communications LLC (ECL) was on a visit to their offshore development center in Bangalore. He
was immersed in thoughts as he walked into his office in the buzzing industrial suburbs of Bangalore, India. ECL
was always proud of its track record in development and maintenance of integrated hardware and software
solutions and associated services. ECL was known for its reliable deliveries in line with their commitments.
However, recently, one customer account had been bothering Karamath. He began the workday by checking his e-
mails, a routine he diligently followed. He glanced through the weekly status reports sent on the previous day by
his client engagement manager. The e-mail on an escalation alert from one of his client managers caught his
attention. It was yet again the same customer account.

The latest complaint was about the slow progress of system testing of a software package being developed for that
client. Furthermore, the e-mail pointed to several defects found during the ongoing system testing phase. This
particular client was demanding and tended to jump the gun at times. What bothered Karamath was that the
frequency of escalation e-mails from this customer was high compared to their other customer accounts. Was this
latest complaint another one-off issue or was there a larger pattern behind this? What could he do to improve the
productivity and reliability of their deliveries to this client?

COMPANY BACKGROUND

Exceed Communications LLC (ECL) was a specialist hardware and software solutions’ company that provided
reliable, high quality, customized business solutions to clients based out of the Middle East, Africa, and South Asia.
The company was founded in 2013. As of January 2017, they had a team size of 36 members spread between their
offices in the Middle East and Bangalore. ECL was rated by the industry analysts as one of the fastest-growing
hardware and software solutions’ company in the United Arab Emirates. They prided themselves on a solution-
based approach to understand the business ambitions and technological requirements of their customers, to
uniquely craft a techno-business solution customized to their clients. They provided a variety of network and
security solutions, hardware cum software development and maintenance services, and digital marketing solutions
to their clients. ECL had a technological partnership with a wide array of global product labels including 3M, Avaya,
Bosch, Cisco, HP, Dell, Microsoft, Samsung, Schneider, and Xerox. As of January 2017, ECL had over 30 marquee
clients spread across the Middle East, Africa, and South Asia.

SOFTWARE DEVELOPMENT PROJECT FOR H&K ACCOUNT

H&K (name disguised) was a market leader in providing real estate and integrated business management solutions
to its clients. Its relationship with ECL began in October 2015. ECL was developing a bespoke software package for
customer relationship management of the clients of H&K. The project was being executed on a Time and Material1
basis, where the client closely monitored both staffing volumes and staff productivity. The project followed a V-
process model in the software development lifecycle (Exhibit 1). This model was an improved version of the
waterfall process model that started with requirement gathering and ended with user-acceptance testing. In this
process model, the test cases for the various testing phases were developed alongside the requirements, higher-
level, and detail design documents. The current project started in August 2016 and on February 6, 2017 was into
the system testing phase in the client’s test environment. The project had experienced its ups and downs, but the
dedication and capability of ECL team members had played a significant role in the progress of the project.
Although the project was delayed at the start due to delay in collecting business requirements, the team had made
good the delay during the development phase. However, in the extremely competitive environment that ECL
operated in, clients such as H&K were increasingly demanding higher quality and quicker deliveries.

1
Time and Material is a common billing model prevalent in professional services where the vendor bills the client for the effort incurred (as
recorded in a timesheet) and actual cost of materials used to service the client.

Page 2 of 8

This document is authorized for use only in PROF. SAJI GOPINATH's Lean Six Sigma- Section B (EOM-02) at IIM Kozhikode - EPGP Kozhikode Campus from Dec 2024 to Jun 2025.
 Exceed: Improving Productivity and Reliability in Delivery of Software Solutions

Karamath telephoned Mohammed Aslam to discuss the latest customer complaint. Aslam was the head of
operations, responsible for all services provided to H&K and was based out of their Dubai office. Aslam sent his
spreadsheet tracker that meticulously recorded all the customer complaints received from H&K since the start of
the project (Exhibit 2). Karamath quickly glanced through the tabulation but was not sure if there was any pattern
behind the complaints. The customer appeared to have frequently complained about the team working too slowly
(productivity complaint) and too many errors (quality complaint). The complaints on excess staffing were a result
of the client micromanaging the project based on the timesheet submitted by the project personnel. However,
Karamath was not sure as to which of these categories of complaints required immediate attention.

PRODUCTIVITY DATA

Karamath had also asked Aslam to send him data on the progress of system testing activity that was underway.
The activity involved the execution of system test cases to verify the conformance of the developed software
package to the specifications. Aslam had collated the statistics of test case execution for every two-hour window
for the past 4 weeks (Exhibit 3). Karamath quickly calculated the average of test cases executed. He felt it was
better than the company average. There were some days when the numbers were low, but according to him, there
appeared nothing alarming to complain about. When he inquired Aslam about the reasons for customer
complaints, Aslam had the following to say:

The latest customer complaint was perhaps referring to testing performed on 12 th January, 17th January or
on 1st February; those were the days of lower productivity. These were one-off discrepancies, and
everything was under control. According to me, there was nothing that required further action.

But was it? wondered Karamath.

TEAM MEETING

Karamath arranged a conference call with the entire project team to discuss the various concerns raised by the
client. All the senior members of the client engagement, development, and testing teams were invited to this
conference call. Karamath listed the summary of the concerns reported by the client and particularly raised the
issue of inconsistent productivity in the execution of system test cases. The attendees came up with various
explanations for this issue.

One senior testing team member stated:

Execution of test cases is not easy, because we have to identify suitable test data for testing the identified
test scenarios. It sometimes requires a complicated SQL code to identify the relevant test data for
execution. This takes a lot of time!

The head of system testing team remarked:

We are not always operating at our full potential. We have an agreement that the development team will
deploy a fresh software package before 9 AM, after fixing of the bugs. But many days their deployment
finishes only by 11 AM and our testing productivity is impacted on account of this.

The other members of the system testing team complained that their computers had insufficient RAM slowing
down the testing process. There was also latency in the network forcing slow access to the client’s test
environment. Karamath had earlier heard complaints about the latency in accessing client network, but he was not
sure whether that was the primary explanation for low productivity.

Page 3 of 8

This document is authorized for use only in PROF. SAJI GOPINATH's Lean Six Sigma- Section B (EOM-02) at IIM Kozhikode - EPGP Kozhikode Campus from Dec 2024 to Jun 2025.
 Exceed: Improving Productivity and Reliability in Delivery of Software Solutions

One of the technical experts in the team called Dashrath who had a background in the automation of testing
processes stated:

Sir, our testing is bound to have low productivity as it is being done manually. We had test automation in
the previous project for H&K and they are perhaps expecting the same standard now. In manual testing
when we have to re-test failed test scenarios, it requires data entry once again. This involves a significant
time overhead.

Although testing automation was an expensive proposition, Karamath thought there was merit in the argument
made by Dashrath. Manual testing relied on the testing personnel to enter the test case/test data on the software
package and verify the results against the specification. On the other hand, testing automation relied on special
software programs called test scripts for automated entry of test cases and verification of test results. However,
testing automation cannot be recommended to H&K as a solution without a proper business case on the cost-
benefits involved. During the informal chat after the meeting, some members made a casual remark about the
ongoing cricket series involving India and how it could be a productivity dampener. Some members also whispered
that on the days when Aslam left office early, offshore team members tended to relax. This was for the first time
Karamath that became aware of the issues of distraction and shirking.

At the end of the meeting, Karamath was more confused than before. There seemed to be several causes to
explain low productivity. However, he started to wonder which one needed his immediate attention.

WAY FORWARD

Karamath walked back to his cabin with mixed thoughts on proceeding further. He had asked Imran Sheik, the
head of the network division in their Bangalore office to provide him with network response time statistics for the
first week of the corresponding period. The report had just arrived in his mailbox (Exhibit 4). The inconsistency in
the network response times was evident.

But was that the primary explanation for low productivity? What about the other issues such as infrastructure,
coordination, communication, and discipline then? Which among these issues needed immediate attention?
Would testing automation solve all the productivity problems so that he could recommend that to the customer?
These questions remained unanswered and Karamath knew the nature of the challenge that lay ahead.

Page 4 of 8

This document is authorized for use only in PROF. SAJI GOPINATH's Lean Six Sigma- Section B (EOM-02) at IIM Kozhikode - EPGP Kozhikode Campus from Dec 2024 to Jun 2025.
 Exceed: Improving Productivity and Reliability in Delivery of Software Solutions

Exhibit 1
V-Process model for software development

Source: Compiled from https://www.tutorialspoint.com/software_testing_dictionary/v_model.htm in May 2019

Page 5 of 8

This document is authorized for use only in PROF. SAJI GOPINATH's Lean Six Sigma- Section B (EOM-02) at IIM Kozhikode - EPGP Kozhikode Campus from Dec 2024 to Jun 2025.
 Exceed: Improving Productivity and Reliability in Delivery of Software Solutions

Exhibit 2
Customer complaints tracker

Serial # Complaint Category Date raised Project Phase2

1 Team Working Slowly 10-Aug-16 Requirements
2 Too Many Errors 19-Aug-16 Requirements
3 Communication Issue 19-Aug-16 Requirements
4 Excess Staffing 24-Aug-16 Requirements
5 Excess Staffing 26-Aug-16 Requirements
6 Too Many Errors 16-Sep-16 High-Level Design
7 Excess Staffing 22-Sep-16 High-Level Design
8 Excess Staffing 26-Sep-16 High-Level Design
9 Domain Knowledge 12-Oct-16 High-Level Design
10 Communication Issue 12-Oct-16 High-Level Design
11 Team Working Slowly 22-Oct-16 High-Level Design
12 Team Working Slowly 26-Oct-16 Detail Design
13 Team Working Slowly 28-Oct-16 Detail Design
14 Team Working Slowly 11-Nov-16 Coding & Unit Testing
15 Team Working Slowly 16-Nov-16 Coding & Unit Testing
16 Team Working Slowly 25-Nov-16 Coding & Unit Testing
17 Excess Staffing 28-Nov-16 Coding & Unit Testing
18 Team Working Slowly 16-Dec-16 Integration Testing
19 Too Many Errors 21-Dec-16 Integration Testing
20 Team Working Slowly 25-Dec-16 Integration Testing
21 Too Many Errors 29-Dec-16 Integration Testing
22 Team Working Slowly 02-Jan-17 System Testing
23 Excess Staffing 05-Jan-17 System Testing
24 Domain Knowledge 06-Jan-17 System Testing
25 Team Working Slowly 06-Jan-17 System Testing
26 Team Working Slowly 10-Jan-17 System Testing
27 Team Working Slowly 19-Jan-17 System Testing
28 Excess Staffing 23-Jan-17 System Testing
29 Communication Issue 23-Jan-17 System Testing
30 Too Many Errors 27-Jan-17 System Testing
31 Too Many Errors 31-Jan-17 System Testing
32 Too Many Errors 01-Feb-17 System Testing
33 Communication Issue 01-Feb-17 System Testing
34 Too Many Errors 02-Feb-17 System Testing
35 Team Working Slowly 03-Feb-17 System Testing
36 Too Many Errors 03-Feb-17 System Testing
Source: Company Records

2
Project phase indicates the incident phase in the V process model (as explained in Exhibit 1) when the issue was raised.

Page 6 of 8

This document is authorized for use only in PROF. SAJI GOPINATH's Lean Six Sigma- Section B (EOM-02) at IIM Kozhikode - EPGP Kozhikode Campus from Dec 2024 to Jun 2025.
 Exceed: Improving Productivity and Reliability in Delivery of Software Solutions

Exhibit 3
System test case execution statistics

Time 3 09-Jan-17 10-Jan-17 11-Jan-17 12-Jan-17 13-Jan-17 16-Jan-17 17-Jan-17 18-Jan-17 19-Jan-17 20-Jan-17
9:00 AM 81 82 81 74 75 81 83 86 88 82
11:00 AM 73 87 83 81 86 86 82 83 79 84
1:00 PM 85 88 76 72 83 82 76 82 86 89
3:00 PM 90 78 84 73 84 88 77 79 84 83
5:00 PM 88 83 82 83 78 81 78 85 85 80
Time 23-Jan-17 24-Jan-17 25-Jan-17 26-Jan-17 27-Jan-17 30-Jan-17 31-Jan-17 01-Feb-17 02-Feb-17 03-Feb-17
9:00 AM 86 86 88 73 84 76 74 85 82 88
11:00 AM 83 83 79 86 85 82 86 85 83 79
1:00 PM 81 78 83 81 81 84 83 82 83 92
3:00 PM 81 80 83 79 87 85 89 77 90 83
5:00 PM 87 83 82 87 81 79 83 77 84 77

Source: Company Records

3 Time refers to the start time of a time window of 2-hour duration for the execution of system test cases

Page 7 of 8

This document is authorized for use only in PROF. SAJI GOPINATH's Lean Six Sigma- Section B (EOM-02) at IIM Kozhikode - EPGP Kozhikode Campus from Dec 2024 to Jun 2025.
 Exceed: Improving Productivity and Reliability in Delivery of Software Solutions

Exhibit 4
Network response time metrics

Response times
Date Time
(in milliseconds)4
09-Jan-17 9:00 AM 222
09-Jan-17 11:00 AM 234
09-Jan-17 1:00 PM 214
09-Jan-17 3:00 PM 211
09-Jan-17 5:00 PM 211
10-Jan-17 9:00 AM 221
10-Jan-17 11:00 AM 215
10-Jan-17 1:00 PM 210
10-Jan-17 3:00 PM 226
10-Jan-17 5:00 PM 218
11-Jan-17 9:00 AM 218
11-Jan-17 11:00 AM 219
11-Jan-17 1:00 PM 229
11-Jan-17 3:00 PM 218
11-Jan-17 5:00 PM 221
12-Jan-17 9:00 AM 235
12-Jan-17 11:00 AM 222
12-Jan-17 1:00 PM 236
12-Jan-17 3:00 PM 230
12-Jan-17 5:00 PM 221
13-Jan-17 9:00 AM 230
13-Jan-17 11:00 AM 215
13-Jan-17 1:00 PM 221
13-Jan-17 3:00 PM 217
13-Jan-17 5:00 PM 230

Source: Company Records

4
Response time is the time taken for a request and response of data packet to travel between the offshore servers of Exceed and H&K client
network. It is a measure of latency in data communication between the two networks.

Page 8 of 8

This document is authorized for use only in PROF. SAJI GOPINATH's Lean Six Sigma- Section B (EOM-02) at IIM Kozhikode - EPGP Kozhikode Campus from Dec 2024 to Jun 2025.