Robert D. Austin, Richard L. Nolan, and Shannon O’Donnell
Buy the book:
Amazon
Barnes & Noble
HarvardBusiness.org
Harvard Business Press
Boston, Massachusetts
ISBN-13: 978-1-4221-3046-9
3028BC
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
op yo Copyright 2009 Harvard Business School Publishing Corporation
All rights reserved
Printed in the United States of America
This chapter was originally published as chapter 14 of The Adventures of an IT Leader, copyright 2009 Harvard Business School Publishing Corporation.
No part of this publication may be reproduced, stored in or introduced into a retrieval system, or transmitted, in any form, or by any means (electronic, mechanical, photocopying, recording, or otherwise), without the prior permission of the publisher. Requests for permission should be directed to permissions@harvardbusiness.org, or mailed to Permissions,
Harvard Business School Publishing, 60 Harvard Way, Boston, Massachusetts 02163.
Do
No
tC
You can purchase Harvard Business Press books at booksellers worldwide. You can order Harvard
Business Press books and book chapters online at www.harvardbusiness.org/press, or by calling 888-500-1016 or, outside the U.S. and Canada, 617-783-7410.
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
op yo Series overview
tC
The Adventures of an IT Leader invites readers to “walk in the shoes” of a new CIO as he spends a difficult year learning effective information technology leadership. Experienced cumulatively, this eighteenchapter story gains dramatic momentum, and later chapters provide opportunities to revisit key IT management issues in more depth.
However, chapters can also be read independently or in smaller batches as suits the needs of particular readers. To facilitate this, we provide the following contextual information.
SUMMARY
Do
No
As the story begins in chapter 1, the IVK Corporation, a midsize financial services firm, is attempting a turnaround following a period of slowing business performance1. The stock price has fallen substantially as investors have adjusted their expectations of the firm’s growth. An aggressive new CEO, Carl Williams, takes over and assigns a new management team. In the process, the former head of Loan Operations, Jim Barton, is appointed CIO. Barton has no background in IT—none at all. The story
1
The IVK Corporation and its staff are fictional, but the contents of the book are based on the authors’years of firsthand experience with diverse companies and managers.
iii
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
Series Overview
follows Barton as he figures out what effective IT management is all about and deals with issues and challenges of the job. These broadly include, by chapter number:
• The challenges of information technology leadership (1–3)
• Managing the IT budget (4)
• Maximizing the value of IT (5)
op yo • Approaches to project management (6)
• Managing large projects (7)
• Prioritizing among a portfolio of projects (8)
• Board-level governance (9)
• Management and aftermath of a security crisis (10–11)
• Communication and interaction with the boss and peers (12)
• Analyzing emerging technologies (13)
• Arranging and managing partnerships with vendors (14)
tC
• Managing highly talented employees (15)
• Investing in infrastructure to move toward standardization and innovation (16)
• Managing risk (17)
Do
No
• Job opportunities for IT leaders (18)
The Main Characters
In order of appearance . . .
Jim Barton: The new CIO of IVK. A talented and ambitious general manager, formerly the head of Loan Operations. Barton knows little about IT; he sets out to learn quickly and to lead the IT department toward renewed growth, stability, and strategic partnership within the company—but not without facing serious challenges.
iv
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
Series Overview
Carl Williams: This bold turnaround CEO is high on ambition and short on patience.
Maggie Landis: A savvy management consultant and Barton’s girlfriend, she often provides Barton with valuable insight, references, and perspectives.
op yo The kid: Wise beyond his years, this twenty-something tech nerd, whom Barton mysteriously meets only at Vinnie’s Bar, proves a useful sounding board and source of surprisingly good advice.
Bill Davies: Former CIO at IVK, Davies was fired in part because he struggled with management-level communication. He tells Barton that he “won’t last one year” in the job of CIO.
Bernie Ruben: As the director of the Technical Services Group and longtime IVK employee, Ruben is nearing retirement and thus mostly immune to concerns about risk to his career. He frequently provides Barton with the candid advice, knowledge, and context he needs to make key decisions.
tC
Raj Juvvani: The director of Customer Support and Collection Systems and part of Barton’s core IT team.
Tyra Gordon: As director of Loan Operations and New Application
Development Systems, Tyra worked closely with Barton when he was head of Loan Operations and takes the lead on several new IT projects under his management.
Do
No
Paul Fenton: Director of Infrastructure and Operations, Fenton manages a large and important domain, including IT security, and is part of Barton’s core IT team.
Gary Geisler: As director of Planning and Control, Geisler works closely with Barton on IT financials.
John Cho: IVK’s outspoken resident security genius, Cho has a distinct fashion sense and provocative musical talent.
Jenny: Barton’s ever-dependable executive assistant.
Several additional characters populate the story, but are described in context. v
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
Series Overview
GLOSSARY OF ACRONYMS AND TERMS
A public-domain, open source Web server
A/P
Applications portfolio
APM
Agile project management; also applications portfolio management CPM
op yo Apache
CRM
Customer relationship management
COO
Chief operating officer
CQ
Competes [versus] Qualifiers
DBA
Database administration
APP
BGP
CEO
CFO
CIO
CMM
Border gateway control
Chief executive officer
Chief financial officer
Chief information officer
Capability maturity model
Common Object Request Broker Architecture
Critical path method
Do
No
tC
CORBA
Aggregate project planning
DBMS
Database management system
DoS
Denial of service
DP Era
Data processing era
EAI
Enterprise application integration
ERP
Enterprise resource planning
IDS
Intrusion detection system
IPO
Initial public offering
vi
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
IR or IRP
Infrastructure replacement project (as used in this book) IT
Information technology
ITIL
Information technology infrastructure library
Java
A high-level programming language developed by Sun
Microsystems
KWYDK
“Know what you don’t know” (as used in this book)
Linux
op yo Series Overiew
OCI
Oracle
OSI
PC
PDA
PERL
Option creating investment (as used in this book)
Large software company, focused on database products Open systems interconnection
Personal computer
Personal digital assistant
Practical Extraction and Report Language
Project Evaluation and Review Technique
tC
PERT
An open source operating system
Pretty Good Privacy
PLM
Product lifecycle management
RFP
Request for proposal
ROI
Return on investment
SaaS
Software as a service
Sabre
Semi-automated business research environment
SLA
Service level agreement
SOA
Service-oriented architecture
SQL
Structured query language (originally SEQUEL: structured English query language)
SSL
Secure sockets layer
Do
No
PGP
vii
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
SYH2DP
“Sometimes you have to duck a punch” (as used in this book)
TCP/IP
Transmission Control Protocol/Internet Protocol
TPM
Traditional project management
Unix
A multi-user, multitasking operating system
UWGDF
“Understand what got Davies fired” (as used in this book) VPN
op yo Series Overview
“You won’t last one year” (as used in this book)
Do
No
tC
YWLOY
Virtual private network
viii
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
chapter fourteen
op yo Vendor Partnering
Thursday, September 27, 10:55 a.m. . . .
Do
No
tC
“I tell you, that’s simply not going to work!” The speaker pounded his fist on the table. Others in the room let loose a collective sigh. Frustration filled the air like an unpleasant smell.
Barton sat at the back of the conference room, listening. He struggled to control an urge to wade in with his opinion. Two camps had formed, and they disagreed bitterly. Twice before in the span of a week he’d sat through unproductive meetings on this or related topics. The group leading the effort to replace IVK’s increasingly decrepit backoffice systems, the IR team, had been basically stuck for months now.
Someone, or something, needed to break the pattern of dysfunction.
But not Jim Barton. He was the one person in the company who could not do it, who had to stay out of it. He’d dealt a major setback to the effort when he’d canceled the NetiFects contract in May. It had been one of his earliest actions, arguably the most decisive. But, also, Barton knew, the most second-guessed. Barton still believed it had been the right move, but his role in slowing the project down made him both desperate to see the group regain forward momentum and certain that he could not again intervene without creating even greater problems.
When Barton had jettisoned NetiFects, the IVK cross-disciplinary, business-led group that had been working on back-office systems replacement had seemed to take it well. They’d had their own frustrations
1
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
The Hero Breaks Through
Do
No
tC
op yo with NetiFects. Many of them felt satisfaction when Barton summarily dismissed smooth-talking Carlton Leopold. But hard feelings toward
Barton lingered among members of the group, borne of the fact that
Barton had made the decision to fire NetiFects without consulting them. His unilateral decision to add IT people to the group’s membership, justified by an assertion that the group needed more technical experts to ensure adequate evaluation of the merits of solutions proposed by vendors, added insult to injury. Everyone knew that technical evaluation was important. But the group felt that Barton had exerted too much influence. They had agreed, in response to Barton’s urging, that they would avoid the paralysis that had set in last time a result of the sheer size of the project, by starting with a major off-the-shelf package. They would not develop a custom system out of small components, using a systems integrator for new software development, as had been the direction evolving under the previous plan. Instead, they’d start with a big system and use a systems integrator to help install it and, perhaps, customize parts of it. This agreement, however, formed a foundation for three new— and major—disagreements.
The first centered on selection of the off-the-shelf package. The group had narrowed an initial list of more than twenty candidates down to three finalists. Much of this work had been straightforward. Choice of a product also meant choice of the vendor that would become a longterm IVK partner, and many companies that submitted initial proposals were, for one reason or another, unsuitable partners (too small, too close to a major competitor). Others had obviously inferior product offerings. The handful that remained after an early cut had to be evaluated more carefully.
That summer, the group had issued a request for proposal. This RFP asked vendors to self-assess the degree of fit between their product and the specification that IVK had developed when the emphasis had been on working with NetiFects to create a customized system. The IVK team then set out to verify the degree of fit claimed by the top seven vendors. They visited vendor sites, watched demonstrations of software in action, discussed experiences with reference customers, and attended a final presentation by each vendor. Every member of the team had also,
2
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
Vendor Partnering
Do
No
tC
op yo in theory, carefully reviewed each of the multivolume proposals submitted by vendors. Upon receiving the proposals, some on the team had joked that the contestants seemed to think they would be chosen based on the weight of proposal binders. The documents made substantial and difficult reading. Gradually, however, the group ruled out candidates from the list of seven. Eventually, three candidates remained, each with significant appeal to a subgroup within the IR team.*
HiOSoft had historically operated primarily as a software company focused on a product line for customers in many more businesses than financial services. Their product fit reasonably with IVK’s requirements, but it was arguably the least suitable as installed out of the box. Also, because the company had not been a services firm historically, its proposal relied a great deal on partners; the IVK team agreed that HiOSoft probably ranked third in maturity of service offerings. But HiOSoft’s product had one compelling strength: a very open architecture, which would arguably make the customization IVK needed to do in the future much easier. Still, though the product fared well in terms of flexibility, it measured up less well in terms of robustness. The company had been late to the market for industrial-strength, high-reliability products, so there were questions marks in this area that disconcerted some. Other, mostly IT, people believed there was no real reason you shouldn’t be able to engineer a very robust system based on the HiOSoft proposal.
On the IVK IR team, HiOSoft and its product had support among the smallest minority, much of it from people who had favored the initial
NetiFects component-based approach; they also enjoyed disproportionate support from technical members of the IR team.
VerxaWeb had been labeled (by its detractors) “the sentimental favorite.” And it was true—people liked VerxaWeb. Its products and services specialized in the financial services industry. VerxaWeb spoke
IVK’s language, and the two organizations fit well together culturally.
Their product fit well with IVK’s specification. But some on the IVK team felt that the considerably smaller and less financially successful vendor tried too hard. They worried about long-term viability of this
*See “Vendor Assessment Matrices” at the end of this chapter.
3
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
The Hero Breaks Through
Do
No
tC
op yo potential partner. VerxaWeb, some worried, might be a partner that
IVK would too soon outgrow.
No one had such worries about ServoLith, by far the largest and most successful company in the group of finalists. Their product focused less on financial services, but fit well with the specification. The concern with ServoLith: whether the vendor’s size and success would make IVK, a relatively small company among ServoLith clients, a lowpriority customer. Many disliked this vendor based on past dealings.
Words used to describe them included “arrogant,” “overpriced,” and
“unresponsive.” During the vendor presentations, some had felt vindicated in their harsh assessments when ServoLith failed to produce a technical expert as promised, instead substituting someone more junior. ServoLith representatives said that the promised expert had been detained because of a “personal matter.” But an explanation obtained through the grapevine, from someone who knew someone who knew someone within ServoLith contradicted the vendor’s assertion. Supposedly, a major company—much bigger than IVK—had summoned the technical expert to their site on the day he’d been promised to IVK; the big client’s demands had won out when measured against the request from a prospective, smallish client. To some, this was a red flag; it might very well be a behavior destined to be repeated if IVK chose to work with ServoLith. Nevertheless, choice of ServoLith, the industry leader, would be defensible to analysts, shareholders, and executives, no matter how you looked at it. Because this vendor had so many large clients, there were few doubts about the scalability or robustness of their products. Questions centered mostly on whether they would be a sufficiently attentive partner.
To Barton, HiOSoft appeared to be losing out as increasingly strident voices gathered around the other two vendors. Advocates of ServoLith accused proponents of VerxaWeb of ignoring hard facts in favor of mysterious “intangibles.” VerxaWeb advocates accused proponents of ServoLith of taking people factors and past experience with that vendor too little into account. Barton thought positions appeared to be hardening, and he saw no indications that the group would break through to consensus anytime soon. If he’d dared enter the debate, Barton would have urged that they focus less on deciding and more on
4
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
Vendor Partnering
Do
No
tC
op yo coming up with an agreed method for making a decision they’d all consider reasonable and by which they’d all be willing to abide. But he didn’t dare enter, at least not yet.
The second major topic of debate had to do with the proposed contract structure that would define the relationship with whichever vendor might eventually be chosen. In this discussion, the divide had formed around what Barton thought of as “hard-line” and the “soft-cooperation” philosophies. Hard-liners advocated a service level agreement with, as they put it,
“real teeth.” If the vendor failed to deliver services as contracted, on the schedule laid out in the SLA, then the vendor would pay significant and painful penalties to IVK. Those who favored this approach believed it would assure maximum effort from the vendor and strong alignment of the interests of the two parties. This group was also more likely to be interested in front-loading the vendor contract in IVK’s favor, so that benefits would be realized by the vendor only after the partnership had been in place for a while and the benefits of the partnership to IVK had become obvious. In this sort of scheme, IVK would pay discounted prices to the vendor at first, which would become more lucrative for the vendor as the initiative progressed.
The soft cooperators argued for a more collaborative and less adversarial approach to contracting. They believed the SLA should be in place to eliminate ambiguity in definitions of vendor “performance,” but they did not see the point in severe penalties that might, at least in the case of
VerxaWeb, actually harm the vendor partner (there was significant overlap between the people of this opinion and those who favored VerxaWeb). This contingent also opposed an adversarial approach to the payment schedule to the vendor, and instead proposed setting up a contract based on fairness and profitability for both parties.
The debate about contract structure interacted with the debate about choice of vendor. In particular, some observed that IVK’s ability to dictate contract terms would be considerably diminished if they chose ServoLith. That large company, which could, in the final analysis, easily afford to lose IVK’s business, would be less flexible on contract terms and much else. Probably IVK would have to work with them on relatively standard terms that the vendor would disproportionately
5
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
The Hero Breaks Through
Do
No
tC
op yo dictate. The standard ServoLith contract would be designed, no doubt, to avoid any serious teeth that might impact the vendor. In contrast,
IVK would have a great deal of contract negotiation leverage with VerxaWeb, whose managers seemed especially hungry for the IVK deal; IVK would be able to put big teeth in the deal, but teeth too big might chew hard into VerxaWeb. The coalition of vendors specified in the proposal from HiOSoft offered flexibility in contracting, but also complexity that might be hard to manage. In the HiOSoft proposal, there would not be one SLA but many, each with a different partner vendor.
The third major disagreement centered on what the group had labeled the service delivery model. The three vendors had proposed quite different ways to meet the requirements in the IVK specification, and very different structures of ownership.
ServoLith, for example, had proposed a total solution model, in which the vendor would own the data centers, as well as all the equipment and software in use, and provide services to IVK business personnel and clients over network connections. IVK would pay for these services in a monthly fee, the way they bought power and paid rent.
Some members of the group positively loved this idea; they argued that
IVK should not be in the business of running data centers or developing basic software systems. They allowed that IVK might want to run its own software in certain areas key to maintaining competitive advantage, but moving nonessential services to a vendor would make it easier, they believed, to focus on the IT services that provided competitive advantage. For most of what ServoLith proposed to do, IVK would never be as expert as the vendor, which had vast experience with many other clients and many more specialized resources than IVK could ever bring to bear.
Those who disliked the “software as a service” model proposed by ServoLith tended to dislike what they saw as the loss of control inherent in the model. The vendor would keep IVK data inside vendor facilities and equipment, and control IVK’s access to it. If the vendor proved unresponsive, the result could cripple IVK. In addition, part of the ServoLith strategy of operating this way appeared to be based on their intention to offer solutions developed for one client to their other clients. This could mean, in theory, that IVK-specific functionality would be marketed by
6
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
Vendor Partnering
Do
No
tC
op yo ServoLith to IVK competitors, although the vendor had signaled flexibility in this matter—proposing waiting periods, for example—before remarketing IVK custom features and functionality.
VerxaWeb proposed a solution with very different features. Their representatives went to great pains to reassure IVK managers that they would remain in charge of their own systems and operations. VerxaWeb proposed installing IVK systems on IVK equipment running in
IVK data centers (or space IVK had rented from a third party). VerxaWeb’s proposal included an option to provide operations staff who would manage IVK systems and equipment inside IVK walls, but it was just an option, and it was as far as their proposal went in that direction.
The proposal did not, of course, rule out any future outsourcing, but it represented, on the whole, a more modest service delivery proposal than what had been offered by ServoLith, and one that was also priced more moderately. And VerxaWeb explicitly assured IVK that they would not remarket solutions to IVK competitors. They promised, in fact, not to work for any of IVK’s top competitors.
HighOSoft left the question of service architecture completely open.
Their proposal had assumed that IVK would be contracting separately with other partners, and the service architecture would be determined by how IVK decided to set up these partnerships. Whatever IVK wanted to do about partnership and contract arrangements would be okay with HiOSoft; they were mainly looking to sell a software package. The only aspect of the solution on which this vendor was inflexible: the broad technical specifications of the service-oriented architecture at the center of their proposal, which set up a standard framework that all the components of the solution would plug into. This approach provided IVK with maximum flexibility, but also maximum relationship overhead. It risked forcing many decisions back onto IVK and causing the same kind of paralysis that had been the norm in the NetiFects era.
Unlike the solution proposed by the other two finalists, this one did not explicitly offer “one throat to choke” in its support arrangements. In other words, if something went wrong in the ServoLith model, there’d be only one party to call: ServoLith. The VerxaWeb proposal had a similar feature, and they would likely feel the grip of IVK’s metaphorical fingers much more pressingly than would ServoLith. HiOSoft would
7
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
The Hero Breaks Through
tC
op yo not object to IVK setting up a single point of contact for service, a single throat to choke, as part of its web of involved partners, but this arrangement wasn’t inherent in their proposal.
Barton was not at all confident that he knew the solution to all these problems, but he grew increasingly impatient with the debate. He worried, too, that none of the vendors might be able to do a really good job on such a big infrastructure replacement, at least not on schedule and budget. Barton had read the grim story of FoxMeyer Drug, a $5 billion pharmaceutical distributor that had gone out of business when it switched over to a replacement operating infrastructure and discovered that the new one couldn’t handle the required volume of transactions.
And that was just the tip of the iceberg. He’d also heard horror stories about candy companies unable to deliver shipments for major holidays and legal battles between companies over whose fault some major failure might be.
As the latest acrimonious meeting ended, Barton thought far too much remained up in the air. It felt to him like a form of torture, following a discussion headed nowhere without being able to intervene.
But it was torture of his own making, the result of his own decisions that had seemed right at the time but now cast a long dark shadow. Perhaps he had acted too hastily. Perhaps. But what was done was done.
Barton believed he now needed to do something really hard but really essential—trust the team to work through its own difficulties.
Do
No
Monday, October 1, 11:00 a.m. . . .
As Barton stepped into Maria Navarro’s office for their scheduled meeting, he felt himself relax. Maria had that wonderful ability to seem like anyone’s confidante. Probably that was a major reason for her success as an HR executive. Certainly it worked with Barton, even though he knew he should be wary. The HR VP was also involved in firings, demotions, and other adverse outcomes. She was, Barton knew, very tough under that warm exterior. But she and Barton had known each other for a long time and had worked together often, so he tended to feel at ease with her.
8
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
Vendor Partnering
Do
No
tC
op yo “How’s it going, Jim?” she asked, as she motioned him into a chair.
He shut the office door before he complied.
“It’s getting better,” Barton said.
“Williams has been pretty hard on you, I imagine.”
“It’s been tough with Carl for a few months, but I’m pressing on. It’s getting better,” he repeated, not altogether convincingly.
“How about with your peers?”
“You mean do their conversations still stop when I walk up on them?
Yes, sometimes. Less often than a month or two ago. I really think we’re building back some credibility.”
Navarro, in that maddening way of HR managers, was noncommittal. She looked sympathetic, but offered no corroboration that would stake her to a position on how Barton was doing. Then she changed the subject, which was just as well.
“You’ve got a hiring issue you wanted to talk about, yes?”
“Yes,” said Barton. “You remember the NetiFects situation.”
“You terminated their contract, as I recall,” said Navarro.
“I did. A decision with aftershocks I’ve been suffering personally.
Among the issues we’ve had to deal with in the aftermath of the divorce: we lost access to some of their very good people. One of them,
Ash Srinivasa, is now signaling to us his availability and desire to work for IVK. We’d like to hire him.”
“What’s the situation with NetiFects? Is that resolved?”
“No. They’re still maintaining that we owe them several hundred thousand dollars. I’m arguing that they failed to perform. It won’t come to anything. They’ll keep billing us, we’ll keep refusing to pay. But it won’t go any further.”
“Unless there’s an issue around hiring one of their key people.”
“Exactly. Sometimes hiring from a vendor is no big deal. Some vendors expect a certain amount of that. Others make you agree not to do it. NetiFects is the latter kind, although we no longer have a contract with them.”
“What kind of contract does—what’s his name?”
“Srinivasa.”
“What kind of contract does he have?”
9
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
The Hero Breaks Through
Do
No
tC
op yo “I think there’s something in it saying he can’t go to work for a client for some period of time without NetiFects’ official permission.”
“Probably not enforceable. But it could result in a skirmish. How good is this guy?”
“By all accounts, he’s very, very good. Worth a skirmish, probably.”
“I think we probably go for it, then. They’re likely to try to intimidate him into not taking the position. They’ll go at him, not us. Or they might just let the whole thing slide. But let me ask you another question,” Navarro said. Barton nodded, and she continued: “Why do we need someone this good?”
“You can never have enough people this good,” said Barton. “They’re very rare.”
“But aren’t we headed toward more outsourcing in the future?”
“Probably.”
“How many very, very good people will we need in that future?
Won’t we just need smart contract managers?”
“It’s a good question,” conceded Barton. “I know my guys think we need expertise in-house even when we outsource, so that we can manage effectively. In a way, that’s what went wrong with the NetiFects relationship. We didn’t have people on it who could supervise their work effectively, and they were taking advantage of that.”
“But will we even be able to retain people like this Srinivasa in the long run? How interested will he be in a job that’s mostly contract management?”
“Also a good question. My guess is we’ll always have some need for really good technical specialists, but I’m not sure how we stay attractive to them if we outsource more of our IT. We need to think about a future when we’ll manage more vendor contracts. What does vendor management look like in the long run? What kinds of people will allow us to do it effectively? We’ll need to work our way into answers for questions like these, by trying different things, seeing how they work.
Minimizing the cost and risk of our experiments. Meanwhile, I think we could use this guy. My people think he’d really be able to help us.”
Navarro nodded again. “Let’s move forward with it then. Just let me know if you need any more help from me.”
Barton stood. “I will.”
10
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
Vendor Partnering
Do
No
tC
op yo “Take care of yourself, Jim.” She winked at him. Barton felt his defenses falling, then made a deliberate effort to pull them back up. “And use some of that vacation time you’re accruing,” she called after him, as he started down the hall.
“I’ll do that. Thanks, Maria.”
Vacation? Geez, it’s already October! thought Barton. He hadn’t meant to wait so long without going somewhere with Maggie. When he’d cancelled their summer vacation—two weeks in Italy—he was amazed she hadn’t dumped him. But he just couldn’t see leaving IVK at such a critical time. Who knew what kind of revolt might have arisen in his absence? In a particularly memorable nightmare, Barton had returned from
Europe to find Davies reinstated as CIO, sitting in his chair.
In the first few months they’d dated, Barton and Maggie had mastered the romantic getaway: skiing in Aspen, cruising to Cabo; even an impromptu weekend in Paris. They’d always compensated for their awkward schedules, the time apart necessitated by two high-powered careers, by taking really fabulous trips to exotic places when they could get away.
Now they were long overdue for one of these. Hadn’t Maggie just said something about scuba diving in Bermuda? Belize? Maybe he could make something happen before the month was over. The matter definitely needed tending, Barton thought. Just the possibility of a vacation lightened his step as he headed back to his office to do some research.
11
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
rP os t
The Hero Breaks Through
REFLECTION
Should acquisition of infrastructure replacement services be an arm’s-length transaction or a close partnership transaction?
Which vendor, contract structure, and service delivery model should IVK choose? op yo How important is “control”to a company like IVK when outsourcing IT? How should control be maintained?
Do
No
tC
How should IVK hire to support contract management? Will very, very good technical employees be needed in a future characterized by high levels of IT outsourcing? 12
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
Vendor Assessment Matrices
rP os t
Vendor Partnering
Percentage of fit
Functional criteria
HiOSoft
VerxaWeb ServoLith
Electronic application
100
100
96
98
98
99
97
Credit scoring
82
Decision processing
75
Workflow management
78
Customer service interface
97
Financial systems interface
86
100
97
100
98
100
87
94
89
100
100
100
Design for robust operation
94
93
99
Scalability
94
91
100
88
89
95
op yo 94
Client management functions
Disbursement
Other
Rating
HiOSoft
VerxaWeb ServoLith
Long-term relationship potential
M
H
Research and development capability
M
MH
H
Understanding of IVK culture and challenges
M
H
M
Financial viability
M
M
H
Openness of architecture
H
M
M
Cost
M
ML
H
Technical support capabilities and plan
M
M
M
Dependence on subcontractors in proposed solution H
M
L
Compatibility with existing IVK infrastructure
(DBMS, etc.)
H
MH
M
H
M
M
Do
No
tC
Qualitative criteria
Compatibility with existing IVK applications
M
Confidence in project management capability
ML
M
MH
Contract flexibility
H
H
L
Likely attentiveness to IVK concerns
M
H
M
Training capability and plan
M
H
M
Financial service industry experience
M
MH
M
Intangibles, comfort factor
M
H
M
Market position
Second
Third
Leader
13
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
THE ANSWERS YOU NEED,
rP os t
WHEN YOU NEED THEM
DOWNLOAD
BOOK
CHAPTERS
NOW
op yo NOT ALL BUSINESS
CHALLENGES ARE
CREATED EQUAL.
Some require detailed analysis and others demand a thoughtful solution—but in a quick and easily accessible format.
Now you can get instant access to the answers you need by downloading individual chapters from our most popular books, including:
Blue Ocean Strategy by
W. Chan Kim and Renée Mauborgne
Leading Change by John P. Kotter
Groundswell by Charlene Li and Josh Bernoff
And many others
The solutions to your toughest challenges are just a click away.
Do
No
tC
LEARN MORE ABOUT HARVARD BUSINESS PRESS CHAPTERS:
www.harvardbusiness.org/press
This document is authorized for use only by Paolo Neirotti until September 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860.
