analysis a case

________________________________________________________________________________________________________________ Neeraj Bharadwaj, Visiting Assistant Professor of Marketing, University of Chicago Graduate School of Business, and John B. Gordon, Director of Channels Marketing for IBM SOA and WebSphere Software, prepared this case solely as a basis for class discussion. This case is not intended to serve as an endorsement, a source of primary data, or an illustration of effective or ineffective management. The authors thank Mike DeRocco, Diane Badame, Greg Bonner, Derrick Boone, Craig Chapman, H. Sujan, and Mike Hanssens for their insightful feedback. This case, though based on real events, is fictionalized, and any resemblance to actual persons or entities is coincidental. There are occasional references to actual companies in the narration. Copyright © 2007 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685, write Harvard Business Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. This publication may not be digitized, photocopied, or otherwise reproduced, posted, or transmitted, without the permission of Harvard Business School.

N E E R A J B H A R A D W A J

J O H N B . G O R D O N

Atlantic Computer: A Bundle of Pricing Options

I. Overview

On an unseasonably cold afternoon in Amberville, New York, in early October 2000, Jason Jowers leaned forward in his chair, impatient for the meeting to end, so that he could get started. The youngest product manager at Atlantic Computer, a large manufacturer of servers and other high-tech products, Jowers had joined the company just four months ago as a newly minted MBA. Now, fresh from a management rotation that had included projects in product management, strategy, and business development, Jowers was getting his first shot at some real responsibility. He would be responsible for developing the pricing strategy for the "Atlantic Bundle" (i.e., the new Tronn server and the PESA software tool).

Jowers knew that it would not be easy to find the sweet spot, but he was confident and eager for the challenge.

There were just four of them at this marketing kick-off meeting: Jowers; Chris Matzer, head of the server division; Emily Jones, director of the division’s R&D team; and Harry Fowler, director of new product marketing.

Matzer had started the meeting with a brief overview of the new server. The Tronn, he explained, had been developed specifically to meet an emerging U.S. marketplace opportunity. “But,” he had said, “We really feel that the key to making this product a success is going to be our ability to sell the server with our new software tool.” Jones had taken it from there. The tool, called the “Performance Enhancing Server Accelerator,” or PESA, would allow the Tronn to perform up to four times faster than its standard speed, she explained. It was specifically designed to make frequently requested information (e.g., files, data, pictures, etc.) extremely accessible. “The ‘Atlantic Bundle’ is the sale we want.”

Next, Fowler went through a slide presentation, reviewing key industry characteristics and projecting sales for the server (see Exhibit 1 for projected sales). “Ultimately, there are two main market segments in the server industry,” he said, clicking to a new slide. “The largest segment, High Performance Servers, represents the traditional use of servers to run complex applications. We’re

2078 M A Y 2 8 , 2 0 0 7

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.

2078 | Atlantic Computer: A Bundle of Pricing Options

2 BRIEFCASES | HARVARD BUSINESS SCHOOL

talking supply chain management, enterprise resource planning, business intelligence, that ilk. This segment is expected to demand approximately 200,000 units next year and is projected to grow around 3% annually the following two years.”

He clicked over to another slide. “Atlantic has been a strong player in this segment and has captured 20% of the revenue market share. Emerging with the growth of the Internet, however, is a newer segment for Basic Servers.” He paused, looking at Jones. “It’s not all the bundle, Em; we all know that.” Fowler looked back at the slide and continued his presentation. “Companies increasingly need basic computing capability to perform simple, repeatable tasks, such as showing website information on the Internet. This market should demand about 50,000 units in 2001, and register about a 36% compound annual growth rate through 2003.”1

Jowers nodded, taking notes on the hard copy of the slide deck he had in front of him. He wished that Fowler would wrap it up; thankfully, he appeared to be winding down.

Matzer cleared his throat and jumped in. “Thanks, Harry. So that’s where we stand. This was a useful review. It’s good to know that we’re all on the same page as we go forward.” He turned to Jowers. “Jason, you have an opportunity to make a big splash here. Our new server, the Tronn, and the PESA software tool are going to help us enter a new market—one that has been dominated by Ontario Computer for longer than I care to remember. “But we need to get our pricing right for the ‘Atlantic Bundle.’ You're going to have to figure out which businesses are most likely to benefit from our new offering. You will also need to take into consideration how customers are likely to react and how competitors are likely to respond to your recommended pricing strategy.”

He looked to Jones and Fowler for agreement, and seeing their nods, continued talking to Jason. “Let’s set a deadline two weeks out. I’d like to see a pricing structure in place for the trade show at the beginning of November.” Jowers nodded as well. He knew that Matzer would want him to be ready for that show. Officially called “The Small and Medium-Size Enterprise Systems Solutions Trade Show”—SME for short—this event was a big deal for Atlantic Computer. Getting their pricing strategy down cold before the show was a must.

II. The Necessary Background

The meeting adjourned; Jowers headed back to his office. Seated in front of his own computer, he pulled up some secondary research that already existed within some internal company reports. He decided that it couldn’t hurt to take some time and just review the market—and Atlantic’s piece of it—in broad strokes. In business school, he had often found himself caught up in the details of one or another case, and had been pulled up short, more than once, by a professor saying, “Jason, you’re clearly winning the battle, but you’ve already lost the war. Get your head up and look out, look back and look forward. You ignore the big picture at your peril.”

He first reviewed a document detailing the company's history. Atlantic Computer, Inc., was the largest player in the overall computer industry. It had been competing in the server market for 30

1 We offer some information on servers for the nontechnical reader. For one, servers are just big computers. Just like personal computers run software for individuals, servers run software that is shared by a number of people. For example, when many different people visit a given website, servers assist in managing the interface between each visitor and the website. When customers wish to check their account balances at an ATM, the banking software that tracks account balances across a large set of customers runs on servers. Second, servers differ in terms of the operations that they perform. Some servers are basic. They perform simple, repeatable tasks such as displaying the same information to everyone who goes to a website (e.g., www.espn.com). Others are more complex (i.e., high-performance) and may run highly sophisticated models such as calculating the age of the universe or the makeup of DNA.

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.

Atlantic Computer: A Bundle of Pricing Options | 2078

HARVARD BUSINESS SCHOOL | BRIEFCASES 3

years by selling its high-end performance servers, called Radia, to large enterprise customers. The firm in general and its Server Division in particular had come to be known for providing top-notch, highly reliable products, and had developed a reputation for providing high quality, responsive post- sales assistance. This was a direct result of their overarching strategy being based on customer intimacy and product differentiation.

He then read a memo written by Matzer conveying that the Internet had yielded a market for low- end systems, of course. Although the server market was historically characterized by complex, robust systems capable of handling critical business applications, the Internet and the proliferation of applications in the late 1990s had changed all that. For instance, a company could now build a website by utilizing several smaller basic servers that shared the work.

Jowers pulled up a report from March of 1999, also prepared by Matzer. Given the projected growth for low-end, basic servers, Matzer (a 20-year veteran of the computer business) had decided to stretch the division’s product line downward with the goal of introducing a model called the Tronn by the end of 2000. His strong belief was that high performance servers and basic servers would not be viewed as substitutes by customers.

Matzer had focused a large chunk of his report on competitors—Ontario in particular. Jowers reviewed this section with interest. Then, on top of Matzer’s report, he opened up Atlantic’s current internal analysis of Ontario and the basic server market.

Ontario Computer, Inc., a firm that concentrated its efforts solely on the low-end server market with its Zink product line, currently claimed 50% revenue market share in the basic server market. Ontario’s Zink servers performed at approximately the same level as Atlantic’s Tronn. The rest of the basic server market was made up of many smaller vendors. Unlike Atlantic, the majority of Ontario’s sales were generated online. Jowers made a note of that.

Given that Ontario’s business model was based on operational excellence, the company had been able to drive out many nonvalue-added costs and, therefore, compete largely on price. Allen Capps, CEO of Ontario Computer, had been heard saying, “Our business model is not to be the leading innovator in product technology. Rather, our business model is to provide leading technology to customers via the most flexible and innovative supply chain strategy possible.”

Jowers opened a file on the PESA’s beta field tests and reviewed the results.

The bottom line was that Atlantic’s Tronn was going to compete directly against Ontario’s Zink. And the beta test had confirmed that the PESA allowed Atlantic’s low-end servers to perform up to four times faster than their standard speed when loaded with the PESA software tool. That meant that a business customer could conceivably receive the same level of performance by buying one Tronn loaded with the PESA as compared to buying four basic servers. Jowers knew that it was a no- brainer to buy Atlantic’s offering, but he also knew that getting the pricing right was crucial. Customers were funny when it came to weighing short- and long-term investments. What was the right strategy?2

2 In this case, software tools are being distinguished from software applications. While many business customers see value in paying for software applications (such as financial systems, procurement systems, etc), this is not always the case for software tools (such as those that are intended to monitor the health of a server or enhance the server’s performance). The widely held belief is that software tools are just “part” of the server and should be provided for free.

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.

2078 | Atlantic Computer: A Bundle of Pricing Options

4 BRIEFCASES | HARVARD BUSINESS SCHOOL

III. Marketing the Tronn Loaded with the PESA

There were several key factors influencing the pricing strategy for the “Atlantic Bundle” that Jowers had to keep in mind. First, he thought about the Server Division’s traditional focus on hardware. This division had placed only limited emphasis on developing and selling software tools that helped to enhance the performance of their servers.

For example, when the division had designed a small software management tool to help server administrators monitor the health of their systems and be more responsive regarding repairs, that tool had not been viewed as a differentiator to win deals because the sales force had historically given away software tools.

That approach reflected Matzer’s belief that software tools should generally be provided to customers for free. Jowers wasn’t at all sure the PESA should be handled the same way. Still, Matzer was a force to be reckoned with.

Second, the division had long relied upon cost-plus pricing analysis (the standard approach to pricing in the industry) to determine the prices of their servers.3 For the Tronn, Matzer had followed convention and established the price of the basic server at $2,000.

Third, Jowers realized that establishing the gains accruing to customers from the PESA software tool would require some careful consideration of customer segments. Based upon internal performance test results (see Exhibit 2), customers in the web-server and file-sharing application segments appeared to be the ones that would benefit most from the tool.

Jowers thought back to a quick conversation he’d had with Matzer the previous day. Matzer had come up to him at the coffee machine, and the talk had quickly (and not surprisingly) turned to the Tronn.

“The thing is, customers will usually dedicate their basic servers to one application,” he had said. Upon reflecting on some of Matzer’s other comments, Jowers recognized the importance of conveying to prospective customers that the first-order savings effects from purchasing the “Atlantic Bundle” derived from the need to purchase fewer servers, and the second-order savings effects included lower annual electricity charges, software license fees, and labor costs (see Exhibit 3 for cost and pricing information).

Jowers closed all of the reports arrayed on his desktop and sent a quick email to Jairo Cadena, director of sales. The reply was back almost instantaneously, confirming what he had already known. Cadena was looking forward to manning Atlantic Computer’s SME trade show booth with Jowers, and was also eagerly awaiting Jowers’s plan. He would also be happy to meet with him any time and act as a sounding board.

Jowers read Cadena’s reply and smiled, shaking his head. Cadena wouldn’t be just a sounding board, he knew. Still, it would be good to hear what he had to say. Truth be told, Jowers wasn’t sure at all why Cadena hadn’t been included in the marketing meeting. He would have to ask Matzer offline whether that was a significant omission.

3 Cost-plus pricing is a method of determining the price of a product or service that uses direct costs, indirect costs, and fixed costs whether related to the production and sale of the product or service or not. These costs are converted to per-unit costs for the product, and then a predetermined percentage of these costs is added to provide a profit margin. The resulting price is cost per unit plus the percentage markup.

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.

Atlantic Computer: A Bundle of Pricing Options | 2078

HARVARD BUSINESS SCHOOL | BRIEFCASES 5

He thought about the company’s approach to sales. Traditionally, the Server Division had relied upon a high-touch direct sales channel. Cadena’s sales force compensation structure was roughly a 70% salary and a 30% commission. Clearly that fact was going to color Cadena’s view of the pricing strategy. Maybe that was it. Still, it was important to hear Cadena’s point of view; his people would eventually be the ones primarily in charge of generating demand, fulfilling demand, and providing post-sales assistance.

At the trade show, though, Jowers would get a shot at talking to prospective customers himself. He would be the point person assigned to discuss the new server and software tool with visitors at Atlantic's SME booth. The information gained from the trade show would allow Cadena to eventually develop a sales script and other marketing collateral (e.g., sell sheets) that his sales staff and his third-party business partners could leverage when offering the servers to small and medium- sized business customers. But he had to go into the trade show ready to test a theory; otherwise, they wouldn’t learn anything tangible.

IV. Planning the Strategy

The phone rang; it was his boss, Harry Fowler, offering his assistance. “Jason I don’t yet know you well enough to know how you prefer to work, but I usually find it useful to think out loud. If you’d like to take some time and brainstorm in front of a white board, I’d be glad to help.”

Jowers thanked him and suggested meeting early the next morning. No lack of people willing to pitch in, he thought to himself. But having too many opinions at this stage wouldn’t be useful, either for the Atlantic Bundle or for Jowers’ career. Everyone wanted to help; everyone wanted to influence his thinking. Not that that was a bad thing, but it was clear that tradition and position were flavoring how Jones, Fowler, Matzer, and Cadena were thinking about the Tronn and PESA. Jowers had to be the one with the clear head.

Jowers began to contemplate the major decision for his October 15 presentation to Matzer and the management team. He thought that it would be most effective if he were to identify an exemplary customer from the most viable segment in discussing his proposed pricing strategy. In reviewing some of the sales leads that Cadena had passed along, Jowers decided that DayTraderJournal.com (see Exhibit 4 for a customer profile) could serve as an exemplary customer that would be visiting their sales booth. From the completed sales lead sheet, it was evident that this customer was seeking four basic servers, and that minimizing initial purchase costs and subsequent possession costs were two very closely ranked considerations (i.e. "key buying criteria"). Jowers started to contemplate the first-order and second-order savings that this customer would enjoy by purchasing the Tronn server in conjunction with the PESA as opposed to buying four basic servers. He knew that Matzer and his colleagues were quite conservative, so he decided that it would be best to compare two basic servers loaded with the PESA software tool versus four basic servers.

He thought about the alternative approaches that could be utilized to develop the pricing strategy for the Atlantic Bundle for the exemplary customer. The “usual suspects” included status-quo pricing, competition-based pricing, and cost-plus pricing. Then there was value-in-use pricing—an alternative he’d read about in a working paper just a few months ago.4 This latter option would

4 Please see “Customer Value Propositions in Business Markets” appearing in the March (2006) issue of Harvard Business Review. In that article, Anderson, Narus, and van Rossum outline the importance of being able to demonstrate and articulate the savings to the customer from a vendor’s offering and provide a means by which to do so. They offer that a value word equation “expresses in words and simple mathematical operators (for example, + and ÷) how to assess the differences in functionality or performance between a supplier’s offering and the next best alternative and how to convert those differences into dollars.” (p. 96)

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.

2078 | Atlantic Computer: A Bundle of Pricing Options

6 BRIEFCASES | HARVARD BUSINESS SCHOOL

require him to demonstrate the savings to the customer from purchasing the Atlantic Bundle versus buying four Ontario Zink servers (i.e., identifying the savings the customer would realize from having to buy fewer servers, utilize less labor and electricity, and secure fewer software application licenses).

Jowers stood up, walked around his desk, and tried to stretch out his back. Ultimately, he thought, he could go one of four routes:

1. Stick with company tradition by charging only for hardware and give the PESA software tool away for free.

2. Charge a price equal to what the customer would pay for four Ontario Zink servers.

3. Charge a price based on a cost-plus approach to pricing PESA (based on software tool’s development costs).5

4. Charge a price based on value-in-use pricing.6

He sat back down to identify the critical tasks that he needed to complete. First, he needed to calculate the price that he should charge DayTraderJournal.com for the Atlantic Bundle utilizing the four aforementioned approaches. He would then have to recommend one of those alternatives for this customer. This would require him to not only convey the pricing strategy that would optimize value capture for Atlantic, but also describe any implementation issues that could possibly arise. For example, he recalled that the industry norm was to give away software tools. If he were to recommend one of the other alternatives, he’d better be ready to answer how he could break from what was already a well-established tradition in the industry. In addition, he knew that he would have to think more broadly about the top-line revenue implications from each of the four alternative pricing strategies. He knew that he could undertake a quick “back of the envelope” calculation utilizing the total estimated sales for the Atlantic Bundle over the upcoming three years to effectively convey the gains from his recommended pricing strategy.

The second agenda item would be to sketch out for Cadena how to get the Division’s sales force to charge for (and not give away) the PESA. What could he recommend to get Cadena’s hardware- oriented sales force to understand and sell the value of the PESA software tool effectively?

He recalled again the working paper he had read on value-in-use pricing. According to the authors, business managers in the United States and Europe were reporting that the phrase, “We can save you money!” had become a generic blanket statement. Almost every prospective supplier was saying it. Jowers thought about the possibility of developing a simple role-playing exercise in order to convey to his senior management team how an Atlantic salesperson could effectively demonstrate the value created for DayTraderJournal.com, and overcome the potential objections that the customer was likely to raise. He knew that his calculations could easily be generalized to evaluate the savings

5 For the cost-plus approach, some assumptions will need to be made about the expected sales volume, the PESA attach rate, the time period, and the margin. Given Atlantic’s production constraints, the firm will only be able to produce a limited number of basic servers in the near term. Assume that the firm will be able to sell all of the Tronn servers it can produce, and that Atlantic’s resulting share of the basic server segment (in units) will be 4% in 2001, 9% in 2002, and 14% in 2003. On these shipments, assume a 50% attach rate (i.e., half of all of their basic servers sold will be loaded with the PESA) since this is an entirely new concept and some basic servers are used for applications that will not benefit from PESA. Assume that Atlantic’s software development costs for the PESA will be paid off over three years. Last, target a 30% markup above costs. 6 Value-in-use pricing is a method of setting prices in which an attempt is made to capture a portion of what a customer would save by buying a firm's product. For this case, please assume a 50-50 sharing of the savings gain with the customer. Also, please base your calculations on one year of savings (e.g., annual electricity savings equal $250 (Exhibit 3)). Although the average life of an Atlantic basic server is estimated to be three years, please use the conservative per annum estimate.

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.

Atlantic Computer: A Bundle of Pricing Options | 2078

HARVARD BUSINESS SCHOOL | BRIEFCASES 7

likely to be achieved by other customers who were seeking to utilize the basic servers for running websites or file sharing would experience.

The final item that he needed to address in his presentation stemmed from a comment made at the orientation meeting regarding anticipating the main competitor’s likely reaction. He thought that he could address that by preparing a response to the following question: How is Ontario Zink’s senior management team likely to react to the threat posed by the Atlantic Bundle?

Jowers stood up once again, grabbed his jacket, and headed toward the door. It was time to get a sandwich and a beverage and contemplate the situation. He had a good outline, he felt. But now the hard work was going to begin. It was time to start filling in that outline with details. Over the next week or so, he was going to have to make some decisions.

Exhibit 1: Projected Market Volumes by Segment (in units)

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

2001 2002 2003

High Performance segment

Basic segment

50,000 70,000 92,000 Basic segment (Units) 200,000 205,000 210,500 High Performance segment (Units) 250,000 275,000 302,500 Total Estimated Market for Servers (Units)

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.

2078 | Atlantic Computer: A Bundle of Pricing Options

8 BRIEFCASES | HARVARD BUSINESS SCHOOL

Exhibit 2: Test Results from PESA Software on Tronn Servers on a Variety of Application Types a, b

Atlantic Internal Performance Testing Results July 22, 2000

Tronn Tronn Application Tested with PESA without PESA

High Performance / Compute Intensive Workloads 180 187 (e.g., running complex scientific models)

Graphics Applications 280 281 (e.g., creating custom graphics / video)

Enterprise Applications 264 254 (e.g., supply chain management)

e-Mail Applications 341 300 (e.g., Microsoft Outlook, Lotus Notes)

File Sharing 812 404 (e.g., shared storage and data backup applications)

Web Servers 2222 542 (e.g., running websites such as www.espn.com)

a The performance results above capture the test output that Jones’s team received when evaluating the Tronn server running various applications both with and without PESA software loaded. The ordering of the applications in Exhibit 2 is from the most complex server application to the least. We note that the numbers appearing in the two right-hand columns per se are not important for this case. Rather, it is the ratio between the results with and without PESA that is most meaningful. Thus, a score of 200 with the PESA in relation to a score of 100 without the PESA means that the server running PESA is performing at the level of two standard Tronn servers.

b As stated in the case, the PESA was specifically designed to make frequently requested information (e.g., files, data, pictures, etc.) extremely accessible. This means that those applications that require frequently requested information (e.g., web servers that just show the same web page to everyone who visits the website and file sharing) ought to benefit from having basic servers loaded with the PESA. In some applications, there is less repeated information so the benefits will not be as pronounced (e.g., email and enterprise applications). In other applications (e.g., graphics and high performance applications), loading the PESA onto the server can actually slow down the server because scarce computing resources have to be diverted away from the complex, unique calculations to run PESA; therefore, PESA may have detrimental effects when loaded onto high performance servers.

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.

Atlantic Computer: A Bundle of Pricing Options | 2078

HARVARD BUSINESS SCHOOL | BRIEFCASES 9

Exhibit 3: Salient Pricing Information

COST INFORMATION (per server) Electricity

(annual cost)

Cost of Application

Software Licenses

Labor (# of servers an "administrator"

can manage) Basic server $250 $750 40 High Performance server $400 $750 20

Atlantic Computer Price per Server Cost per Server Tronn (Basic) $2,000 $1,538 Radia (High Performance) $11,000 $7,586

Ontario Computer Zink (Basic) $1,700 $1,214

Other customer expense information:

• Cost of Electricity: Charge for heating and cooling servers, etc.

• Cost of Software Application Licenses: Assume application software is licensed “per server.” The average per server application software license is $750. Examples of Application Software are supply chain management software, customer relationship management software, web application software. PESA is a software tool, not application software.

• Cost of Labor: A server administrator's annual salary is $80,000.

R&D Costs for Atlantic:

• PESA Software development costs = $2,000,000

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.

2078 | Atlantic Computer: A Bundle of Pricing Options

10 BRIEFCASES | HARVARD BUSINESS SCHOOL

Exhibit 4: Customer Profiles

Sample “Web Server” Basic Server (New SME) Customer Profile:

Customer Name/Size: DayTraderJournal.com (15 employees)

Address: 11325 Magic Bean Drive, San Jose, CA

Business Description: With the increasing popularity of day trading fueled by the growth of the Internet, Day Trader Journal has identified an unmet market need for providing training information and tips to prospective day traders. Day Trader Journal online will generate revenue by selling advertising space to other companies who are attempting to reach the day trader market.

Server Need: Web servers to host the company’s new website, DayTraderJournal.com, where day traders can review articles and relevant training information.

Top 3 Requirements: (i.e., ranking of “key buying criteria” for basic server purchase) 1. Minimize our acquisition costs 2. Minimize our possession costs (i.e., total cost of ownership) 3. Allow our website to process many information requests

Sample “File Sharing” Basic Server (New SME) Customer Profile:

Customer Name/Size: Look Sharp Advertising (65 employees)

Address: 580 Madison Avenue, New York, NY

Business Description: Look Sharp Advertising designs customer-facing marketing communications materials such as brochures, websites, and advertisements. It works primarily with business clients building integrated marketing communications programs.

Server Need: File servers to help layout designers share graphics, text, and layouts.

Top 3 Requirements: (i.e., ranking of “key buying criteria” for basic server purchase) 1. Minimize our possession costs 2. Make highly requested information among employees readily available (e.g.,

graphics and images) 3. Availability of top-notch post-sales support

Sample High-Performance Server Customer Profile:

Customer Name/Size: Human Genome Institute for DNA Modeling and Mapping (HGIDMM) (100 employees)

Address: 100 Massachusetts Avenue, Cambridge, MA

Business Description: HGIDMM is attempting to map the human genome and model the potential effects of various experimental medications on humans. With the increasing expense of medical trials, big pharmaceutical companies are looking for better ways to evaluate the potential efficacy of new drugs before going to trial. HGIDMM wants to help these companies expedite the process and thereby reduce new product development costs.

Server Need: They currently own a Radia to perform complex, one-time model simulations of the effect of medication on individuals with various DNA specifications.

Top 3 Requirements: (i.e., ranking of “key buying criteria” for high performance server purchase) 1. Capable of maximizing speed of mathematical computations in complex

model simulations. 2. Availability of outstanding customer support (24x7) 3. Flexible payment terms

This document is authorized for use only in Kyuseop Kwak's SPR_2019_24760_Pricing and Revenue Management at University of Technology Sydney from Jul 2019 to Sep 2020.