Read this article and Answer the following questions:

Timberjack Parts: Packaged Software Selection Project

On the morning of December 15, 1995, Jim Utting, general manager?Parts at Timberjack Corp. in Atlanta, pondered the decision that he and his project team were about to make. The project team, which included three members from Atlanta and three from a sister parts operation in Sweden, had been given the task of selecting the same packaged software to be used on both sides of the Atlantic. While both operations were very similar in terms of sales volume and number of personnel, Utting couldn't help but recognize the differences in opinion with regard to software requirements. While Utting had overall responsibility for the selection process and was in a position to make a final decision, he felt that his project team should come to consensus.

Clearly a consensus would be reached only by compromise, and the group had already made so many to get this far. While selecting software used by excellent distribution or wholesale companies was the goal at the outset, the requirements of strong local support on both sides of the Atlantic, as well as a UNIX platform forced the list to include mainly enterprise requirements planning software companies which specialized in manufacturing software. Sahlqvist, president of Timberjack Parts AB, frequently asked during demonstrations, "Do we have to start manufacturing to use this system?"

While Utting and Sahlqvist tried hard to form a single team, geography?and to a lesser extent, culture?had worked against this. As Utting explained: "When the Swedish group came over here, and we went out on site visits together, and spent time together, by the end of the week we would be one team. After a few weeks on our own, we would quickly drift apart."

Industry Background

Timberjack was the world's leading manufacturer of heavy equipment for the professional logger, with an overall market share of roughly 25%. Its owner was Rauma Oy, a Finnish conglomerate that was listed on both the NYSE and the Finnish stock exchange. In 1995, Timberjack had sales of 627 million USD, net profits of 88 million USD, and roughly 1,600 employees. Timberjack and a few other large competitors such as John Deere, Blount, Caterpillar, and Valmet, had maintained the dominant share of the market over the long term. Many smaller firms also participated with specialized equipment that met a niche market. The Timberjack group was formed through a number of acquisitions during the 1980s and 1990s.

The evolution of logging equipment had dramatically changed the way that logging was done worldwide. While felling trees with chain saws and using animals to haul felled trees was still practiced, these methods were disappearing. In North America, trees were cut down using machines called feller bunchers, then dragged to the roadside by skidders, where they were trimmed by delimbers, and then lifted on to log trucks using log loaders. In Scandinavia, harvesters would cut and delimb trees in the forest, and then a forwarder would carry the logs to the roadside. This method allowed the logger to selectively harvest the forest rather than clear cutting.

The industry was extremely cyclical, as it was directly tied to the prices of pulp and lumber, which were in turn highly dependent on the overall strength of the economy. During severe downturns in wholesale pulp and lumber prices, annualized machine sales could fall by 75% with very little advance warning. A typical downturn lasted one to two years, and often many of the leveraged dealers, contractors, and manufacturers would not be able to weather the storm. For this reason, many capital projects were completed during the prosperous years.

Timberjack Parts Operations

Timberjack operated two service parts operations to service its equipment worldwide; one in Marsta, Sweden, which was located just outside of Stockholm, and one in Atlanta, Georgia. Both facilities were opened in 1994 and involved the relocation of personnel from other locations. In 1995, each facility had approximately 35 employees and sales of roughly 35 million USD. The European operation sold most of its parts directly to company-owned retail stores in Scandinavia; in North America, Timberjack sold exclusively to third-party heavy equipment dealers. In Scandinavia, Timberjack's parts operation relied on a software system known as DAIM which was acquired in 1987 and ran on an AS400 platform. Timberjack in Atlanta relied on Hewlett-Packard's MM3000 software designed to run on HP3000 hardware. This system, originally acquired in 1981, included a custom-coded dealer module developed in the early 1990s. The dealer module allowed dealers dialup access to electronically transmit orders and inquire on the price and availability of parts using a PC acting as a dumb terminal. Roughly 80% of all Timberjack North American orders were placed on the dealer system.

The Need for New Software

In early 1995, Sahlqvist and the Marsta group were actively investigating the acquisition of new computer software to run their parts operations. The main impetus behind the decision was the instability of their existing systems. Heavy modifications of their source code over the years had led to frequent system failures. As a result, the users could not always rely on the data that was presented to them. For these reasons, Sahlqvist actively sought a new computer system, with the intent of installing the system by early 1996 at the latest.

At the same time, at the Timberjack North American headquarters in Woodstock, Ontario, Canada, Coopers & Lybrand was helping top management outline a future information systems strategy. This project included the integration of manufacturing, the dealer network, as well as parts and service. While Woodstock was in no hurry to replace all of its systems, it did face the year 2000 problem; more importantly, Hewlett-Packard had notified customers that it would no longer be supporting its aging MM3000 software. After several months on site in Woodstock, Coopers & Lybrand prepared a strategic IS plan. The strategy concluded that Timberjack would benefit by selecting best-of-breed software to run the various units within the North American operations. Rather than relying on one package to run manufacturing, parts, and service, it was proposed that a strong "distribution company package" would best serve parts, and a separate manufacturing system would best suit the factory.

The Coopers & Lybrand recommendations to the North American headquarters were subsequently presented to the board of directors in Helsinki, Finland. Mikko Rysaa, president of Timberjack Group, concluded that a more global software strategy would be beneficial to Timberjack. The headquarters office in Helsinki continually struggled to compare financial results of the various units, with only marginal success. One financials package would help solve this problem. Integrated systems implemented worldwide would also make it easier to send employees overseas for a few years, as there would be no need for them to learn new computer software. Global selection would also add leverage to negotiations with software and hardware vendors. For these reasons, Rysaa felt that Timberjack should embark on a global software strategy.

The service parts organizations were subsequently given the task of selecting one system to run the parts businesses in Atlanta and in Marsta. Rysaa appointed Utting to head up the project, and instructed Sahlqvist to halt the process of choosing new software independently. Instructions to Utting were fairly clear; he was to gain consensus on one package for both sites, and he was told "There will be no disasters." If the two groups failed to reach consensus, then the Timberjack board would decide. Rysaa had warned Utting in his usual dry humor, "The intelligence level does not always go up when the decision is made by the board."

Over the next week, Utting and Sahlqvist decided on project team members, as well as a rough timetable. On April 17, 1995, the joint project team had its first meeting in Woodstock, Ontario. The team members present were Christine Smedjer, IS manager in Marsta, Jorma Nikkinen, materials manager in Marsta, Mark Gonzalez, information services coordinator for service parts, Darryl Romanow, materials manager in Atlanta, as well as Utting and Sahlqvist. Coopers & Lybrand appointed Ilya Bahar to the project, although Timberjack had not yet decided on the extent of the Coopers & Lybrand involvement.

Both organizations were in considerable transition at the time the project began. In April 1995, the team members were at various stages of relocation to the new parts facilities, in Atlanta and Marsta. Utting, Romanow, and Gonzalez lived near the Woodstock, Ontario factory. Utting was to relocate to Atlanta in May; Romanow was moving at the end of June; while Gonzalez had not yet decided whether he was going to move to Atlanta. While the new parts facility in Atlanta was opened in February 1994, it consisted of the distribution center staff, one buyer, and three accounting staff, all of whom were new to Timberjack. Marsta was in a similar situation, as Timberjack had closed its parts operations in Tampere, Finland, and Alfta, Sweden, and combined the facilities into one in Marsta. While Smedjer was recently hired from the Marsta area, Nikkinen continued to commute on weekends from Tampere (central Finland), and Sahlqvist commuted from Gotenborg, a three-hour drive from Marsta.

In the spring of 1995, the North American operation had the core group of parts buyers, customer service, and accounting staff still working in Woodstock. Although most were offered positions in Atlanta, very few were contemplating the move. Most members of the group were actively searching for new positions in the factory operation. Utting felt that he had a narrow window of opportunity to gain input from this group of people who averaged 20 years' seniority. Utting's primary concern was that the commitment to the project might be limited as many of the people would be leaving the parts department.

At the end of the first meeting, it was decided that Bahar would be invited to bid on the software selection contract and that the requirements definition would be done in Woodstock, with a sign-off in Marsta. Romanow and Gonzalez would work with Bahar during May and June, with a request for proposal (RFP) to be mailed to vendors in late August 1995. In order to free up time for Romanow to work on the project, Utting decided that an interim supervisor, reporting to Romanow, would remove him somewhat from the day-to-day activities. Jim McGregor, a 40-year employee and a senior buyer in the parts department, was given the task. McGregor had already decided that he was not moving to Atlanta and that he would likely retire at the end of the project.

RFP Process

In early May, Utting received a quotation from Bahar, which included the services to be offered and a fee schedule. The bid covered the RFP creation, list of potential vendors, vendor and site visits, consultation on the final selection decision, as well as advice on negotiation of the contract. The bulk of the work would be during the RFP creation, as a number of Coopers & Lybrand specialists would be involved in the process. Bahar guaranteed that the fees, including travel costs would not exceed $100,000 CDN (75,000 USD) which would then be split between Marsta and Atlanta. While Utting and Sahlqvist thought that this fee was high, both realized that the team could use the expertise as well as the resources necessary to choose the right software. Utting signed the contract, and work started soon after.

Bahar, Romanow, and Gonzalez put together a schedule of meeting dates to complete the majority of the work on the RFP prior to the middle of June. Sessions were to be scheduled with several members of each functional area within the parts department. The sessions were scheduled to take a half-day for the initial meeting, a half-day to review procedures from the first session and to develop a comprehensive requirements list, and a series of short follow-up meetings to finalize requirements and gain approvals. The final draft of the RFP was to be ready by mid-August, with a final review set for the week of August 25, 1995.

During the first meeting, three or four functional representatives would outline all the business processes used at Timberjack within their area. One Coopers & Lybrand expert would be the facilitator, and an alternate kept notes. Gonzalez and Romanow, who attended all of the sessions, would work with the group to try and ensure that all processes were covered in enough detail that formal requirements could be drafted. Special attention was paid to the most important processes at Timberjack, regardless of whether or not the functionality existed in standard packaged software. Each group, and its manager, was also asked to consider any additional software functionality that would improve the efficiency of their area.

After each session, Coopers & Lybrand would prepare the minutes for review at the next stage. At the second meeting, each business process was broken into a number of abbreviated requirements statements in the form of a question, such as "Does your package support EDI?" or "Does your package calculate forecasts automatically?" Once all the Timberjack requirements were determined, Coopers & Lybrand presented an exhaustive list of requirements that they and other clients had developed over the years. This list was compared to the existing Timberjack list, and occasionally new items from the Coopers & Lybrand list were added to Timberjack's list of requirements.

Sessions were held for accounting, finance, purchasing/planning, customer service, pricing, distribution, as well as information systems. All of the meetings were completed prior to the middle of June, and the preliminary lists of requirements were sent to Marsta for review. Members of the Swedish team were not present during the RFP meetings in Woodstock. Marsta approved the requirements lists quickly and added requirements related to multilanguage capability, electronic funds transfer for payments to suppliers, and Intrastat reporting, which was used in Europe to track the movement of goods.

A benefits review was also started in Woodstock. This benefits analysis was an attempt to tie potential improvements in software functionality to savings in operating costs or improved customer service. In Woodstock, all capital expenditures required a capital appropriation, and approval was based on expected return on investment. In the European operations, capital expenditures were not as formalized. Romanow explained the rationale for the benefits analysis:

We recognized that from a project approval standpoint, the benefits analysis wasn't necessary; we had been asked by the head office to implement the new system. We were doing the benefits analysis to weight the potential functionality improvements. This would help us decide which software had the most potential to improve our operations.

To complete the RFP, Romanow and Nikkinen gathered basic business statistics such as the number of part numbers, number of sales orders processed daily, and the number of customers. Gonzalez and Smedjer outlined the number of users, and the preferred hardware and operating systems. Gonzalez included a description of the North American dealer system, which linked dealers directly to Timberjack's legacy system via modem. By the end of June, most of the work was done, but final approvals were not completed until the middle of August. Summer vacations, as well as Romanow's relocation to Atlanta, slowed the approval process. It included over a thousand requirements, copies of standard reports, and instructions to vendors on the bidding process. Exhibit 5 shows an excerpt of the Timberjack RFP. Each requirement was listed as critical or noncritical, and space was available for vendors to indicate if their software would comply, comply with minor modification, or require a major modification. A minor modification was defined as having an estimated cost of $2,500 or less. Fifty hard copies were produced, and the RFP was also copied in electronic format to allow the vendors to respond as expeditiously as possible.

At this stage in the project, it was clear that there were opposing views on the approach taken on the RFP. The Swedish team was very much in a hurry to select and install a new system. Smedjer explained:

If the RFP were done in Sweden it would have been roughly five pages. Not at all as specific, just mentioning the functions that are required, and not anything of how it should be done or what it should look like. In the end the document was a lot of work for the suppliers to answer all these questions on such a detailed level.

The North American group was committed to following the Coopers & Lybrand methodology. Utting defended the RFP process:

The strength of the RFP process was that we ended up with a clear idea of what we wanted in detail. The second benefit was that we got a lot of people involved. We created some buy-in, not perfect, but some buy-in due to the fact that people were involved. This avoided the problem of people saying afterwards "No one asked me about that," so I think the RFP established a form of consensus.

By the end of the RFP process, the Swedish side was quite frustrated with the time required to put the RFP together. While Utting felt there was value in the RFP, he recognized that in the end it had created a wedge between the groups in Sweden and Atlanta. Sahlqvist commented:

The RFP process was far too long for us, as we were much more squeezed than the other guys. We had a system that didn't work properly, so we were very anxious for a quick process. And Atlanta was quite satisfied with its system, and was more thrust into the project to have a common system. It didn't make compromises to shorten the process; it didn't think about everyone else.

Vendor List

The list of potential vendors was developed by Coopers & Lybrand, with input from Sahlqvist. Several constraints limited the potential suppliers. First of all, the system had to be UNIX based, as specified by an internal standards committee within Timberjack, called the Data Standards and Integration (DSI) group. The DSI group was made up of managers of the various information systems departments from Timberjack facilities around the world, including Gonzalez and Smedjer. The UNIX requirement ruled out many of the excellent packages on the market designed to run on other platforms.

Secondly, the DSI group stated a preference for an Oracle database as a corporate standard, while Progress was also accepted as a second alternative. Gonzalez commented:

The corporate standard of Oracle was intended to give Timberjack the most flexibility in choosing other software and allowing for maximum integration between other Timberjack sites. While Progress may be an easier product to administer, its market share was a fraction of the Oracle market share. A Progress database may make future integration with Woodstock more difficult, and it may be difficult to find qualified programmers.

The package also had to have a presence in both Europe and North America. There were strong distribution packages used in Sweden, but not in North America, and vice versa. By specifying local presence in both markets, the list of vendors was narrowed to large enterprise requirements planning (ERP) packages, which were used predominantly by large manufacturing firms.

Coopers & Lybrand originally recommended strong best of breed packages for the business units. While the parts operations on both sides indicated that they wanted to purchase the same package with strengths predominantly in distribution, compromises were necessary. The vendor list was effectively narrowed down to manufacturing systems with secondary strengths in distribution and service.

In total, 13 companies were sent the RFP and the RFP diskettes. They were sent out on September 15, 1995, three weeks behind the schedule originally drafted in April. Shortly thereafter, a meeting was held in Atlanta to answer any questions that the vendors had regarding the RFP. The vendor meeting was well attended, with most of the invited vendors present. Many of the vendors were interested in the impact of the dealer system on the project, and they also raised several questions relating to the overall budget. The Timberjack team did not feel comfortable disclosing the budget number at that time, as it may have influenced the bidding process.

The vendors were given one month to respond, and in most cases, those with an intent to bid did submit on time. Others required some coaxing to take the time to respond to the RFP. The vendors which did not initially respond included SAP and QAD. QAD's initial failure to bid was a disappointment to the Swedish side and a surprise to Utting, as QAD had recently been awarded the contract to supply the Timberjack retail stores in Finland. Sahlqvist eventually convinced QAD in Sweden to respond to the RFP. In the end, there were a total of six bids; Lawson, Oracle, Computer and Associates, QAD, Scala, and American Software.

When the responses came in, the team assessed the relative strengths and weaknesses of the various packages with respect to functionality. Some packages were very strong in the financials area, while others were strong in the purchasing/planning area. None of the packages was able to do everything that was hoped for. Sahlqvist and Utting agreed that the most critical functionality was in the planning and customer service areas, followed by warehousing, and then finance.

Creating the short list

Once all the bids were received and the results tabulated, the group went through a fairly quick process of eliminating several of the bids. Scala offered a PC-based package; but while its costs were quite low, it lacked in functionality. American Software showed excellent promise in the purchasing/planning area, but very few of its other modules were written for UNIX at that time. As a result, the Scala and American Software packages were dropped from the list of potential vendors.

With the quick elimination of two bidding vendors, the team had the opportunity to concentrate on the remaining four vendors. Initially, separate meetings were held with the vendors in Stockholm and Atlanta. Most of the meetings were held at the vendor site and, in some instances, other members of the parts management team attended. Additional participants in Atlanta were Ron Kipp, manager of Customer Service and Pricing, as well as Gene Torbush, manager of Distribution. Bert Westin, customer service manager in Marsta, added depth to the meetings in Sweden. The meetings typically lasted most of the day, with a portion taken up with presentations by the vendor, as well as Timberjack questions related to the RFP responses, or the amount of local support for the software. According to Romanow, while the RFP was very detailed, it still left a great deal of room for interpretation. Romanow explained: "The vendor visits were used to gain a better understanding of the details behind the vendors' responses. This was accomplished by asking each vendor to demonstrate how its software accomplished various tasks."

While in Atlanta, the group expected strong local support of the packages; in Sweden, a considerable amount of time was spent investigating local support capabilities as well as the expertise of the vendors' implementation consultants. Questions there focused on the number of installations in Sweden and the number of staff who had prior experience installing their software. The Swedish team had experienced problems in the past finding competent vendor support during implementation.

In Sweden, there were a few installations of the QAD package MFG\Pro, and the software was widely used in Europe. Origin, a division of Philips in the Netherlands, was a worldwide partner of QAD for implementation support. Origin had offices in the United States and Sweden. The Swedish team favored QAD in terms of implementation support.

The early Oracle presentations in Europe were not well received. Smedjer remarked:

We attended a sales presentation with a number of other companies in attendance. To us, the presentations indicated that the staff was not fully conversed in the functionality of the software. There were numerous technical problems during the presentations, which left us with an uneasy feeling about the software.

The Computer and Associates (CA) software package that was presented was MANMANX. The Atlanta group attended the sessions with CA, while the Swedish team did not. The main concern of the team was the strategy that CA had taken, which basically involved software acquisition rather than ongoing development. The Timberjack group was concerned about support and the long-term commitment to the package.

At this time, the project team relied heavily on industry articles to better understand the company attributes as well as the software functionality. Timberjack in Woodstock and Helsinki subscribed to the Gartner Group publications, which frequently covered the major players in the ERP software market. The Gartner Group articles, according to Utting, "were like a Consumer Reports for software." At the time, the Gartner group put SAP and Oracle as the leaders in this market according to technical attributes and ability to implement.

Narrowing the list to two

In November 1995, Sahlqvist, Smedjer and Nikkinen came to Atlanta to join the Atlanta team for meetings with the four remaining vendors. The purpose of the visit was to gain more insight into the packages and the vendors that supported them, and to narrow the choices to two vendors.

At the meeting with Lawson, the sales representatives spent a great deal of time focusing on the technical competencies of the software. Lawson emphasized its ability to interface easily with other best of breed packages to fill in for potential gaps in its functionality. Nikkinen stated:

The Lawson package was presented to us in Sweden as well, and we were very impressed with the drill down capabilities of the software; technically it seemed to be very strong. While the package was missing some of the functionality that was important to the Atlanta team, these items were not as critical in Marsta. We saw some real potential in the software.

The Lawson representative was asked to provide a quote for a forecast package to round out its offering, as the other packages offered their own integrated forecasting module.

The QAD representative, Rohit Lal, stressed that the time required to implement its MFG\Pro software was usually six months, which was very important to the team from Sweden. Gonzalez, however, felt that the software did not appear to be based on the same flexibility that he saw in their legacy software and some of the other vendors the team had looked at. Gonzalez remarked: "We're used to a very flexible system?a system which allows you to go in and make changes without affecting source code. MFG\Pro did not appear to have that type of technology." Gonzalez felt that for these reasons the Timberjack dealer system would require significant modification on a MFG\Pro platform. Gonzalez was also concerned that MFG\Pro was written in Progress, and that QAD was having difficulties migrating it to an Oracle database.

The Oracle presentations in Atlanta involved an Oracle financials representative, a distribution representative, and the director of the Implementation Consulting group. Gonzalez remarked:

The package was clearly more complicated than the others, with a large number of flags and switches to allow the user the flexibility to change the software without programming. Inquiry screens were based on their Smart Client technology, which allowed the user to customize screen layouts easily. This technology was based on their growing use of object oriented programming.

During the meeting, the Swedish team started to press the Oracle presenters on their implementations in Europe. Sahlqvist took the list of clients that the Oracle representative in Sweden had provided and recounted the conversations that he had had with the individuals at the Oracle client sites. Sahlqvist explained his concerns:

One install failed; the company's systems representative said it couldn't make the software work, although Oracle claimed it was a successful implementation. Eight more had only installed the financial modules, not the whole manufacturing suite. Not a single distribution type of company in Sweden had chosen the package, and only two or three companies were in the process of installing the full package.

The Oracle presenters were surprised by the hostile questioning, stating that they believed that of all the software companies that Timberjack was considering, Oracle had the most in-depth presence in Europe. They felt strongly that most companies recognized Oracle as a global player and it was one of its strengths. The Oracle representative, Mark Fine, promised Sahlqvist more information on European installations and reiterated that it had a staff of seventy-five in Stockholm to support several hundred client sites.

After the Oracle meeting, the group met briefly to discuss next steps, as the Swedish group had to catch an airplane. It was decided that Smedjer and Nikkinen would continue to investigate QAD and Lawson, while Romanow and Gonzalez would focus on CA and Oracle. The team felt that the list could be narrowed to two within a week. The main sources of input at this time were the license fees for software, the functionality comparisons, implementation consulting costs, and the Gartner Group articles.

After the week was up, a conference call was held to determine the final two packages. The Swedish group showed strong preference for QAD, while the Atlanta group showed a strong preference for Oracle. The decision was made to go forward with Oracle and QAD, with both sides to investigate the possibilities of each implementation.

Timberjack Customizations

By late November 1995, the project team had a good idea of functionality shortfalls in the QAD and Oracle software. The QAD package would require customizations on the dealer system, order type, and automated safety stock, while Oracle needed work on the order processing module to match the current dealer system functionality. The order type functionality allowed Timberjack dealers to specify the level of importance placed on a given sales order, while the safety stock program was used to automatically adjust inventory levels to optimize service to the dealer at the lowest cost. While other functionality shortfalls did exist, the Timberjack team did not consider them as critical for running the business.

Timberjack was interested in obtaining a fixed-price contract on the software, the implementation consulting, as well as major modifications. The Atlanta team had both Oracle and QAD customization representatives on site to see a demonstration of the dealer system. The team from Sweden was committed to implementation in Marsta without any customizations. In Atlanta, detailed documentation was provided on the dealer system, and source code was made available. Both QAD and Oracle were also given the opportunity to dial in to the dealer system via modem to better understand its functionality.

In addition to the dealer system modifications, QAD was given documentation on the automated safety stock program, as well as order type functionality requirements. Oracle was able to meet these needs in the base package. The total costs for the MFG\Pro modifications were roughly $200,000 USD, and for Oracle they were $30,000. While Utting expected to modify the software and believed that the major modifications had been identified, he wondered what other customization costs would be uncovered during installation.

Implementation Consulting

Implementation consultants from Oracle and Origin, who represented QAD, were brought on site to further investigate the user environment and to present to Timberjack their implementation methodology. The methodologies of both firms were fairly similar; however, the Oracle plan involved more consultants over a longer duration. Early estimates from Oracle showed that consulting fees to install the system in both Timberjack sites would exceed $500,000 USD. While Utting believed that the Oracle system was more complicated, it was hard to understand the huge difference in the relative costs of implementation consulting.

Over the next several weeks, the Atlanta team met with Oracle's implementation consulting group to discuss the consulting requirements. Each of the tasks within the implementation process was discussed, in terms of duration and complexity. While the costs of the Origin bid were not disclosed, the number of days to complete tasks was compared. Within each of the tasks, the Oracle group spent roughly two to three times the resource. Utting wondered if the package was really that much more complicated or whether the consulting was used to generate extra revenue.

Given the amount of time that Timberjack had already spent on the RFP, Utting pressured the consulting groups to lower the time required to implement. This was based on the assumption that Timberjack would not have to spend as much time outlining business processes and requirements definitions. Fine worked with the Timberjack group to leverage the consulting arm of Oracle to make concessions on the consulting services. Because of the quality of the Timberjack RFP, he believed there was clearly less need to define processes at the beginning of the implementation. The Oracle consulting bid was eventually reduced by roughly 25%.

While Utting was pleased that the Oracle estimate was reduced, he wondered how much extra work would fall on his people. His organization did not have a large support staff, and most of his people were on the phone with customers and suppliers all day. It would be hard to imagine spending months in conference rooms trying to understand the new package. Utting commented, "Right or wrong, the team felt that in terms of implementation complexity, QAD would take roughly six to seven months to install, and Oracle would be at least a one year program."

Utting relayed the implementation timeframes to the Oracle implementation consultant, who responded: "All packages are tough to install, there is no way around it. Is Oracle hard to install? Absolutely. But in no way is it twice as hard as QAD. You still have to complete all the data conversion, and truly understand how the software works through key processes."

Utting then asked "Why does it takes twice as many hours for your consultants to do roughly the same thing." The Oracle representative responded:

In everything you get what you pay for. If you want someone to come in and spend very little time understanding your business, it's up to you. Essentially they will tell you which settings to use, and away they go. We call that "brains on a stick" consulting, and we just don't run our business that way.

Sahlqvist and the Swedish team were even more concerned about an Oracle installation. In an effort to reduce these concerns, Mark Fine flew to Sweden to meet with the Timberjack team during one of the Oracle sessions. By this time, the Oracle Sweden group had enough time to gain more positive information about the installations to date in Scandinavia. During the meetings in Sweden, Fine extended guarantees on behalf of the Oracle consulting group in the United States; he committed that if suitable consultants were not available in Sweden, consultants from the United States would be arranged at no extra cost for the duration of the install. Due to the urgency of the install in Sweden, he also committed that full implementation would be completed within a one-year timeframe, with noncompliance penalties to apply to Oracle. Fine was able to offer the guarantees based on the fact that other Timberjack and Rauma sites, particularly in Woodstock, were going to be selecting new software in the near future.

Site Visits

The guarantees offered by Fine to Sahlqvist made Oracle a viable option for the group in Sweden. Romanow stated: "Until that point, we felt that we were pushing a rope to get them to seriously consider Oracle as a solution." As the year was drawing to a close, the Swedish team was pushing for a final decision. Sahlqvist indicated to Utting that the week of December 12 was the last chance before Christmas to travel. Sahlqvist also indicated that client site visits might be appropriate if they could be arranged. Several key people from QAD, and particularly Oracle, were already booked for that week, but arrangements were made for meetings. Initially, site visits had not been arranged, as neither vendor had adequate time to prepare.

When the team arrived at QAD, a company representative indicated that it was possible to see Matrix, the salon shampoo manufacturer located in Cleveland. Sahlqvist was most interested in visiting a distribution company site that was running MFG\PRO, but none were available. The team agreed that a site visit to Matrix would be a good use of their time, however, and booked flights to Cleveland. Romanow called Fine at Oracle and indicated that the team was going to visit a QAD site. Later in the day, Fine called Romanow with the news that True Temper, a manufacturer of golf club shafts located in Memphis, was willing to do the Oracle site visit.

Utting commented on the two site visits:

The day at Matrix was well organized, and a number of individuals from Matrix were available for comment. It was interesting to note that Matrix was running on version six of MFG\PRO, which was a much earlier release. Matrix had also modified extensively, especially in order entry and in the pricing and promotions areas. We wondered if there was a correlation to the modifications and the fact that Matrix was still running on version six, while Timberjack was looking at version 8.4G.

The day with True Temper was also very informative, and their project team was very gracious with their time. True Temper had managed their install in steps, rolling out the financials first, then purchasing, and so on. A True Temper person from each functional area spent time with the Timberjack team to better understand the implementation process. We could not help but notice the number of people that had worked on the install. Financials had a team of six people, with several of them spending 60% of their day on the project. The project had taken more than a year to complete, although there were breaks along the way.

On the way home from True Temper, the group had some time at the Memphis airport and stopped for a drink. Sahlqvist, Romanow, and Gonzalez discussed the site visits, and felt that the week had given them new information. Gonzalez made a comment on the fact that Matrix was still running on an earlier version of MFG\Pro and speculated that the modules were hard to modify without affecting the core system. Romanow added, "I believe our decision is one of short-term pain through an Oracle install versus the long-term gain of more flexible software. Are we willing to pay a premium for that flexibility?" Sahlqvist replied:

You guys don't know what you are talking about, because you are so convinced that Oracle is the way to go. Why should we go with such a huge system, and all the headaches, when software is changing so fast that we may install something entirely different in three years. We are more like Joe's garage than General Motors, so why do we need Oracle? We would not support paying one dollar more for Oracle.

Final Selection

The next day, Utting and Sahlqvist decided that they would exclude themselves from the final selection meeting. If the rest of the group could not reach a consensus, then Utting and Sahlqvist would try to make a decision. Smedjer, Nikkinen, Gonzalez, Romanow, and Bahar met in the conference room to try to make a recommendation to Utting.

Utting and Sahlqvist worked together in Utting's office to try and reconcile the latest financial costs of the software, hardware, consulting and modifications. The project team had recently received new quotes from QAD, Origin, and Oracle, and Utting wanted to make sure that they were all comparable. Currency translation, and notes in Swedish regarding the changes made the comparison difficult.

Meanwhile, the rest of the project team met with Bahar, who acted as the facilitator for the meeting, and who remarked that he sincerely understood that each side was coming from a different place, and the criteria were different. How could one average the Swedish view?don't complicate things too much, we want a system that will work in six months time, and by the way, we want it cheap?with the other view?we are not in a big hurry; we want a good system, and if we can bring it in at more or less a similar cost, why not go with the better technology, the better system?

Bahar worked with the group to try and decide which factors would be most important going forward, attach a weighting to each, and then gain consensus on an overall score. Exhibit 8 shows the factors and corresponding weights that the project team used to evaluate the two packages.

Utting drifted briefly from his discussion with Sahlqvist and wondered how the team was doing in the other room. He hoped that they were having a better time of it than he and Sahlqvist were having trying to gain consensus on the latest cost projections from Oracle and QAD. Given the tension in the air at the Memphis airport, both Utting and Sahlqvist believed that their involvement might stifle free discussion amongst the team. Besides, it was the project team who would have to implement the final choice. At that point, Bahar stepped into Utting's office and announced "The group has reached a consensus."

The Qeustion: Talk about the process of preparing RFP in Timberjack Company.

Part 2.

7. Transportation costs and comparative advantage The following graph shows a fictional world economy that consists of only two countries, Golikia and Bulvia. Both countries produce boats under increasing-cost conditions. Note that the left-hand part of the diagram is a mirror image of a standard supply-demand diagram, and therefore the supply and demand curves slope in directions opposite their usual directions. Golikia Bulvia 30 Se 27 24 21 -18 PRICE OF BOATS (Thousands of dollars) 15 12 De DG 10 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10 BOATS In the absence of trade (that is, autarky), the equilibrium price in Golikia is $12,000 , and the equilibrium price in Bulvia is $24,000 . (Hint: Enter all monetary values in full. For example, $7,000 rather than $7.) In the absence of trade, which of the following statements is correct? Golikia has the comparative advantage in production of boats. Golikia and Bulvia are equally good at producing boats. Bulvia has the comparative advantage in production of boats. Now suppose both countries open up to international trade with each other. For each country, use the previous graph to compute the equalizing price, consumption and production at that price, and the quantity of exports and imports when there are no transportation costs. Enter these amounts into the first two rows of the following table. Equalizing Price Consumption at Equalizing Price Production at Equalizing Price Exports Imports (Dollars) (Boats) (Boats) (Boats) (Boats) Without Transportation Costs Golikia Bulvia With Transportation Costs Golikia Bulvia Now suppose that the per-unit cost of transporting a boat between Golikia and Bulvia is $6,000. For each country, use the previous graph to compute the equalizing price, consumption and production at that price, and the quantity of exports and imports when transportation costs equal $6,000. Enter these amounts into the last two rows of the previous table. Compare free trade in the absence of transportation costs with the case when transportation costs are included. Which of the following statements about how the trade between Golikia and Bulvia differs in the presence of transportation costs relative to trade when then are no transportation costs are correct? Check all that apply. Golikia will produce less, consume more, and export less. O Bulvia will produce less, consume more, and export less. O) Golikia will produce more, consume less, and import less. O Bulvia will produce more, consume less, and import less