|
|
 
|
This page contains fairly detailed descriptions of some of the larger consulting projects I've overseen.
Commission Tracking and AdministrationClient ProfileThe client company is a Fortune 50 telecommunications company with a long history in the industry. The project was done for a division of the corporation which has responsibility for the design, assembly, sales, leasing and maintenance of computer systems and directly addressed the needs of the staff administering the calculation, reconciliation, tracking and payment of commissions to the commissioned sales staff.
Situation AnalysisThe corporation had been in business for many years, long pre-dating its involvement in the sale of computers. The system that was currently being used to administer the compensation needs of the division was originally built to track the commissions of communication services sales people and was, in many respects, ill-suited to the needs of a merchandise-sales division. This was most apparent in the lack of agreement about the exact nature of customer accounts. The way the system tracked customer subsidiaries and locations was so at odds with the way the division wished to incent their sales staff that it was virtually impossible to ensure that any given salesperson was credited with the sales of his or her customers. Moreover, the amount of administrative overhead in the system was staggering, with the result that commissions were routinely paid several months after they were due. The final weak point was that, as the system was still owned and maintained by the communications portion of the corporation, all system enhancement requests from the computer division were treated as exception requests which were implemented only with great delay and even greater expense.
Project SummaryI was half of a two-person advance team which began the project with detailed interviews with compensation policy designers, current-policy administrators and representatives from the actual sales force. After extensive analysis of the weak points in the current operations and projections of potential trouble from upcoming policy changes, I led a much larger team which began building a system which was small enough to be economical for the computer division to operate on its own, friendly enough to allow the administrators to perform the jobs quickly and accurately, flexible enough to allow the policy designers to create the plans they wanted to (rather than the plans their system could support) and compact enough to be maintained by a small in-house systems development/administrative team. Built as an Oracle application running under Unix System V, the system was prototyped on a small, dedicated minicomputer with user access via regular dialup lines into a bank of modems. When development and testing were completed, the application was ported to an Amdahl mainframe which was being used to run a number of the division's applications. Access to the Amdahl was via a high- speed Wide Area Network. The system, as finally put into production, was notable in many respects:
Communication Service ProvisioningClient ProfileThe client is a major distributor of network communications services. These services are marketed directly to major corporate customers who use the services directly as well as to long-distance services resellers who offer the services to small and mid-sized business customers. Situation AnalysisFederal Communications Commission regulations are very stringent and occasionally quite complicated on the issue of switching telephone service from one long-distance carrier to another. As a result, this client performed extensive research on most of the orders that were received from resellers. This research was mostly comprised of numerous checks against existing-service databases on a number of different computer systems (both internal and external) to verify that the reseller's order accurately reflected the wishes of the reseller's end-user. This research was done by manually logging on to a number of different computer systems and making numerous handwritten notes on a physical copy of the order. Very complex cross-referencing was often necessitated by discrepancies across the reference databases but the extremely time-consuming nature of resolving these discrepancies meant that that research could rarely be done. As a result, the client was rejecting a very large number of orders, was receiving frequent complaints of errors from end-users, resellers and the FCC and was as much as four months behind on tens of thousands of orders. Moreover, the client was attempting to introduce newer, simplified services, directed solely at the reseller market which would attempt to process all orders within one to two weeks. Although the client was willing to make a fairly substantial one-time investment in technology, a further goal was to keep head-count from growing (or to allow it to reduce through attrition). Project SummaryI was brought in as project manager of a team of consultants which was to work both with product managers and with services-provisioning managers and staff to assess ways in which state-of-the-art technology could be coupled with a total redesign of the client's business processes to produce order of magnitude increases in both the quality and quantity of orders processed. The project was divided into a number of large phases. In the first phase, the new, reseller-only product was leveraged. In addition to interviews with representatives from product management, operations management, front-line terminal operators and supervisors and customer-resellers, we conducted time-and-motion studies, cost-benefit analyses, exception-processing analyses and an overall technology survey. With this information, the team built a prototype system for the processing, tracking and provisioning of orders. This prototype replaced all users' terminals with personal computers accessing a database on a local area network. Also connected to this network were several communications servers which ran dozens of scripted login sessions as background processes performing all necessary research and provisioning against external mainframe systems. These servers were so much faster than the manual research processes that it was now possible to research every single order, including the complex but critical cross-referencing functions. After a refinement process in which the prototype was in active use by a special test-group, the next phase of the project was the roll-out of this environment to the entire product group. During the course of the prototype and subsequent roll-out, a number of additional modifications were made to the project, including many more quality checkpoint criteria, several dozen management reports, extensively detailed order status inquiry facilities and the ability for the resellers to submit their orders and to receive routine status reports via electronic mail. By this time, the system was so highly automated (electronically-submitted orders literally could proceed through the entire process with no human intervention), that the average time to research and provision an order dropped from approximately five hours to an average of four to six minutes and front-line sales representatives were free to spend most of their time taking care of customer sales and inquiries instead of digging through a backlog of orders. The final phase of the project consisted of rolling a very similar system out to the remainder of the organization. Since the organization as a whole dealt with a number of very different network products, a number of additional analysis sessions were required at this point but, due to the knowledge and experience that had been gained in the previous phases, this process went extremely rapidly. By the time the entire project was completed, the client's workplace had changed dramatically. When my team first arrived on the site, orders were received on paper via fax machines and regular mail and researched by terminal operators logging on to numerous external systems. At times, up to 40% of all orders were sent back to customers for additional information and backlogged orders could sit on people's desks for months at a time. Management was frequently forced to bring in veritable armies of temporary workers on a round-the-clock basis to keep the backlogs from getting even worse. By the end of the project, the client had moved all computing functions on to a large, multiple-server, multiple-platform Local Area Network with 350 Windows-based workstations, both unix and DOS-based servers and fully automated connections to several different mainframe computers supporting a number of different communications protocols. Operator interaction with the order pipeline was now mostly by choice rather than necessity, virtually all orders could be serviced within a matter of days and temporary help was a thing of the past. In addition, the presence of such a distributed processing environment also provided a number of indirect productivity enhancements. The most significant of these was that, because the client's third-party electronic mail service was supplemented with an in-house communications facility, internal communications increased greatly...managers and staff became much more aware of overall status, problem issues were communicated and resolved more quickly and the expense of upper management's internal e-mail dropped greatly since these no longer had to be routed out the third- party mail service, only to be sent back to the same site.
Facilitation of Pension ReportingClient ProfileThe client is a Fortune 50 manufacturer in the automotive, machinery and aerospace industry. This project was done specifically for the computing services organization of the company's Retirement Benefits Administration. The company has nearly 70,000 employees participating in a number of different pension plan options. Situation AnalysisThe client was in the process of shifting its record-keeping from one pension-tracking company to a much larger banking firm. Although this move was cost-effective overall, the new bank provided only a certain number of pre-defined monthly reports. The client had a number of other reports which it had been receiving from the previous record-keeper and which were critical to its own internal administrative requirements. Although the bank was willing to discuss generation of these reports, both the time and money they were demanding to do so was unacceptable to the client. Project SummaryI was brought in to work with representatives from the internal user-groups, from the internal data processing staff and with the senior manager at the bank who was supervising the transition. It was quickly determined that the bank was willing to transmit a bulk dump of their database to the client on a monthly basis. The client was amenable to this and we began looking at ways to store this data locally in a format that would be readily accessible to the internal user groups, most of whom did not have a great deal of experience with database applications. In addition to this lack of comfort- level, a number of other constraints were added. There were plans to upgrade the division's LAN server at some point soon but, other than this, there was essentially no budget either for extensive development efforts or for new hardware purchases. The new reports had to be accessible within a very short time frame. Finally, the users were requesting that any new reports they might require in the future be readily obtainable from this system. The one positive budgetary issue was that, because of the client's size, they had bulk licensing agreements with a number of major software vendors and consequently, a great deal of software was accessible to us at little or no cost. Because the budget and time constraints precluded the development of a customized system, it was decided that we would research various off-the-shelf solutions and determine the degree to which these could meet the users needs. It was also decided that, given various elements of the corporations standard operating environment, we would use an OS/2-based server on the corporate LAN as an Oracle database server. The application package would have to run as a Microsoft Windows application, talking to the database server via Oracle SQL*Net. The remainder of the project consisted of a number of subtasks. First, because the monthly feed was enormous (and well beyond the storage capacities of the database server we were using), it had to be reduced. The extreme size of the feed was due in part to there being a large number of data elements which did not interest the internal administrators and in part to the fact that it contained a great deal of individual transaction information which was more detailed than the users required. It was therefore decided to transmit the data to one of the corporation's mainframes where I designed and supervised the coding of a filter-program which stripped out unwanted detail and rolled transaction data up to a higher-degree of summary. It was the output of this feed that we actually received on our Oracle server. In conjunction with this step, a database was designed on the Oracle server to hold this data. The structure of the database was well-documented, names were carefully chosen to be clear and meaningful to end-users and details of the database were carefully documented both online and in printed documentation so that users could be quickly educated on where to find the information they required via ad-hoc queries. Overlapping this work was an in-depth survey of inexpensive reporting tools that met the users requirements (particularly the ease-of-use criterion). A number of these packages were obtained on approval and hands-on testing proceeded with these. Although all of these packages involved some degree of compromise, one product determined to be acceptable to the users once I subjected it to some custom configuration for this user community. The final subtasks were the installation and custom configuration of this reporting tool and a training phase in which the users were instructed in the use of this application. From this point on, the users not only could pull pre-defined reports on an as-need basis, they also had access, directly from their desktop, to a detailed "what-if" processor that allowed them to investigate a number of criteria for current plan participation. Moreover, because the product selected can be linked to other Windows applications via Dynamic Data Exchange, a great deal of related activities, from the production of mailing labels to complicated spreadsheet projections, are readily accessible from this inexpensive, user-friendly system. |
|
[Home] [Travels] [About Me] [Guestbook] [Map] [Search]
|
