At the end of 1996, Central CMI came up with the idea of developing a database application for the data sources the department distributed. At the time, the customers of Central CMI received most data via hard copy and some data via e-mail. The department recognized that both methods had several shortcomings. Delivering in hard copy implied delays because one would have to wait until the full report, usually referred to as a "book," was printed. Producing and printing these "books" was a time-consuming and costly process because of their size and number. Further delays were introduced by the delivery method of hard copy, particularly when destinations such as Sao Paulo or Singapore were involved. It was also very difficult, if not impossible, to make the necessary adaptations once the "books" were printed. E-mail often caused attachments to arrive in mutilated form because of the usually complex graphics included. Also, the department often ran into problems because of the size of the attachment. E-mail also involves risks of security. Reasons such as these induced the department to develop a system to handle these problems.
Early in 1998 the Comate system that resulted from this idea was put into operation. Comate was built on IBM's Lotus Notes functionality and was offered to users on TopTech's intranet via the Domino system. Comate consisted of the following five applications:
Market Data: offers processed data and analyses in the form of presentations concerning markets, market shares of competitors, distribution, price movements, market predictions, and socio-economical and technological trends;
Research Projects: contains the results of research projects completed by internal and external investigators;
Project Informer: contains information about planned, current and completed research projects run by Central CMI;
Let's Japan: provides a monitor of technological developments in Japan and follows the main competitors and their investments in consumer electronics, research and product development in that country;
CMI Contacts: contains organizational charts of the TCE organization, and a knowledge map of the connections of Central CMI inside and outside the TopTech organization.
Access to Comate has to be authorized by Central CMI. The home page of the system, which is accessible to all TopTech employees, offers a registration form to request permission to use the system. At the time the project team led by Johan van Breeveldt started its work in the spring of 1999, some 250 people all over the world were granted this permission. The first two applications mentioned, Market Data and Research Projects, were the most popular in Comate. To illustrate the functionality of Comate some examples from Market Data will be presented. The application can be regarded as a collection of search tools on top of a large set of documents, with some additional functionality loosely linked to search actions. Search actions for documents or their authors usually start by selecting one of the categories "Product," "Region," "Contact," and "Publications," with an additional entry "New Publications." Clicking, for instance, the option to search for documents related to specific products offers a taxonomy of products at several hierarchical layers, based on the standard classification of TopTech with which all employees – in varying degrees of detail – are familiar. New layers will appear when users zoom in on a specific class of products (or if they choose at any point in the hierarchy to "expand all"). Documents are typically connected to the base categories of the taxonomy. Apart from the hierarchical menu system organized around products, regions, etc., some additional search functions are offered. Most of the additional functionality in Comate is introduced for the purpose of stimulating communication among Comate users. For all documents, additional meta-information is stored, including the names of the authors. A typical example is the response button that is connected to every document. Clicking this button will open a new window allowing the user to send remarks or questions to the authors in question. When the user files his or her comments, an e-mail message is sent to the authors to notify them. To read these comments, they have to log in to Comate and navigate to the document to which the comments apply. These comments and reactions are accessible to all users of the system, allowing them to contribute to the discussion.
With regard to this case study, it is important to note that the Comate system was developed on a top-down basis. Central CMI, and particularly Hans Broekmans who considered the project "his offspring," pulled all the strings in the project. Its customers, the intended intermediate and end-users of CMI sources, were hardly involved in its development and implementation. Also, when the system needed to be expanded or adapted, no customers were involved. No systematic consultations with people outside Central CMI's development staff ever occurred. This may appear as more surprising than it actually is; the system was conceived primarily as an extension of the work of Central CMI, and not as an aid to make life easier for the customers of Central CMI. It was intended to help streamline existing procedures and speed up current routines in the work of that department. The rationale was that if requests for information could be processed faster and at less cost through Comate, this would be to the benefit of all parties involved.