Data Resource Management at Semantic Software.com: Chaos to Efficiency
by Scott Tavelli & Rajeev Kaula

Semantic Software.com was a startup software development company. It focused on creating practical solutions to improve operational and strategic knowledge of other corporations. The dynamic nature of the Internet environment and its expectations however prevented Semantic Software.com from organizing its own corporate data efficiently. This case details the data management chaos that existed at Semantic Software.com and the steps the company undertook to improve its management of data and bring efficiency toward its operations. (The company name has been changed for legal purposes.)

Introduction

Across industries today, there is a much greater awareness than ever before for cost control and cost reduction by means of improved information management. As a result of the evolution of market competitiveness, tremendous pressures have been set upon information technology businesses to provide additional value at lower costs.

As a startup software development company, Semantic Software.com addressed this need by designing and developing state-of-the-art enterprise information systems. It stood out from its competitors by offering high quality Internet-based solutions at competitive prices. Its goal was to become a leading software solution provider across the European market. It intended to accomplish this goal by offering well-designed products, which met client needs more effectively than any other competitor, and by applying the latest available technology in software engineering.

Semantic Software.com wanted all prospective clients to know that it was a firm that was focused exclusively on developing practical solutions. Such solutions provided operational and strategic knowledge where it was needed, when it was needed, in the form it was needed, and that it served this market better than any other competitor. It wanted customers to regard it as one of the industry's innovators, and a creative company working hard to find new ways to make information ubiquitous. When an organization wanted novel leading edge solutions, Semantic Software.com intended to be the first name that came to mind.

Data Management Chaos

The dynamic nature of the Internet (dot com) environment and its market expectations however prevented Semantic Software.com from organizing its own corporate data efficiently. For example, billing information was mostly paper-based. Work reports were for the most part handwritten as personal notes. Contracts were scattered throughout the IT network. Salaries and payments were centralized in a proprietary accounting software. Project requirements and specifications were dispersed in email boxes and on project managers' desks. Time was tracked using a simple intranet-based solution, but this data was never used. Strategic data planning was non-existent. Strategy was an unknown concept from a company-wide data administration perspective. Corporate knowledge was individualistic and volatile.

Such data management chaos led to stressful situations. As bills were prepared for the company's first customers, secretaries had to rely on project managers' memory, or on personal notes of software engineers, or on internal emails. Information was many times incomplete and incorrect. Multiple versions of customer contracts existed on the network, which led to confusion. Several signed original contracts got lost in the company's 2000 square foot facilities due to such chaos. Project requirements and specifications were rarely documented, let alone signed-off by the customer. Software engineers relied on project managers to provide technical assistance, communicate user expectations, and cope with critical situations. Further, software development activities were collected daily in a relational database for which no one took ownership.

Data administration in such an environment was sporadic and instinctive. Valuable data was inadequately organized as a standard but not communicated properly. In addition, top management requested exact on-the-spot predictions for product completion and project delivery. As a result, everyone relied on project managers, corporate memory, and sort of "red eyed" heroism. Nevertheless, amid this apparent chaos, Semantic Software.com was creating novel and unique products.

Data Management Efficiency

A considerable change started taking place when the company's Chief Technology Officer (CTO) came back from a university on-campus session. A concept was devised, a plan was elaborated, and a data management revolution was put in motion.

Strategic goals were defined to perfect the quality of services and products; to improve accuracy of proposals and plans; and, to consolidate corporate knowledge of the software development process in a central data store. Accordingly, multiple actions were taken to fully document all aspects of projects, both internal and customer specific, and standardize daily activity data into a new knowledge warehouse.

A data model was developed as the foundation of the company's new knowledge warehouse. Detailed reports and numerous summaries now could be generated at will using an intranet-based application. While the data and the tools appeared to be quite simple, they provided a wealth of information and value to project managers and executives.

The improvements came not only from new goals and reengineered processes, but from the leadership, ownership, conviction and passion the CTO demonstrated throughout the process. The CTO took the necessary time to generate knowledge out of the information that was gathered on a daily basis. He knew he could improve his understanding of software development projects, learn to recognize problems early, identify their origins, grasp the underlying mechanics, and improve existing processes and organizational issues.

Data Management Initiative

One aspect of consolidating strategic knowledge focused on structured data that measured software development costs. This was necessary to improve project estimation and project planning capabilities. The core element of this process was an activity, which is a single unit of work qualified by time, purpose, origin and location. An activity is performed by an employee during a limited period of time. It is associated with a cost center, or project, and is fully commented and characterized by a generic type. This new data structure allowed project stakeholders to collect work activities based on corporate standards, and conduct comparative analysis across projects and time.

After analyzing management and operational needs, three groups of cost centers were proposed, each having its own activity types:

  1. Product development centers consisting of component development, application development, configuration tools development, testing/quality assurance, and project management. Such centers were considered as investments in software assets.
  2. Customer-specific project centers consisting of testing/quality assurance, on-site installation, training, and project management. These cost centers were billed monthly to customers based on detailed accurate data.
  3. Administration centers consisting of secretarial work, accounting, human resources, training, management, sales, etc. These centers were considered as operational overhead.

The setup of project cost centers helped improve understanding as well as management in the following areas:

  • Human resources.
  • Software development process.
  • Finance and accounting

Another aspect of the strategic initiative focused on unstructured data associated with product development as well as customer-specific projects. The objective of this initiative was to facilitate and promote communication within Semantic Software.com so as to improve the quality of deliverables and increase customer satisfaction. It was also aimed to keep written records of project successes and failures.

This required projects to be documented, standards to be defined, requirements centralized, specifications properly signed off, and streamline information flow at an operational level. Cross-departmental information needs had to be understood and improved.

The solution was aligned with the new corporate data structure and the directory-based intranet application. It made employees responsible for maintaining a standardized home page, describing their past experience (companies, activities, technologies), their current assignments (with respect to customers, projects, and technologies), their skills, and their interests. Also project managers were required to describe project goals, risks, requirements, deliverables, and milestones according to new standards.

From a process standpoint, mockups and prototypes were now systematically developed and signed-off by the customer, the project manager, the product specialist, the applications architect, and the platform architect. As soon as features and functionality of a product were approved, technical plans were designed, work breakdown structures were established, detailed functional and technical specifications were written, workloads were estimated, resources were assigned, and deadlines and deliverables were formally scheduled. The entire process was supported by up-to-date documentation in the company's knowledge warehouse. All documents were consolidated, and all versions were available.

Conclusion

The premier goal of Semantic Software.com was to provide great software solutions at competitive prices. While development dynamics were essential to support the rapid growth it was going through, data management pandemonium had to be gradually replaced by structure, standards and processes. The corporate knowledge warehouse, envisioned by the CTO and implemented by his team helped to improve the reliability of software development estimations. Structured efficiency replaced chaos, at last.


Scott Tavelli is a member of the MSCIS program at Southwest Missouri State University. He works as an Software Development Consultant at a large company in Europe.

Rajeev Kaula is a professor in computer information systems at Southwest Missouri State University. Dr. Kaula does research in the areas of open information systems, computer integrated manufacturing, data management and decision support systems. He is the contact for additional information by e-Mail: rak549f@smsu.edu .

Many more articles on Leading Change, Performance Improvement and eBusiness in The CEO Refresher Archives

   


Copyright 2002 by Scott Tavelli & Rajeev Kaula. All rights reserved.

Current Issue - Archives - CEO Links - News - Conferences - Recommended Reading