the User's Definition of Success
 |
 |
The first step is gathering cost of downtime data to decide how much infrastructure reliability, availability and maintainability is enough. Total Quality Management (TQM) is a proven approach for assembling information and formulating an appropriate plan of action. With TQM, the focus shifts from what is convenient and customary for the supplier to what is needed and valuable to the customer. In our business, the customer is the end user of information who is very clearly saying he is extremely unhappy with costs, availability, and the timeliness of new applications. End users are moving to PCs and workstations believing that they will have more control at less cost. Outsourcing is growing tremendously because senior management is increasingly upset with a perceived lack of Information Systems results.
Several years ago, the information users in one very large company confronted IS with their growing disgust and desire to do something dramatically different. IS responded with the standard answer: their mainframe availability was the highest in their industry. The users then accused the IS people of dreaming if they thought the users availability was "world class." Out of this confrontation came a commitment to apply TQM concepts, and the results have been remarkable.
Instead of the usual posturing and finger pointing, the company initiated a massive data collection project which quickly revealed how truly awful user availability really was. In fact, glasshouse and the communications network availability could be perfect and end users would still believe that availability was unacceptable. The manually assembled data made it clearly apparent that the end user was his own worst enemy! Less than 10% of lost workstation hours were being caused by problems with the host or with communications. Yet, lacking information to the contrary, the end user blamed all of his own problems on his glasshouse information supplier.
The problem here, and in other companies, is one of proper availability measurement. IS typically measures availability at the mainframe. Communications is concerned with the network. The end user is concerned with terminal or workstation issues. What the ultimate customer experiences is application availability on his screen. End-to-End availability measurement identifies problems that would otherwise be overlooked and allows scarce resources to be focused on the most fruitful areas.
Another interesting conclusion derived from the data underscored the growing importance of site infrastructure reliability. A single site infrastructure failure at the glasshouse level was equal to the company's total lost workstation time (from all other causes) for an entire year. And, unlike information outages limited to a single workstation or clusters, a glasshouse infrastructure failure affected the entire company simultaneously, preventing the orderly conduct of business. This impact on revenue, productivity, and customer satisfaction can only escalate as manual systems become computerized and computer application become ever more tightly integrated.
Many years ago someone said true wisdom is in not necessarily knowing the answer, but in formulating the right question. End-to-End is the right question, and now is the time to create the necessary measurement tools and collect hard data. The results have already proven to be shocking and profound, and could save your company millions of dollars!
