The observation that computers create work is nothing new. Admittedly, it has become less common nowadays when computers and communications have become “e” and are properly credited with creating opportunity. They have always created opportunity but only ever been the tools, and using them to exploit opportunity has always created work. As the nature of the opportunities and work has changed over time, so too have the demands on the corporate information that is central to the projects.

Many early computer systems focused on data recording, processing and reporting within fairly limited parameters. Output created huge volumes of paper, reflecting businesses' thirst for data analysis, in the hope of improving their decision making and hence their performance.

Many early systems mimicked the manual systems that they were replacing. Efficiencies were gained (hopefully) by increasing the speed and scope of reporting, and automating previously laborious manual tasks. Those efficiencies in due course reduced costs, which flowed through to show improved bottom line profitability.

At the time, the financial sector relied heavily on clerical staff, and eagerly grasped the opportunity to automate much repetitive, labour intensive and costly work. As tools for the painstaking and exhaustive analysis of the data had never existed, data quality was only considered in so far as it was necessary to support contemporaneous reporting. Poorly formatted fields, and a high proportion of free text, characterised systems of the day.

The term reporting is often used to cover all output from systems, but there is a difference between reporting and analysis.

  • Reporting can be defined as carefully formatted data which is actually reported, perhaps to a regulator, board, or owner. Reporting is regular both in production frequency and style. Its repetitive familiarity allows readers to assimilate rapidly any changes since the previous version. Its intention is to show how the business is operating within some kind of predetermined framework.

  • Analysis is the examination of data and the drawing of conclusions from it. It is a more flexible tool than reporting, allowing the business to be dissected in minute detail, and showing patterns, leading to more questions being asked of the data.

    The right analysis project enables new patterns to be seen and allows businesses to make better informed decisions. However, where reporting has led to the creation of strong threads of information through the database, analysis requires a solid matrix of reliable information, descriptive of all aspects of the business.

    Successful analysis necessitates emphasis on data control and quality, factors that are often forgotten in the quest for speed and functionality. Achieving and maintaining data quality is a huge challenge, particularly in a market seeing so many mergers and acquisitions. Separate halves of newly married enterprises may have evolved completely separate analysis codes and techniques. With the development of the internet and HTML formats, paper volume is perhaps reducing, but the output isn't. Web pages hosted on a single server can be viewed by millions. This massive distribution is another driver for better data quality.

    The food retailing industry provides a good example of analysis, with the discovery that not only the stock on offer, but its physical layout within the store influences demand. This has led to sophisticated product placement techniques for performance gain. More recently, with loyalty schemes enabling accurate measurement of each customer's spending pattern, retailers are rumoured to put different products alongside each other at different times of the day and on different days of the week, fine tuning their business for the different groups of their customers that visit the store at different times. Improved loyalty, repeat orders and marginal growth are critical in a highly competitive and price sensitive market.

    The element connecting food retailing and insurance is business intelligence systems such as on-line analytical processing (OLAP): the use of computers to determine patterns in data which would otherwise lie unrecognised, enabling organisations to take better decisions to improve their performance.

    In commercial insurance the return on investment provided by improved decision making is potentially vast. Spotting a trend in data that allows business to be declined, re-rated or reinsured can easily recover the cost of implementing the sophisticated software necessary to provide the answers. As the information provided improves, it can be used to better not only bottom line performance but also top line growth - enhancing the relationship with each customer by defining customised products and services.

    Freedom of information provided by the internet is leading to many previously complex financial services becoming commodities, increasing the competition and reducing the margins available. Poor decisions are going to be more damaging in businesses which have less room to manoeuvre.

  • Marcus Broome is chairman of Room Underwriting Systems.