A Computer Called LEO: Lyons Teashops and the World's First Office Computer

Citation

Whitley, E.A. (2004), "A Computer Called LEO: Lyons Teashops and the World's First Office Computer", Information Technology & People, Vol. 17 No. 1, pp. 102-104. https://doi.org/10.1108/09593840410522206

The British Company J Lyons and Co, who ran a series of tearooms and bakeries in the UK, developed the first business application of computer technology over 50 years ago. This delightful book by Georgina Ferry provides a fascinating insight into the way the company developed and used LEO in its business operations and even sold computing services to other companies. It also tells the story of how LEO lost its edge over its competitors and eventually came to an end. This is a fascinating story of the history of our field and of the beliefs of the early systems developers that still resonate today. It should be an essential read for anyone who is interested in information systems.

Founded in 1887, the Lyons company grew steadily and by the 1930s had 250 high street teashops and a workforce of over 30,000. Lyons had a formidable back office operation to support them: hundreds of clerks checked sales and takings, ordered supplies and managed inventories. Lyons had a tradition of “scientific management” and even had a specialised “Department of systems research” whose role was to improve processes in the back office.

Following the Second World War, the computer began to enter the popular imagination and Lyons Comptroller John Simmons and his staff began to consider how computers might support many of the activities currently undertaken by the clerks. Simmons encouraged some of his staff to visit computer installations in the USA, where the early systems were being used for complex mathematical calculations. They also learned about a system being developed in the UK, at Cambridge.

The Cambridge team, however, were not developing a computer for commercial use and Lyons soon realised that the requirements of business computing were very different from those of scientific computing (for example, many simple calculations versus a few, very complex calculations; extensive versus limited input and output) and so, in 1947, Lyons signed an agreement whereby Cambridge would help them build their own computer, a computer that would be modified to make it suitable for business type operations.

Over the next few years, engineers from Lyons and Cambridge worked closely on LEO, addressing the many technical problems that business operations required, ranging from increasing the storage capabilities of the computer, to handling large numbers of records and outputting detailed results on line printers. They were also concerned about the reliability of the system and so implemented a modular design whereby racks of valves could easily be repeated. They also designed languages with which to program the system. In November 1951 LEO took over the “Bakery valuations” process. By 1953 LEO was also doing payroll which involved calculating employee pay, taking into account tax, national insurance, company loan repayments, holiday and sick pay. It even calculated the quantity of coins needed to fill the pay packets.

With the success of the first payroll system, LEO was soon programmed to perform ever more tasks for Lyons (for example, monitoring sales and even making recommendations for individual teashops). Leo also began providing services for other organisations and, for a while there was talk of the British revolution in business computing leading the world. But, as Ferry carefully analyses, this didn't succeed and by the late 1960s LEO's reign came to an end.

A Computer Called LEO, like other popular histories of science and technology, deserves to be widely read as it conveys much of the enthusiasm and excitement that working on new technology to tackle novel problems contributes. However, I was also struck by how many of the issues facing information systems development have been around from the very start of business computing.

For example, LEO provided its owners with a major competitive advantage, yet the impetus for the system didn't come from academics or consultants, it came from those who had a detailed knowledge of the problems they were trying to solve. Similarly, the LEO development team held many of the values now being rediscovered by information systems academics. The principles of scientific management in place at Lyons meant that the implementation of the system was carefully thought through in terms of its reengineering existing business processes; even then they realised that doing the wrong things super efficiently was not worth doing.

The LEO developers also emphasized the importance of what is now called human factors. The system removed much of the drudgery from many tasks performed by the teashop managers and, for example, allowed standardised orders and just‐in‐time order changes to be implemented.

The book also provides fascinating details about the way early computing was done. For example, binary code (the so‐called first level language of computing) was soon replaced by code that was more understandable to humans. Contrary to the popular history of computing, this changeover was almost immediate rather than gradual. One of the main reasons for this, of course, was that the first programmers for LEO had no experience with computers. Instead, they came from a variety of backgrounds and were taught to program from scratch.

I would encourage everyone to read this book and to reflect on their own views of computerisation as they do so. I am sure they will find the experience informative.