Lessons from LEO the Lyon

With global politics and economics taking a turn for the worse you may be forgiven for having missed a special birthday reported in the back pages of the newspapers. In the last weeks of 2001, business computing turned 50.

It takes a touch of genius to realise military computers can be used for something as simple as counting cakes. But Joe Lyons, head of Lyons & Company, the king of UK tea-time treats, had just such a brainwave when he decided to spearhead the development of computer technology for business after the Second World War. The company needed to automate its complex supply chain and accounting system (sound familiar?) and the Lyons electric office bakery valuation project (LEO) was born.

Lyons’ team went to the US in the 1940s to have a look at the new-fangled machines that were being used to split atoms and make calculations for ballistics. They never saw them – their meeting was cancelled – but they came back to the UK, hired a team of university academics (including a young Margaret Thatcher) and set to work on their own machine. LEO’s valves were first warmed up and used in anger in November 1951, and the LEO computers kept running until the mid-60s.

One of the most important lessons from LEO’s inception is not only the simplicity and practicality of the application (Lyons needed to control stock so he built a computer to do it – and nothing else), but the way the business was more important than the IT. Instead of consultants and tech-wizards, clerks and office workers from Lyons helped write the programs and install the machines. They knew the business best.

The aim of today’s financial reporting and accounting software is much the same as LEO’s, as David Caminer, LEO’s head of research noted in a speech* delivered at the Guildhall in London to mark LEO’s birthday. “The whole objective was to enable management at all levels to take action without having to go searching for needles in a haystack of the Lyons profit and loss accounts. So many companies, even now, don’t know where their profits are draining away,” the 84-year-old said.

As with so many great British inventions, LEO was superseded by work in the US. Faster, card punching, IBM machines were heavily marketed while the Lyons team failed to advertise their success. End-users began to lose control of how business computing evolved. It remains to be asked what lessons the IT industry has learnt in the past 50 years from LEO’s practicality and understated success. Not very many it seems. Today IT is based on scale – speed and capacity come first and the functionality seems like an afterthought.

The need for speed took hold in the early 1980s. One of my colleagues remembers the first business computer at stockbroker Philips & Drew in 1985. The spreadsheet functionality blew him away, but the big button on the keyboard marked ‘TURBO’ was its killer feature. Pressing it meant the computer would run at 12MHz rather than the paltry standard 8MHz.Yet most of the software available wouldn’t run at the higher speed.

Most home PCs now run at over 1GHz and every couple of years the amount of memory needed to run business applications doubles. We’ve come a long way from Bill Gates’ apocryphal 1981 statement that “640k should be enough memory for anybody”. But just as LEO was 16ft long (and much has been made of the bulk and clunkiness of early computers) systems departments are still bursting at the seams. They are crammed with all sorts of boxes with flashing lights, many of which are under-utilised, out of date or even redundant.

So while businesses are still using IT to monitor stock levels and p&l accounts, we now also have a plethora of computing capacity that engenders a must-have approach to software. Instead of company executives seeking solutions to existing business problems, multi-million pound ERP systems and performance visualisation tools are implemented on the dubious promise of immediate and sizeable ROI – only to become legacy systems as technology makes its next leap.

Certainly finance departments will always need accounting and reporting software, and the excitement surrounding data aggregation technology such as XML is justified. But, in these straitened times, when IT vendors and consultants come knocking and claim to know your business better than you, sit back with a cup of tea and a slice of cake and think before you buy.

LEO and the Computer Revolution by David Caminer is available at https://is.lse.ac.uk/leo