New Technology and EPOS

In its early days one of the things that set Woolworth apart was the excellence of its merchandising systems. In the 1930s, without the benefit of the technology that we take for granted today, a system of word-of-mouth and paper forms ensured that every week the Buyers at headquarters received regular, timely feedback from the shopfloors of the rapidly expanding chain.

The advantage was lost in the Fifties as the range grew rapidly and margins fell back. The Board held out against the tide as rivals moved to self-service. Instead money was pumped into computerising the depots and accounting. Despite an efficient central supply chain, the stores continued to do everything in long-hand until the mid 1980s. Individual order forms were used until 1977 (top left) when they were replaced by 'pre-print' forms (bottom). These were completed with a felt-tip pen in-store and then sent by post to the Central Accounting Office, where they were scanned into a computer for processing.

Stock counts and the requirements for the next eight weeks (known as 'stock commitments') were all calculated in-store, using cases and fractions of a case. For example, if a store sold 27 mugs each week and the mugs came in an outer case of 144, their commitment would be 8 x 27 cups (176), divided by 144, expressed as 1 case and 32/144ths of a case.

When the firm was taken over in 1982, a key challenge for the new Board was to update the systems and start exploiting modern technology. Group FD Geoff Mulcahy took control of the Woolworths sytems portfolio, and soon launched a Merchandising Systems Project. Microcomputers were introduced into the Marylebone Road headquarters in 1984, and were replaced by PCs when these became readily available for the first time. He also initiated analysis work in-store, looking for ways to simplify the bureaucratic approach to ordering.

A team led by new IT Director Chris French chose Telxon Hand-Held Terminals to automate the order capture process (right). Bar codes were stuck into the old Binders. Stores scanned them and keyed in the quantity required, based on preset re-order points and quantities which they had calculated themselves during the set-up. The supply chain was adapted to make more items available individually rather than in outer cases. All orders were placed in single units.

The hand-held terminal project was a big success, proving popular with the stores as well as improving efficiency. Before long a number of other applications were installed to help manage the space and prices and sales in-store. The hand-helds were successfully integrated with Electronic Cash Registers from Anker Data Systems. The acronyms 'HHT' and 'ADS' soon entered the stores' vocabulary. The devices captured and reported department-level sales and gave Managers an insight into how each range was performing.

This technology was very different from the equipment that shops use today. The devices were much less powerful than a basic mobile phone and had to be plugged together like an ipod to transfer files, before bluetooth and RF wireless technologies came along. The HHTs connected to a 14.4 kilobytes per second modem to send and receive data from the computer centre. They were 'dumb devices' and could only be reprogrammed by removing their storage chip ('EPROM') and taking it away for updating with an expensive, specialist machine.

Chris French and his successor Dan Bernard worked tirelessly to drive the technology forward. They introduced pre-printed Range and Order books in place of the old barcodes and binders and simplified the hand-held terminal apps. In parallel new central systems packages were integrated, including the World Wide Chain Stores System, which was used in the supply chain from 1987.

The next challenge was to overcome the difficulty in rolling out new applications to the hand-held terminals in-store. The process required engineering visits to update each of 5,400 tills and 2,000 terminals manually, which was costly and time-consuming. In 1989 each store was sent a back office PC. Initially this was used as a 'parent' for the hand held terminals, allowing software to be distributed to the stores much more quickly. Later specialist applications were added.

In a sign of how much computing has changed since, the Philips PC-compatible machines of 1989-90 had the maximum 640 Kb of Memory, 80286 chips running at 4 megahertz (1% of the speed of a top-spec machine today), and 'huge' 40 megabyte hard disks (about a twentieth of a £5 memory stick today).

Despite this the computers went on to revolutionise many aspects of store operation. In 1970 Store Managers had carried out their duties with only word-of-mouth from their staff. They had to wait for the results of annual stock counts to find out what had sold. By 1990 they were able to see sales, stock and financial information on screen, on demand, and to manage by exception. The PC initiative was such a success that within a year, the Company decided to go a step further. It began trials to implement EPOS scanning tills across the chain. Kingfisher was the last major retail group to embrace this technology.

The original Philips 80286 Back Office PC   |   Primary and secondary 80386sx Servers with racked 286 clients  |  The ORS560 Point of Sale

One of the key challenges in the move to EPOS was to get manufacturers to add barcodes to the packaging of the products.  The chain's eclectic mix had grown to more than 30,000 items. Each barcode had to be captured, validated and added to a central file. A team of colleagues in the flagship London store in Edgware Road, W2 started to capture these as work to select a supplier continued. Olivetti Systems and Networks won the £18m contract in collaboration with Post Software International. The initial system was piloted in the London store for a year in 1990/1 before moving into roll-out across the chain between 1992 and 1994. The total roll-out cost, including store training, building works and bolt-ons like credit card payment, exceeded £30m.

By the time the project's two Director sponsors, Dan Bernard (Systems, holding the ribbon, centre) and Martin Toogood (Retail Operations, right), cut the ribbon to mark the completion of the project in the Autumn of 2004, EPOS had already driven big changes in-store.

Transactions were faster and goods were no longer individually priced. Instead shelf edge labels told the customer the price and also helped to simplify stock control. There was also an automated replenishment system behind the scenes.

The simplification also allowed the chain to reduce staff numbers by natural wastage. Between 1990 and 1994 the number of employees on an 'equivalent full time' basis fell by a fifth, from 25,000 to 20,000. During that period the total headcount across the chain remained about the same, reflecting a move away from full-time contracts for most new joiners in the stores.

The business case for the £30m investment had anticipated these savings, which were duly delivered by the Retail team. However it proved much more difficult to persuade the Buying Teams to adapt their purchasing and product-management processes to exploit the richer and more up-to-date sales data.

The cost was incurred at a difficult time. Despite a bright new concept store in Houslow, West London, which was crammed with new technology, all was not well. Unforced errors in the supply chain caused shortages on the shelves at Christmas 1994.

The profits dip was accentuated by a shortfall in the anticipated commercial benefits of the EPOS investment. It took three years to force through the changes in buying practices that had been envisaged in the business case.

Some investors expressed concern when Kingfisher announced that they planned the same EPOS system to two of their other brands, B&Q and Superdrug, until they saw the faster and more efficient service at Woolworths for themselves.

The IT and Retail Teams worked together to plug the shortfall. They leveraged the new systems infrastructure to introduce new profit-generators in-store. Woolworths became the only national retailer to complete the preparatory work to sell National Lottery tickets in time for its launch in 1994. Camelot rewarded the achievement by choosing over 400 of the chain's stores to receive the tightly-rationed terminals.

A pilot loyalty card scheme 'Street Value Plus' was also initiated quickly and cheaply and helped executives to understand customer demographics and purchase patterns. In parallel new 'EPOS Promotions' were introduced. These rapidly boosted promotional sales from 12% to 20% of the firm's total turnover. The Buyers had recognised the potential, and had engaged with the initiative. By the launch they had negotiated supplier funding for all of the three for two, link save and Buy One Get One Free offers.