Listen to this article
When Somerfield, the UK supermarkets chain, analysed what staff in its 200-plus stores were actually doing each day, it found that for 30 per cent of the time the answer was not very much.
“We looked at store activity,” says Neil Kinnerley, head of finance for the group’s convenience division, “and almost one-third of the hours worked added no value to operations at all”.
Closer examination found very different patterns of demand across the business: at 3am the branch at Manchester University is in peak trading mode, for example, while in commuter areas 5pm to 6pm is the busiest time. By analysing the workload and then building staffing levels and skills to match, Somerfield expects to cut its staffing costs by 0.5 per cent of net sales.
The company – in this case using IT applications from Torex Retail – is one of the first in Europe to embrace workforce optimisation (WFO), a new approach to staff scheduling that integrates demand forecasts and employee skills to create a “best fit” model to reduce costs while maintaining service levels. It is a complex technology demanding sophisticated algorithms and a willingness to spend at least two years on implementation.
“Workforce optimisation is intellectually challenging,” says Ian Lenagen, chief executive of Workplace Systems, “and people tend to undervalue the degree of complex mathematics that is involved. A lot of projects simply don’t get the algorithms right, but if you do then the savings are massive – easily 10 to 15 per cent of the labour costs.”
Many such systems originated in the US where flexible labour laws have led to such developments as “zero contract hours” allowing staff to be called in for work as required without any guarantee of regular employment. The US bias, however, limits the software options for others: “A problem for Europe,” says Judy Sweeney, vice-president of research at Boston-based AMR Research, “is that there are currently not enough IT vendors capable of handling EU-related employment rules.”
Mikael Bisgaard-Bohr, retail industry director for NCR Teradata’s European operations agrees: “Optimised scheduling is not so easy in Europe. Instead there is a lot of interest here in trying to improve staff productivity by analysing sales transactions to identify the high performers, encouraging and retaining them, while retraining poorer staff. Even a 0.5 per cent improvement in productivity can have a major impact on the bottom line.”
German cash and carry business Metro is implementing systems from Workplace Systems throughout its operations in central and eastern Europe – adding still more employment law complexity.
It is a long-term programme not due for completion until well into 2007 but already the returns are significant. Initial benefits have included average increases in productivity of 6 per cent – up to 20 per cent in some stores – with a 50 per cent reduction in customer queues and a 0.4 per cent cut in wage costs per hour.
“Employees also have the possibility to influence their own schedules via a web kiosk,” says Berthold Steur, Metro C&C senior department manager, who has been responsible for the project. “And in the longer term we expect to decrease wages per hour still further as well as define future employment contracts in terms of workload.”
WFO is best suited to sectors with variable but largely predictable demand and with a large workforce – typically 2,000 or more. “In the US retail and call centres are leading in workforce optimisation,” says Mrs Sweeney. “There is also growing interest from other service providers including fire, police, hospitals and transport.”
US vendors such as Kronos, BlueCube and Reflexis are all moving into Europe and adapting their systems to cope with EU directives. Reflexis has already sold systems in the UK to retailers Marks and Spencer and B&Q with other European contracts expected shortly. International retailers are also driving activity. BlueCube, for example, supplies optimisation systems to fast food chain McDonald’s.
“They wanted to go global with the system and so we’ve set up offices in Europe,” says Martin Pashley, BlueCube’s UK director.
As all companies appreciate, skilled and experienced staff are a valuable commodity – expensive and difficult to replace.
By matching skills to tasks and demand companies can not only save money by eliminating that non-value-added time, but staff satisfaction also improves so those valuable workers are less likely to move on: a win for everyone, it seems.
Organising for optimisation
A sophisticated staff schedule may be the final result of a workforce optimisation project but it is like the tip of an iceberg – there is an awful lot of underlying material.
Analysis of the tasks involved and the time each takes is needed – in some cases it is classic work study material.
Toys R Us in the US used RedPrairie’s labour management system to calculate exactly how long each task in its warehouse should take. This then drove task allocation and performance targets for staff that led to a 15 to 20 per cent improvement in productivity.
Demand also needs to be accurately assessed and monitored. Traditionally, demand forecasts are based on analysis of historical performance but increasingly real-time inputs are also involved.
In retailing, sales data is often fed back to central systems at five to 15 minute intervals rather than batched once a day. This can trigger staff activity – changing shelf-fillers to cashiers, for example – in response to real events.
Kronos, the US vendor, talks of “demand driven workforce management” with North American retailers now defining staffing requirements in 15 minute intervals. In large retail parks, for example, one person may work a number of very short shifts for several retailers during the course of the day but each business will be achieving optimum staffing levels as a result.
There also needs to be an accurate database of staff skills, performance and availability – and, ideally, flexible employment contracts to allow working times to be changed easily in response to changing demand. WFO also requires some clever and very complex algorithms to find the best match between demand levels and suitable workers.
The algorithms must take account of employment rules – such as the European Working Time Directive – and also create practical solutions giving employees realistic shift patterns and not rostering overtime on days when they are unavailable for work.
Systems can also be personalised so that high performing staff, who the organisation is eager to retain, are given shifts that match their preferences more often than less valued employees.
It also takes internal resource and commitment: typically it takes seven to 10 minutes to schedule a supermarket optimally. Extend that to perhaps 2,000 stores in a chain with a weekly schedule and the requirements stack up quickly.