Success or failure of any system, ‘e’ or not, depends to a massively underrated degree on the attention you give not just to your business issues, but specifically to your people. John Dwyer hears it from those that learnt the hard way
One day in late summer a phone rings in the office of an anonymous northern industrial academic. The caller, a production director, had read of his work on agile manufacturing and wanted to implement it. “By Christmas.” The academic rages, then loses the power of speech: “The idea that you’re going to implement a major change like that in three or four months!”
His view is as valid for ERP or e-commerce as for agile manufacturing, and his despair is understandable. The technology can be tricky enough. Even harder is to persuade users of its importance, not just for the business but for them as individuals.
Gerry Moran, general manager of electrical and electronic connector maker FCI of Fermoy in Cork, Ireland, says persuasion has little to do with reward systems. FCI’s 300 employees meet between 1,700 and 2,300 orders a month. In adopting enterprise software vendor QAD’s Mfg/Pro ERP system and moving to consultancy Oliver Wight’s definition of Class A business performance, Moran says, “Reward was never discussed. If you want to change something, you must be changing it for a reason. It’s your job to sell that to your people, and [show how] they will benefit from it as well.”.
The measures tell the story. FCI’s customer delivery performance now averages 95%. Inventory accuracy, stuck at 85% in May and June of last year, now averages 98%. Adherence to master production schedule is 95% or above, and so on.
Commit 100% to it
Moran confesses that, at first, managers and supervisors were the biggest obstacles. When Moran and Oliver Wight held a three-day, off-site meeting for 17 key staff early in 1999, one middle manager proved so hostile Moran sent him back to the plant in a taxi to clear his desk.
Moran proved his commitment to the project by attending all three days. But, “Just because you provide training for people, that doesn’t mean they have bought in,” he says. After all, Moran recalls, “We thought we were pretty good...” FCI already met 91% of its delivery promise dates. What FCI hadn’t faced was the heroic scale of fire fighting needed to achieve this. And that is a key point. Moran acted: “We had the teams. We decided to allocate a mentor to listen to what each team had to say and encourage them to move forward and make decisions faster.”
Another OW client, £232 million international metal-products company Luxfer Gas Cylinders (LGC), also had a taste for heroic fire fighting. LGC makes cylinders for fire extinguishers, oxy-acetylene equipment, medical gases, soft drinks and underwater breathing apparatus. The firm decided late in 1996 to pursue OW Class A MRPII based on SAP ERP software to drive all the businesses forward. Andy Butcher, director of LGC’s European operations, says adopting Class A and sales and operations planning (S&OP) has “transformed our business performance” over the last five years.
When LGC first surveyed its customers it found a number of problems. Butcher recalls that LGC’s quality was ‘best in class’ but its lead times and delivery performance were the worst. And there was sharp downward pressure on prices. None of LGC’s five UK and US sites had consistent manufacturing or measuring processes, and each put great value on fire fighting. There was inventory everywhere, Butcher recalls. Managers devoted much energy to reacting to the surprises their forecasts hadn’t warned them of.
Luxfer had a visionary CEO, Ian McKinnon, prepared to lose his UK managing director rather than compromise the MRPII ‘mandate’. The MD, sceptical about MRPII, “was soon pursuing a career elsewhere,” says Butcher.
The story was similar at Moss Plastic Parts, a Kidlington, Oxford-based supplier of protective plugs and caps, masking products, electrical grommets and blanking plugs, cable ties, washers, spacers and nuts and bolts. Five years ago Moss started rolling out SSA’s BPCS ERP system across over a dozen European sites. The main driver, says Sarah Cobb, Moss’ business systems director, was to see where all the company’s inventory was.
Most of Moss’ business is standard components, sold off the shelf for orders taken by fax, phone, or Moss’ website. Its catalogue holds over 5,000 parts available from stock for delivery the next day: “Stock availability is critical to us,” says Cobb. “If a customer wants it they’re not going to wait three weeks for a grommet – they’ll go somewhere else.”
Inevitably, Moss’ processes changed. Warehouse staff used to wait for a delivery note to come out of the printer. Now that they could see forward orders they had to plan their work forward as well. Cobb recalls: “We did have one or two casualties [at the supervisory level] who just couldn’t cope with the change.”
Moss had begun by appointing a full-time project team. Cobb had been head of logistics before taking up her present role. FCI didn’t do this at first. Then it took two managers away from what they were doing. FCI logistics manager Alan Burdett remembers being told, “Don’t do anything on the shopfloor today.”
LGC had to learn the same lesson. It first appointed an ‘engineering and IT manager’: “This didn’t work at all,” Butcher says. Too frequently he was “diverted from important long-term work to deal with the latest short-term crisis out on the plant.” Butcher’s appointment followed, and that of a full-time, cross-functional project team.
Moss’ approach to software was to install it in two small pilots – distribution centres with six or seven employees. Then it installed at Kidlington and paused to draw lessons. LGC, on the other hand, was determined to tackle the processes first. S&OP had been functioning for two years at LGC by the time the IT went in. Duncan Banks, LGC’s British-born US ERP manager, says: “Unless you are ready to go, don’t touch the software... [We decided that] all measures of performance would be class A before going live. That was key.”
Both LGC’s US and UK operations faltered when launching the Class A project to their workforces. In the UK LGC used an interactive presentation to everyone in the business, and posted updates, a planner and minutes of meetings on a notice board. Surveys showed employee awareness was poor. A common view was that “MRPII will go away. It will be TQM again, then Kaizen again.” LGC’s managers then adopted quarterly briefings, one-to-one sessions, posters, a newsletter, free gifts and other incentives to take an interest. Surveyed again, the UK workers showed not just excellent awareness but a belief that this project would be seen to a conclusion.
LGC is now reaping benefits. The 50 steps to produce a purchase order have now been cut to 10. It used to take five days at month end to close the month’s sales. Now they know the month-end totals by midday on the first day of the following month. Where it once took 30 days for material to go through a factory, it now takes seven. In 2001 over 80% of business can be despatched in 72 hours or less. Even more important, during 1999 the number of on time and in full deliveries in the UK business, which had been running as low as 25%, consistently reached 100%.
Banks and Butcher conclude that Class A releases all the benefits advertised for it “and more”, provided that you give first priority to the people involved in the business, second to the processes, and only last to the systems. A salutary statement.
Tell it like it is
For FCI, perhaps the most important turning point was the installation of a machine which they could only ramp up to half its expected workrate. FCI had given customers schedules based on the higher, unattainable rate. For six months managers argued about whether they should accept the low rate and tell customers or do more firefighting. Eventually, they levelled with the customers and, from then on, only promised what they could really do – a 95% delivery to promise rate.
Now Moran never rejigs schedules for the customers or anyone else. He refers such requests to schedulers to examine the impacts on other orders. “Everyone has to go through ‘the proper channels’,” says Moran. “If you can give them what they want by working a weekend or whatever, fine, but you’re not going to hurt any other customer as a result.” Moran’s advice is: “In simple terms the commitment and involvement must be from the top. You’d be very fortunate to get it to work otherwise. It will work – people are natural innovators and they want to achieve an improvement – but it will be much slower, the net benefit will take longer and will probably cost you more.”
Moss, meanwhile, had another cultural jolt in the last year when it adopted e-commerce: “Customers can go on the net, place an order and it is directly as if they were keying it into BPCS and were seeing stock,” says Cobb. She did several presentations about what this would mean. At one, “I told [the warehouse that] customers were going to see stock and place orders. There was a shocked silence, then one of them said, ‘That means they’ll be able to see our stock on the Internet!’. ‘Yes.’ ‘You can’t do that: it might be wrong!’.
“People have to get through their heads that, if they’re putting something up on a system, and they have the pan-European visibility that we do, then they have to be absolutely sure that people really buy into the fact that the information is correct. So data integrity is paramount.” Cobb sums up: “However much time you think you’re going to need to spend training, make sure you double it.”
Treat training holistically
LGC goes further. All its employees had at least eight hours of training. Even so, it “totally underestimated the training we needed.” Based on this experience, LGC describes the system training as “totally insufficient”, and advises, “Think of a number for the total hours required for training and multiply it by between five and 10!”.
Other pitfalls are less obvious. Cobb says too much of the burden of change can fall on too few. At one pilot site, users referred all system problems to a user who had taken to the system easily. So they didn’t learn, and the system wasn’t as effective as it was meant to be because the ‘super-user’ suffered stress. Cobb adds, “We learned a lot not just about how the system should work from a system point of view, but how the system should work from a process point of view.”
An example was that people were taught the sequence by which they could override a credit block on a customer’s order. But they weren’t told that they should only use the sequence in certain conditions: “It’s not just how to do things, but when to do them,” says Cobb. “It’s the difference between how to do it and why you’re doing it.”