How to turn a nightmare into a dream ticket

This story is about an engineering/manufacturing company like thousands of others: at the big end of small, but the small end of big – effectively a minor multi-national that faces all the challenges of the large corporates, but without the resources or budget to do what they might. It didn't want to be named, and we respect that wish – but it's recent experience and resulting decisions are too useful to miss in this strategy issue. Just to paint the picture, we're talking about a company that serves the process manufacturing sector, with a massive, highly configurable product range, and with a couple of manufacturing sites in this country and a couple more in convenient locations around the world plus sales, warehousing and distribution through associates globally. It's turning over in the low hundreds of millions of pounds and employs 3,500. It's recently restructured at a management level – and that has been one of the keys to its improvement. But prior to that the company was founded on three global divisions for sales – Europe, the Americas and the Far East – with three sales directors and a supply director. Also, each of the associate companies abroad was run as an autonomous business unit. These and the offshore production sites grew semi-independently over many years, so our company has a considerable variety of systems, instances and revisions, supporting the different cultures, managements and processes. From an IT perspective, it was everyone's nightmare. Systems in the portfolio included BPCS (SSA Global), Mapics (now under Infor), Scala, Sage, Baan (also now SSA) and JD Edwards OneWorld (now PeopleSoft EnterpriseOne under Oracle). As our man says: "They were all different: some of the BPCS sites ran on IBM AS/400s, but others were on Hewlett-Packard. "We had several abortive attempts at ERP systems – we would claim to be world experts! For example, because of our management structure, we chose JD Edwards OneWorld in the UK for a new system. Three months later the French made the same decision. But although JD Edwards suggested we treat the French implementation as a roll-out from the UK, we didn't. The result was it was way over budget and time, costing more than £2m in the UK and £1m in France." Part of the reason for that was that the company also allowed each set of users to implement their own systems, so each tried to build one that matched as precisely as possible what it had been used to doing. "The result was a monster of customisation," says our IT man. And even more diversity to support. But the good news is that our manufacturer has in the last couple of years transformed its approach and is transforming its IT, aligning it with the new business needs – and doing so without too much pain, just determination. In effect, the company has developed a templated approach not just to its ERP systems, although that's one of the fundamentals, but to its implementation methodology – including the business analysis and the business processes, based on best practice. It's a profound culture change, it's working, and that's what's so instructive here. Backtracking for a moment, this has only been achieved by appointing a group IT manager in more than name only: ensuring that the role carries responsibility for corporate IT governance across the board. That is what's made it possible to develop a holistic IT strategy and, from that, also a group manufacturing strategy. Everything else has followed, with standard, workable solutions that aren't going to suffer either the politics or the familiar problems outlined. Says our man: "We are now going for one world-wide system that we have outlined in detail, and that system is PeopleSoft EnterpriseOne. The board has signed up to it despite the experience of our earlier ERP problems. And from a manufacturing point of view the group strategy is to go for lean manufacturing." Trial the strategy In 2003 he put the strategy and choice into effect when his team put PeopleSoft ERP into its Spanish affiliate more or less out of the box. "We were two months late on go-live on a six month project and 10% over budget. But the cost was around £350,000 for a 60-user site," he says. A far cry from the previous experience. "It's now our flagship ERP system; we've rolled it out to two more sites and now there's an avalanche of the other sites wanting it." There are some other critical success factors worth examining, the first of which is the management team's clear understanding of what matching the IT to the business strategy means, and how to make it stick. "Our IT strategy sets out our core set of agreed processes first and the core set of applications from that, second," says our man. "We did the overall business process mapping, but we do another business process mapping exercise for any site we're going to implement – so that the users feel responsible for their own changes. "From the business process mapping we go to the conference room pilot; we get a gap analysis against their existing system with the key people involved. And importantly, it's up to them to explain to us why they can't use our template solution – but if they can't we'll add the new function for them and for everybody else... We also get ISO 9000/2000 out of the documentation that process generates and it provides the core material for training as well: it steps people through the system and their roles and responsibilities." Lean thinking But this company is not content with its ERP success alone. "The next big thing is we want to put in lean manufacturing. We need to invest in lean to reduce our lead times and stock. People used to think that stock means service, but no-one can afford to think like that any more. Think of stock as lettuce – perishable items. Service and manufacturing have to work in sync to produce only what the customer wants on demand. "We're hitting set-up times hard to get lead times down, and that means some skill changes too. We can't afford to have dedicated production lines with some working their socks off and others idle. So we have to change to make all products on any of the lines. That's the agreed manufacturing strategy world-wide now." And he means it: process mapping has largely been done, trials are underway and the system platform to support it all is agreed. "We've already had success with this in our American associate, which used JCIT for lean manufacturing, implementing their DFT [demand flow technology] – but on their BPCS ERP system. Throughput there has been increased to the point where they can offer same day delivery on some products for a premium: it was three to four weeks before." How's it done? "They now only need to plan production two days ahead, and they're using BPCS MRP only for longer lead time materials. It's almost a manual system on the shopfloor – driven by customer orders with JIT to get components in. There's no more paperwork in the factory, no more booking at each operation – just backflushing. It's been staggering." Next it will come to the UK: "Over here we're still in the planning phase, although we have done tests. We're upgrading to EnterpriseOne 8.11 mid 2005 – we don't want to be the first to go live on the new system. We're going to put it into manufacturing in a controlled environment first, test it there and then roll it out." He admits to some very good fortune. "I have to tell you, we were very, very lucky. PeopleSoft EnterpriseOne contains DFT. I nearly fell off my chair when I found that out. It meant that all the parts of our business, IT and manufacturing strategy came together. That's another reason why EnterpriseOne is our ERP system of choice."