98 per cent wrong?

“The majority [of software vendors] are still not talking in terms of ‘bottom line’. They use terms like ‘better visibility’, ‘easier to implement’. How can that improve the bottom line? They’re not even thinking the right way.” And thus they are failing the manufacturers they purport to serve. So says Dr Eli Goldratt, the somewhat self-styled father of advanced planning and scheduling (APS), and author of the management best seller ‘The Goal’, the book that catapulted him to guru status. Goldratt retired four years ago at the age of 50 after more than 20 years of fathoming what’s wrong with manufacturing and its supporting IT. Born in Israel, he had majored in physics at Tel Aviv university, and gone on to gain a masters and doctorate before moving to the US and developing his theory of constraints software and methods. He set up shop in 1979 as Creative Output selling OPT. “That was ‘optimised production timetables’,” he recalls, and adds, “it was the first APS in which routings, bills of materials and the rest generated a schedule.” It’s a fascinating story. “OPT was very successful: in 1983 Creative Output was one of the six fastest growing companies in the US,” says Goldratt. But in 1985 he decided there was a huge problem that couldn’t be ducked. In companies that implemented his revolutionary methodologies the results were outstanding: but with manufacturing so much more productive, the result quickly became rapidly growing inventory and longer lead times. “The constraint moved on to another function somewhere else,” explains Goldratt – warehousing, distribution, sales, engineering, whatever. “Then if they trimmed back on people everything collapsed.” He says that unless the user’s sales were growing at an amazing rate and everything else was going well, this was invariably the outcome. So as chairman of the company, he instructed his sales staff not to sell OPT unless it was clear that the prospect could safely double sales in short order – while he developed software to sort out what was clearly becoming an embarrassment. “There weren’t many of those,” he observes wryly, “so we started to lose sales. I said ‘that’s fine; we have enough cash’. But the other shareholders said it wasn’t fine – and I found myself out.” It was a telling moment in Goldratt’s career, and you sense the frustration. Today he is sanguine: “the rest is history,” he says dismissively. He founded the Goldratt Institute in 1986 – and for the next 10 years devoted his and his staff to developing software and methodologies to deal with the issues across the business functions. In 1997 he finished, and captured his achievements in satellite broadcasts and a series of videos in 1999. Now, his new book, ‘Necessary, but not Sufficient’ – the digestible form of his conclusions – is again rattling the establishment. So why exactly does Goldratt think we are still failing? Because, he says, we’re missing the point – too many are still going for departmental benefits, local optimisations, or treating IT as a panacea. And none of these is enough. “They fall at the first hurdle – that of seeing things from the business perspective, and making that perspective holistic,” he says. “Unfortunately, 98% of manufacturing is still in the same position as it was 20 years ago,” he says. “I remember my amazement then. I went to companies like the Gatling Gun plant in Belmont, RCA Television, Sikorski. The software they had was basically computerised JIT and Kanbans with buffers everywhere. And I always found one resource 100% utilised and the rest 50—60% utilised: everything else was waiting for it. Now I know that that was the bottleneck, the constraint. They were telling me everything was 98% utilised, but it just didn’t work out.” And sadly it’s still happening. “You need different rules, policies, measurements – they’re just different words for the same thing,” insists Goldratt. His point, admirably made in the new book, is that you should only deploy new technology if it adds value – “if it reduces today’s limitations”. And if it does that then why stick with the old ways and behave as if the limitation still exists? “Take ERP: the limitation is ‘I don’t have the data that I need now’. So manufacturers’ rules have always been ‘do whatever you can’; ‘do local optimisation’. But those rules have become embedded in the culture: even with new ERP the ‘rules’ get built right back into the software.” And this is what has to change, he declares. So they have to change the rules: that’s common sense isn’t it? “Let me tell you something. If you want to win the Nobel prize for physics you have to write just one article such that other physicists reaction is ‘On shit, it’s common sense’. If it’s not common sense it’s probably not right.” And he continues: “The Goal is like a bible: everybody believes it, but very few implement it. Look at it this way: if 2% of the world’s industry is using it, that’s still thousands of companies – but still 98% are not. People are trained to think that to achieve major changes they have to go through ascent after ascent after ascent. All these things – JIT, TQM, cost accounting and all the other methods. There is no silver bullet. But Archimedes said the opposite: give me a leverage point and I’ll move the earth.” They can’t believe that doing something relatively obvious and small can make such a difference, he asserts. “This [changing the rules] is the silver bullet. But they still don’t do it: they’re thinking ‘it can’t be [that] the whole world is wrong, so it must be a fairy tale’. And how many people don’t care what other people think? Probably one or two per cent!” Does it work? “For sure. Four years ago Texas Instruments’ internal news letter spoke of savings equivalent to the cost of two wafer plants: that’s over $1 billion! GM never talks about it, but this year it said the bottom line benefits from theory of constraints assembly line and all production plants was $1 billion. Lucent did the same thing, but on their engineering and product development site, the fibre optic engineering division, and found that in 1998—99 they reduced the time to develop to half, tripled product output and improved on time delivery from 60% to 90%. Is that a result?”