According to studies carried out by INSEAD Professor of Strategy Gabriel Szulanski, in conjunction with Sidney Winter from the University of Pennsylvania and Robert Jensen of Brigham Young University, companies need to identify and validate actual examples that have been shown to produce results. These ‘templates’ or benchmarks of best practices should, the authors argue, then be copied with minimal if any modifications or adaptations, at least in the initial phase of the knowledge transfer.

Companies “don’t know why what they do really works,” Szulanski says, “and they realize that when try to transfer their practices.”

“They would normally start by copying only those things that they think are important, to discover later on that they leave out important things and also that they copied things that are not only totally unnecessary but also sometimes even comical.” The reason for this, he says, is that there may be a certain amount of ambiguity about factors that are actually driving the success of a particular practice. “One way to really be able to get the performance in the copy is by copying exactly the original and then going back to the original and comparing it to what you’ve done.”

Using templates can produce 'startling' results

“So if don’t have a template, by which we mean a working example, then it’s a lot more difficult to get the results you want,” Szulanski says. And the results you could get can be startling. “In many cases the difference in performance between units of the same organization is incredible. We’re talking about an average of 200 per cent difference between a good case of best practice and not-so-good practice. In some cases, I’ve seen differences of up to ten times!” He adds that nowadays the increase in effectiveness may be around 30-50 per cent, but that’s still impressive.

In a working paper called ‘Template Use and the Effectiveness of Knowledge Transfer’, Szulanski and Jensen investigate Xerox Europe’s efforts to transfer practices within the firm over an eight-year period. The first and third ‘waves’ of transfers were successful, while the second failed, even though all three had equal support from top management. Nor, it seems, was complexity an issue. “It was startling to us,” Szulanski says. “The main reason (for the failure of the second wave) was that they didn’t have a template, that is, they didn’t have a working example of the practice they wanted to replicate.” That made it more difficult to guide employees in other countries and to demonstrate what to do with the new processes.

Szulanski says people may also feign interest to go along with management’s pet project which is seen as the ‘flavour of the month’, especially if a company is in a ‘state of flux’ at the time. Consequently, they ‘may feign enthusiasm and implementation until management’s attention shifts to the next initiative and stops paying attention to this specific one.”

To avoid replicating ‘flash-in-the-pan’ bad practices, which might even be successful for a short while through good luck rather than fundamentals, companies should aim to validate best practices and make sure they are copied to the letter, at least in the initial phase of transfer.

Some adaptations may be necessary, but don't do these in advance

“We believe that ultimately you need to do some adaptations because every environment is different. But what we believe is that it’s not smart to do those in advance because you need to understand the practices in the new context. So they take them somewhere else and find interaction is complicated and they don’t work,” Szulanski says.

For example, an American firm trying to transfer operational routines to China would likely fail if it tries to pre-adapt its practices, especially if it were to ‘read China wrongly.’ “So when they bring the practice to China they make changes and when things don’t go right, they can only compare these with the original example, but they are not sure if the changes are correct or (whether) implementation went wrong. So we advocate you’re better off copying exactly what you have in the original and then seeing what makes sense and what doesn’t and make adaptations accordingly.”

Starbucks is another good example, Szulanski says. When it was set up in Seattle, it copied an Italian model down to details such as opera music and standing areas only. You couldn’t even get coffee with skimmed milk. “So they initially copied everything but when they looked at what really didn’t make sense in the American environment, then they began to adapt. So we advocate a similar process. You have to adapt, but we advocate that first you should take everything, copy the original, see how it works in the new environment and only then make the adaptations.”

The same goes for outsourcing, he says. Toyota, for example, was successful in training its suppliers to meet its standards for precision parts, so it could maintain both quality control and its competitive edge.

When building semi-conductor plants elsewhere, Intel would copy the design of its original fabrication plant, even though this may have been taken to an extreme.  “They copied everything,” Szulanski says, “including a door in the new factory, even if it leads to nowhere (in the original design the door had led to another building). While it may seem a bit extreme, it’s hard to argue with the result,” Szulanski argues.

Before you adapt, know what you're doing

Szulanski says firms should transfer their best practices and make sure they’re carried out to the letter, ‘initially’. “They don’t have to stay that way because learning is important. What we are saying is that before you tailor, change, adapt, you have to know what you’re doing and then have certification to make changes.”

So if a company is looking to make use of the knowledge it already has, rather than creating new one, it needs to have an actual working example of what it wants to transfer, at least a pilot. “Once you have a template, we recommend you copy that template precisely to begin with and once you have experience in the new setting you begin to modify so have the opportunity to compare the original with the copy so if something goes wrong it’s a lot easier to diagnose.”

Szulanski cites the example of a German chemical company, which recently opened a new plant in Thailand. Although it copied all its processes to the letter, it ran into unexpected problems. But because the firm had replicated its processes exactly, it was able to pinpoint quickly the cause of the problems: different water quality in Thailand as compared to Germany. “If they hadn’t copied (the processes) exactly, it would have taken much longer to identify the root cause of the problem and rule out alternative explanations” he says.

“Having a template in the first place is fundamental,” he adds,  “but also it’s a very powerful (method of) persuasion, because if people say it cannot be done, you can say, ‘Yes it can, let me show you’ and that’s the essence of the persuasive power of the template.”