The Business World: Partners who prefer to stay just good friends

WILL ALLIANCES become the conglomerates of tomorrow? All sorts of arguments used to be advanced in favour of conglomerates by clever investment advisers, including the key one that certain particularly talented managers (Lord Hanson, Sir Owen Green) were able to make radical improvements to any business's performance, despite the apparent lack of any synergy between them. But the case was at best a transient one, applicable to a time when there were a lot of sleepy companies about that needed shaking up, and at worst was plain wrong.

Now the current fashionable idea is alliances - contractual links between companies that stop short of ownership (or if they involve an element of ownership, stop well short of control). They come in many manifestations, but are particularly prevalent in the giant network industries such as airlines and telecommunications. Here they are between broadly similar companies with different national bases and are presented as enabling both sides (or all sides, since there are often more than two partners) to gain from the relationship.

The rationale of an alliance within a network industry is that the network becomes more valuable the more people are connected to it. Enormous economies accrue to the group with the most extensive network. This would normally lead to the formation of giant cross- border corporations, such as the oil companies. But in these network industries a combination of regulation and national identity prevents such multinationals emerging. So the companies seek to obtain the advantages of scale without a change of ownership.

Thus in airlines there is the Star Alliance (where the main European partner is Lufthansa) and the British Airways-American Airlines Oneworld alliance. As anyone who travels a lot will know, such alliances mean that you find you are often travelling on a different airline from the one you thought you had booked - which is unsurprising as one of the key elements of these alliances is code-sharing arrangements.

Or consider telecommunications. The surge in international traffic is putting similar pressure on the telecom groups to form alliances, though here the process is less advanced and possibly less successful. There has been one attempt at an alliance involving co-ownership, the Global One partnership between Deutsche Telekom, France Telecom and the US company Sprint. But it is in difficulty, partly because of tensions between its German and French partners.

So what about the potential value of alliances in other business sectors? Here, paradoxically, they may have more of a future. It is almost as though the less obvious the need for an alliance, the more important it may come to be. Why so? Because of the speed of change, and the complexity of customer taste in many of the world's fastest-growing service industries.

To explain: in a manufacturing industry you need a product and a distribution mechanism. But the product you make will be broadly similar in each market. A BMW is essentially the same BMW whether it is sold in Frankfurt, Hong Kong or Los Angeles. But a bank account is not the same in those cities: the terms and conditions will be different and the customer/bank relationship will be different. The same goes for most other services - not all, witness McDonald's, but most.

Service industry companies that want to reap economies of scale can do so by setting up shop abroad, but this is rarely successful. They can take over local companies, but this creates management problems, for frequently they are mainly buying people and the people walk. It is also expensive and risky.

So instead there is a powerful case for service industries forming alliances, where different national corporations seek to learn from each other, and pass business back and forth. A new taste in Hong Kong might, just might, be a pointer to a shift in taste in New York. What alliances offer here is low-cost learning.

Now take the alliance game one stage further. What about alliances between different types of business within countries? This may be the most fruitful area of all.

Take a common example. A large company wants to extend into some new business; how does it do so? Option one is to set up a division. That is the cleanest way to do so, but carries risks, is expensive and may not work. Option two is to buy some other company, but that too carries risks, albeit slightly different ones.

Option three is to build an alliance. You find someone already in that area and get them to test it for you. You help them grow, maybe by giving them access to you customer base, in return for an understanding that if things work you will both share the spoils.

Thus a large management consultancy with no experience in a specialist (but growing) area might find a smaller consultancy and work with it. It can them offer the new service at little risk to itself.

Such arrangements require enormous amounts of trust. You can tie the arrangement up in complex agreements, but if you do that you lose speed and flexibility. You also may upset people, and in businesses where human capital matters more than physical or financial capital this is serious indeed. And as is now widely recognised, the great growth businesses are those that essentially rely on human knowledge.

I suppose the trouble with alliances is that, like conglomerates, the word is used to describe too wide a variety of different entities. Fast- forward 20 years and I suspect that it too will become a "sell" - the big formal alliances will be superseded by other arrangements. The principle, however, may be alive and well at a less formal level. What we really need is a new word to describe business relationships that involve co- operation but not ownership - partnerships, maybe, or even friendships. I remember, many years ago, asking Sir Siegmund Warburg how he had managed to establish a top merchant bank in London, after landing here more or less with the clothes he stood up in. "I had friends," he replied.