I have an idea for a new performance-measurement tool. Companies need a hard and fast rule that will enable them to measure the point at which any specific line of communication has become so stretched that it no longer works. If such a measurement were possible, most companies would find that what they fondly see as lines of communication on some internal chart are actually non-existent.

Or they could simply talk with - or even listen to - their workforce. But of course they don't do that. And the larger the organisation, the truer that is. Take Barclays Bank. As we all know, this organisation exists because, as the advertisements tell us: 'a big world needs a big bank'. Recently I paid a rare visit to a local branch of this Big Bank to pay in a cheque that someone had sent me.

The place was filthy, the queue stretched to the front door and there were only two cashiers on duty. When I eventually reached the front and paid the cheque in, I asked if the cashier could pass the message upwards to the manager, and then further up the line, that by and large the policies they were following were not making them bestloved by the customers. 'No point in doing that - they wouldn't believe me,' she said.

It turned out that she had been drafted in from another branch, so any complaint she passed on would have been seen as special pleading. Doubtless whoever is responsible for communications at Barclays still fondly believes they are getting accurate feedback through the ordained channels, and have no idea that the system is discredited and useless.

Or take the Post Office. Sub-post offices have a system, soon apparently to be dispensed with, of area managers to whom queries can be relayed. These managers have a computer system that provides back-up. But they are so terrified of being second-guessed by managers further up the line that, although they may know the answer to a query from the manager of the subpost-office, they always consult the computer first. This means that the manager asks the question. It is repeated back to him, followed by the sound of a clacking keyboard, and then a pre-scripted answer follows. This can backfire. A request for the number of the human resources centre got the response: 'Why does he want to know that?'

Those are two small examples of large organisations with probably no idea that their internal lines of communication have stretched so far that they no longer work.

The solution that most management theory would produce would be to lay down new lines of communication both upwards and downwards, and, quite possibly, sideways as well. The snag is that this doesn't deal with the problem. The problem is not one of communication. It is one of organisational size.

Organisations are nearly always too big. And the way in which dealmaking is seen as almost the only management skill that gains rewards makes it worse. Almost every research study of the effect of mergers and acquisitions shows that mergers almost always fail to create any extra value. In most cases they lose it. But what they do every time is make an organisation even more unwieldy.

What the business world needs is a rule that shows the point at which efficiency and effectiveness evaporate. Even better would be a rule that keeps companies below a certain size. Anybody in business who can state an honest opinion, which isn't coloured by the need to hype things up for the sake of share options, pensions or bonuses, will tell you that beyond a certain size businesses become a shambles.

Now every senior director knows this. They know how large a team can be before it loses its ability to achieve its goals. The cut-off point is somewhere between 12 and 18, depending on the personalities and skills involved. Extrapolating upwards from that shows you how almost any large business you care to mention has lost the ability to be efficient and fleet of foot. It is very simple. Businesses only thrive up to a certain size.

Using sport as an analogy to point up business and management truths is usually a hopeless and simpleminded wheeze. But there is one business truth that becomes obvious only when the sporting analogy is made. And this is about the size of a team. If you find yourself being treated to some corporate hospitality at a sporting event, listen to the conversation. Are any of the captains of industry present standing there and opining about the size of the teams? Are they announcing that this game of rugby, for example, would work much better if you doubled the teams to 30 a side? Of course they aren't. They know full well that the teams in the game in front of them work best at the size they are, and that increasing it significantly would lead to chaos, confusion and, eventually, breakdown.

But put them back in the boardroom and they would argue the opposite. They would tell analysts that the latest proposal sold to them by an investment bank would create greater synergy, greater shareholder value, and would double the size of the company. They would talk of critical mass to be gained, and so on.

For the vast majority of serious and senior companies, exactly the opposite is what they need to do. Analysts should devise a computer model applicable to different sectors. Once a company approached a certain size, red lights would flash on the screen and 'sell immediately' notes would be despatched to investors. From a shareholder point of view, breaking up a company and floating its constituent parts as separate and smaller companies is almost always seen as increasing the total value. It is quite extraordinary that company chairmen and their boards of directors have never understood the implications. But presumably their own personal communication systems have become so stretched by their egos that they no longer receive coherent signals.