The Hindu Business Line : Intelligence and the corporation

BUSINESS corporations are living beings. They exhibit intelligent behaviour — or at least are supposed to. After a futile search for a theory of intelligence of the corporation with convincing first principles, I decided to evolve one myself.

I have been privileged by my consulting career to research the behaviour of companies big and small. Initially, the episodic observations seemed disjointed. Later on, one was fascinated to uncover some truths that explained the intelligent behaviour (or otherwise) of business corporations. Here are three postulates:

Postulate number 1: The intelligence of a business corporation is inversely proportional to its size.

No business corporation, big or small, whatever originating nationality, appears exempt from this rule. As a company's size grows, its intelligence declines.

Increasing daftness with increasing size is a natural law of the business corporation. The behaviour of the corporation is a function of its decisions. As a company becomes bigger, the decision-making layer gets cut off from the external realities.

The decline in intelligent behaviour is simply explained by the information theory. There is loss of information both ways. First, there is loss of information when it travels from the external realities through the lower layers to the decision-making layers. Next, the decisions that the decision- making layer makes with imperfect information get distorted as they travel from the core to the lower layers during implementation.

The distorted decisions implemented using distorted information manifest as the intelligent behaviour of companies.

Is there a way out? Yes and no. No, the decline in intelligence with size cannot be eliminated or even arrested completely. Yes, the decline can be reduced.

Think of the usual path of information travelling upwards through the layers and of decisions travelling downwards as Path A. The greater the number of layers Path A has, the more the chances of loss of information. The decision-making layer has to resort to the creation of Path B, an alternative path that permits direct exposure to external realities. In a limited sense, using Path B, the decision-making layer gets to do the real work of the company.

Members of the decision-making layer spend a tiny part of their time in developing a product, manufacturing a batch, handling customers' calls, collecting cash from some select customers and so on to get a direct sense of the goings on in the real world.

Think of the more intelligent big companies, and recall the behaviour of their CEOs and senior executives. You will see them talking to customers, participating in developing a series of products, occasionally taking the wheel of a machine to produce a part, and calling a customer to find out how their products worked. Do not mistake this for an act of heroism.

They are not trying to prove that one man can manage it all.

They are creating Path B by which they try and minimise the traditional loss of information they suffer through Path A. The decision-making layer now has large-scale information, albeit distorted, through Path A and reasonably precise, though small-scale information through Path B.

Chances are, tempered by the real world experience from Path B, the decision-making layer takes better decisions.

The problem with Path B is that if it is too small, it be may be statistically insignificant to base decisions on it. If it becomes too large, the decision-making layer is now trying to do the job of everybody else in the company, an impossible task surely.

Theoretically, a large Path B would obviate the need for Path A. It becomes Path A. When Path B is the right size, the probability of taking the right decision is vastly improved. If you think of it carefully, Clayton Christiansen in his famous book The Innovator's Dilemma was contending with this postulate.

Similarly, loss of information in implementing decisions can be reduced by Path B. Path B during implementation could mean that the decision-making layer participates directly in implementing the decision in a limited cross section: In one plant, in one country, in one industry segment, in one technology and so on.

This has the additional advantage of providing the much-needed feedback about decisions so they can be modified, if necessary, before being applied on a larger-scale.

Again, think of the largest and the daftest of the corporations you know of. Study them carefully. There will be little or no Path B activity.

Empowerment is the phenomenon by which operating employees are supposed to take decisions closer to the action. In general, the greater the empowerment, the less the decline with size.

However, as in everything else, Pareto's principle applies to decision-making as well. Eighty per cent of a corporation's behaviour will be dictated by 20 per cent of the really strategic decisions.

For the foreseeable future, unless somebody comes up with a radically different construct for the corporation, such decisions will be taken only by the decision-making layer isolated from real action. See discussion in the next postulate about one such construct: Reorganising a large company as a collection of small companies.

Postulate number 2: The most competitive corporation has the least decline in intelligence.

If decline of intelligence with size is inexorable, certainly the rate of decline can be managed. Reducing the rate of decline constitutes astute management. Between two companies, the one that loses less intelligence with size will be competitively superior.

Leadership consists in designing a statistically significant Path B such that it permits the decision-making layer to do what is contracted to do: Take decisions about the real world without being distracted by operational intricacies.

The origin of every successful large corporation is in a spirited, intelligent, entrepreneurial setup. Later when the corporation grows up and starts acting stupidly, those who have remained with the company since its genesis wonder what incurable illness overtook it that it started behaving so inexplicably? Therein lies the lesson.

The decline in intelligence of a corporation is inversely proportional to its ability to model small company behaviour.

The aspiration to model small company behaviour was one of the original quests of Jack Welch as he remade GE. "Work-out" and other programmes were installed to ensure that if the corporation were to stumble on great insights, it did not pick itself up and walk away as if nothing had happened.

By encouraging small company behaviour, I am not arguing for a kingdom of chieftains which suffers from diseconomies of scale, fissiparous outposts, and associated maladies.

I am talking of a clever orchestrating act that holds together the small companies within the big company by using an appropriate model of federalism.

Remember when you were a small company, you took risks and conducted experiments that other people laughed at. Nobody laughs at you now when you are a big company.

But nothing exciting happens. Decline in human intelligence, it has been pointed out often, happens due to the disuse of reasoning faculties. The same thing applies to business corporations, too.

Postulate number 3: The intelligence of the corporation is less than or equal to the intelligence of its CEO.

The CEO here does not necessarily mean only the CEO though in most instances it does. It can include senior executives who are empowered to take significant business decisions.

The notion of intelligence is not limited to the traditional idea of IQ. It encompasses all leadership qualities: People management, emotional intelligence, technology management, financial management, visionary capabilities and strategic thinking.

Nobody expects the CEO to be a superman excelling in everything. But he should be absolutely distinguished in one or more of the aforesaid qualities.

It is possible that in brief spurts, the intelligence of the corporation actually exceeds that of a mediocre CEO.

Soon enough, the lesser intelligence of the CEO tyrannically imposes itself on the company (the theme of most Dilbert jokes). Over time, the intelligence of the corporation approximates that of the CEO.

The corporation then starts to under perform his intelligence asymptotically. It does not matter that the corporation may be staffed with people of prodigious intelligence at other levels. The most significant decisions are those of the CEO.

A few of his decisions will have a sweeping impact over zillions of other less important though more intelligent decisions made by the operating folks. Sustainable greatness is not the accidental product of mediocrity and ineptitude.

Clearly, more postulates are required to make the theory of intelligence complete. Maybe you should try to add to this body of knowledge. At least, I am trying hard. Watch this space.

(The author is Director-Consulting and Enterprise Solutions, Satyam Computer Services Limited. The views are personal.)