Inside the Machine

Five years ago, the managers of established, old-economy companies concentrated on running their business well: making cars, perhaps, or selling life insurance. They had to contend with constant change, of course, but normally of a fairly predictable kind: costs had to be cut, new products launched, mergers and acquisitions dealt with. Now life has become much more difficult. Change has not only become more rapid, but also more complex and more ubiquitous. Established companies are no longer quite sure who their competitors are, or where their core skills lie, or whether they ought to abandon the particular business that once served them so well. Behind this new uncertainty lies the Internet (which in this survey is used as shorthand to include the whole cluster of technologies that depend upon and enhance it). In the past five years, this has begun to transform managers’ lives.

Why is it causing so much trouble? After all, the Internet as now used by many companies performs familiar functions, although more cheaply and flexibly. The e-mail is not really so different from the memo; the electronic invoice looks much like an on-screen version of its paper predecessor; the intranets that companies install to connect different departments resemble the enterprise resource planning (ERP) systems that many companies bought in the 1990s; even the networks that link companies with their suppliers had their electronic predecessors.

But new technologies often begin by mimicking what has gone before, and change the world later. Think how long it took companies to realise that with electricity they did not need to cluster their machinery around the power source, as in the days of steam. They could take the power to the process, which could even be laid out along a production line and set in motion. In that sense, many of today’s Internet applications are still those of the steam age. Until they make the next leap, their full potential will remain unrealised.

Yet even what the Internet has already achieved is puzzle enough for many managers. Why should it cause more bewilderment than, say, the arrival of the mainframe or the PC before it?

The answer lies in the Internet’s chameleon qualities. It is not simply a new distribution channel, or a new way to communicate. It is many other things: a market place, an information system, a tool for manufacturing goods and services. It makes a difference to a whole range of things that managers do every day, from locating a new supplier to co-ordinating a project to collecting and managing customer data. Each of these, in turn, affects corporate life in many different ways. The changes that the Internet brings are simply more pervasive and varied than anything that has gone before. Even electricity did not promise so many new ways of doing things.

At the root of the changes is a dramatic fall in the cost of handling and transmitting information. Almost every business process involves information in some form: an instruction, a plan, an advertisement, a blueprint, a set of accounts. All this information can be handled and shared far more cheaply than before. That has its drawbacks, of course: a fall in production costs is all too likely to lead to an increase in supply, and plenty of managers now feel that they are drowning in information. But it also brings immense advantages.

In particular, the investments that companies need to make in hardware and software are small in relation to the pay-off. Gary Reiner, chief information officer of GE, one of the pioneers of the Internet, describes how the company set out to build its own electronic-auction site. “It was less expensive than we thought,” he says. “We built the software for $15,000 internally.” Charles Alexander, who heads GE Capital in Europe, makes the point even more forcefully. “What we have rapidly begun to understand is that the incremental investment required is extraordinarily small compared with our overall investment. A $300m investment for a company making $4 billion-5 billion is weeks of cash flow to get a payback which is months away, not years. We thought, if we are completely web-enabled in two years, that’s good. Then we realised we could do it in weeks — and the productivity gain is almost instant.”

Of course it is not really that easy. Companies have to do a lot more than buy terminals and write software. Erik Brynjolfsson, a professor at the Massachusetts Institute of Technology’s Sloan School of Management, argues that software and hardware account for only about a tenth of true corporate investment in information technology. A far larger investment goes into new business processes, new products and the training of employees. Such spending does not show up as investment on corporate accounts. Instead, it generally appears as expenses, such as payments to consultants, and is treated that way by the taxman (unlike investment in old-economy physical goods, which can be capitalised and depreciated). Nor does it appear on national accounts. Yet American companies, Mr Brynjolfsson calculates, have created a total of $1.5 trillion of “organisational capital” in the past decade.

A survey of 416 companies conducted by Mr Brynjolfsson, together with Lorin Hitt of the Wharton School at the University of Pennsylvania and Shinkyu Yang of the Stern School at New York University, identified such organisational capital by picking out companies in similar industries where employees tended to have a great deal of authority over the way they worked; where they were paid for performance; where they were well educated; where the emphasis was on teams; and where most information flowed freely across the company. Financial markets valued organisational capital highly, to judge by market capitalisation, but they were keenest on those firms with both high investment in information technology and a high degree of organisational change.

How’s Your Organisational Capital?

The reason for this correlation, Mr Brynjolfsson believes, is that in order to get the best out of their information technology, companies need to make a host of changes in a co-ordinated fashion. They cannot simply mimic others: if they do, the results may be disappointing or even disruptive. Organisational capital is far harder to reproduce than the more visible, marketable sort.

So the success of a company’s Internet strategy depends on the way the company is run. It is not just a matter of designing a shrewd strategy — although that obviously helps. In addition, a company needs strength in depth. It needs intelligent, empowered employees, a culture of openness and a willingness to experiment, good internal communications and a well-designed pay structure. It also needs absolute commitment from the top. The sheer scale of change is impossible without determined leadership. A company whose chief executive never deals with his own e-mail will not get far.

Many managers began to think seriously about the Internet only once the millennium-bug scare was over. Most large American companies started to develop their Internet strategies a year ago; the benefits will begin to show up only in next year’s accounts. A survey conducted earlier this year by the National Association of Manufacturers found that more than two-thirds of American manufacturers did not use the Internet for business-to-business commerce. If that is true for the United States, it is true in spades for the rest of the world. “Whenever I visit software companies,” says Andrew McAfee, a professor at Harvard Business School, “I get them to complete the sentence, ‘The business-to-business revolution is x% complete.’ The biggest number I have heard is 5%. Many say 1%.”

Some of the remaining 99% will be accomplished only as the network effects of the Internet feed through. For example, only half of Dell Computer’s customers are “online-enabled”, as Joe Marengi, one of the company’s senior executives, puts it: able to use Dell’s web pages to configure their orders. And only 15% of customers place the order electronically. The remaining 35% design the order online and then submit it in some way that requires Dell to take a second step to feed it into the system, such as e-mailing or faxing it. As more companies can submit electronic orders, both they and Dell will cut costs and speed the process. Many more processes involving transactions among companies bring the same double benefits, and ripple outwards as more firms share some particular capability.

The companies that have gone furthest claim to save astonishing amounts. GE plans to cut 15% from its cost base of $100 billion in both 2001 and 2002. That is five times the typical annual growth in productivity, even for this fast-moving firm, of 3-4%. In addition, the company hopes to reduce the prices of the materials it buys by making most of its purchases in electronic auctions. That should save a further $2 billion over the next two years.

Other companies are discovering the same sort of magic. In the summer of 1999 Larry Ellison, boss of Oracle, announced that the company would cut $1 billion from its global corporate expenses of $7 billion. Now it expects to cut a second billion by October next year, and has its eye on a third. “I’ve been in business for more than 30 years, and I think this is by far the biggest productivity advancement I’ve seen in my life,” says Jeff Henley, Oracle’s chief financial officer.

Such figures will have a dramatic impact on the overall productivity of the economy. In America, productivity growth appears to have accelerated sharply in the past few years — although Mr Brynjolfsson would argue that the acceleration would look milder if the costly investment in organisational change could be properly measured. Even if the pioneers of the Internet are exaggerating, a long period of big productivity gains seems to lie ahead.

These gains come from several different directions. Some of them are achieved simply by transferring part of the work from the company to its customers. If you are having trouble recruiting enough staff for your help desk, and you want to spare your customers having to listen to an entire CD’s worth of gruesome music on hold, then well-designed online information may actually benefit everybody. If your customers can submit electronic orders, it saves you the trouble of doing the job yourself, and also greatly diminishes time-wasting wrangles over mistakes. You just e-mail the querulous customer the original order.

Other cost savings come from being able to feed better information to suppliers all the way up the chain, and thus reduce stocks. Paul Bell, who runs Dell Computer’s operations in Europe, the Middle East and Africa, describes inventory as “the physical embodiment of bad information”. His company now measures its inventory in hours rather than days. Still more savings come from being able to buy from a wider market, allowing competition to drive down prices.

These changes allow measurable savings. But there are other, more far-reaching benefits. As companies install software applications to take over many of the tasks that employees now do — such as “running errands” to keep information moving — they alter the balance between the internal and external demands on a company, an issue described in a book called “Harmony: Business, Technology & Life After Paperwork” by Arno Penzias, former boss of Bell Labs. “Though to my knowledge no computer has yet managed to replicate the performance of a single office worker,” he says, “the right combination of computing and communications can frequently replace whole departments.”

The “errands” jobs are not only wasteful; they turn the attention of organisations inwards towards the smooth operation of internal processes, rather than outwards to the customer. Now, though, Internet-based software applications are shifting the balance, shrinking the amount of human time and effort that needs to be spent on internal co-ordination. Companies will no longer need layers of white-collar workers to manage the steps between what the customer wants and what he gets.

All sorts of things can be done differently. Manufacturers can talk directly to their customers, or to their suppliers’ suppliers. Customers can click a mouse and start a production process rolling, far along the supply chain. Training sessions can be carried on a laptop. Sales staff can do presentations to customers a continent away. Companies will often find it difficult to tell whether such measures save money or whether they simply provide a better service, but overall they will increase the proportion of time that goes on keeping customers happy rather than keeping the business running.

Once a company sets off down the Internet track, what does it find? First, a change in familiar boundaries, starting with those of the firm itself. Collaborating with others becomes easier and less expensive, as does linking different operations within and between firms and buying in everything from management skills and innovation to the human-resources department. The boundaries for employees are redrawn too: those between home and work, as people work from home and shop from work, and those between the individual and the company, now that employees’ knowledge and skills may be worth almost as much as the company they work for. Geographical boundaries start shifting as well: within companies, different regional divisions need to collaborate and share customers more than they did, businesses can source more products globally, and new ideas and competition can spring up anywhere on the planet.

Life in the Fishbowl

Second, the Internet brings management out into the open. “You put a piece of glass into your organisation and expose all your internal strife,” says Pete Martinez, who runs the worldwide consulting arm of IBM’s global services. Managerial privacy dwindles: “I have to assume that every bit of information about me is broadcast back to our employees and customers. It’s the fishbowl effect,” says Eric Schmidt, boss of Novell, a troubled software company. If you have failed to reply to an e-mail, you can no longer hide behind “my secretary lost the message”. Pricing is also more transparent, as more deals can be put to the test of an occasional auction. Companies need to allow customers and suppliers “inside the machine”, in the phrase of Peter Martin, editorial director of Internet activities at ft.com. Thus customers can track the progress of their orders, and suppliers are growing used to scooping information straight out of their customers’ databases. Moreover, employees can see what they might earn in similar jobs in other companies.

Third, the Internet increases the importance of standards. Indeed, the glue that holds it together is essentially a set of software standards. Their user-friendly simplicity allows people to use the Internet in many different roles — as customers, suppliers, employees, job-seekers — without needing to be retrained. Electronic commerce needs standards in order to make it easy to transfer information between companies with different systems: hence the importance of XML, a programming language. Companies also need rules about what can be bought online: the aggregation of many departments’ orders for staplers will save money only if all departments are willing to buy from the same standard shopping list. And aiming for a standard technology, and standard look and feel for customers, is a way to reduce maintenance costs and to measure more easily how customers behave. However, standardisation is a force for centralisation in companies. Once the human-resources department decides it needs a single global website to keep employees up-to-date, head office will want to decide what should appear on it.

This survey begins inside the company to see how the Internet affects the way managers communicate with staff, and staff with each other. It then examines links with suppliers and customers. Lastly, it considers how the shape of the company itself may change, and offers some guidelines for good e-management.