If home is where the network is, who pays?

During the next few years, experts at McKinsey say home networks will become a reality. But should it be free to consumers?

Ever since the Internet entered the popular culture, futurists and technophiles have been telling the world that the new medium would transform homes into information-rich hubs of activity.

Refrigerators, they predicted, would someday monitor the expiration dates on milk cartons. The family room would double as a videoconferencing theater. The toaster and the microwave would engage in endless Socratic debate.

Three or four years on, none of this has happened, and some of it may never happen, since consumers are likely to see many gee-whiz applications as more trouble than they are worth. Yet home networking is far from dead: In the past three years, the underlying technology has undergone its own quiet revolution. Big interests are at stake, and companies such as Intel, Microsoft, and 3Com have been diligently working out the bugs; others, including Cisco Systems, Ericsson and Pace, have been testing the new technology in homes to see how consumers react.

Such companies have found that although talking toasters may remain a fantasy, people would even now be willing to use more prosaic applications that make it possible for them to communicate more easily, to find a broader range of entertainment options, and to assume greater control over their homes. Clearly, the major broadband providers--of both cable and digital subscriber line services--should be extremely interested in home networking.

Indeed, the current demand for simple, nuts-and-bolts applications may already be strong enough for broadband providers to think not only about including home-networking hardware in their installation packages but also about doing so free of charge. And if companies in a major industry have an economic incentive to supply the consumer with this equipment at their own expense, any business with an interest in the technology should be formulating plans right now.

Running the numbers
The hardware and software that enable devices in a home to share voice, video and data can reside in a stand-alone unit or be incorporated into DSL modems, cable set-top boxes, PCs and other devices. The equipment either broadcasts short-range radio signals or sends data over telephone or electrical wires, thus reaching appliances and computer peripherals throughout the customer's home.

Software developers, including Microsoft and Sun Microsystems, have created protocols that make it possible for devices ranging from refrigerators to laser printers to recognize and understand one another. Meanwhile, during the past two years, semiconductor companies such as Broadcom, Intel and Lucent Technologies have introduced new integrated circuits that provide for this type of network, using only radio frequencies already allocated for such purposes or the wires already installed in the walls of homes. The capabilities of these chips have made giant increases: Current models can move 10 megabits of data a second through telephone lines or the air, and fairly soon, top speeds are expected to reach 32 megabits or more--enough to deliver several DVD-quality audio and video signals simultaneously. Most manufacturers of computing devices and home appliances plan to start putting home-networking chips into their products this year.

If the trend continues, in three to five years the inhabitants of the developed world will wake up one day to find their homes network enabled, much as they found their television sets equipped with remote controls a decade ago.

To understand the strength of the economic case for unleashing home-networking equipment free of charge to consumers, consider only the most basic, nuts-and-bolts applications: services that people pay for today even without home networks but that become much less costly delivered through them. For example, a home network could easily link two or more computers in a home--a job that would otherwise require a technician to go there and install an Ethernet card and a local area network. However, it does make sense for cable providers to ally
themselves with game makers by installing game capabilities in their
set-top boxes; doing so would raise the cost of a home-networked cable
box by only $40, to $330.

When all the computers in a home talk with one another, they can share printers, scanners and a single Internet connection, thereby realizing big cost savings in homes with a number of systems. Similarly, a home network allows a single set-top box to bring cable service to several TV sets, eliminating the cost of extra boxes. A third basic service is the installation of extra phone lines in homes that have at least one. Home networking makes the job much less costly by allowing new phone jacks to plug in to standard electrical outlets and by creating virtual fax numbers or second phone lines on demand.

Assuming current levels of demand, we estimate that in the United States alone, home networking would cut $14.7 billion from the cost of hardware, installation, and upkeep if it were applied only to in-home computer networks, cable service to additional TVs, additional telephone lines and home security systems. And that figure is merely the average cost savings multiplied by current demand; it makes no allowance for the higher demand that would likely result if prices fell. Even so, the savings amount to at least $140 a year per U.S. household, on average.

In fact, any consumer can enjoy some of the benefits of home networking right now, for a one-time cost of $300 to $400, by purchasing a stand-alone home-networking box. But home owners are reluctant to spend such large sums for benefits they still consider highly speculative.

Broadband companies have the edge
Cable and DSL companies, however, are in a particularly good position to provide those benefits on their own. For them, adding the appropriate hardware to new cable set-top boxes or DSL modems would involve a one-time cost of only $40 and $108, respectively. In view of the possibility of saving $140 per household, installation clearly makes sense from the perspective of overall economic efficiency. In most developed countries, both cable and DSL providers are just beginning to increase new hardware installations. Over the next five years, roughly 40 percent of U.S. homes will acquire digital set-top boxes or DSL modems.

To unveil home networks, cable and DSL providers could include the necessary chips in their new set-top boxes and modems and then continue to install the equipment exactly as they plan to do now. Broadband providers might therefore make an ideal conduit for bringing home networks to the mass market.

Of course, not every product that creates efficiencies comes to market. To finance the installation of home-networking gear, the broadband companies will have to find a way to capture at least some of the money it saves. If all of the money goes into the pockets of consumers, the deployment of home networking may have to wait until they are willing to pay for it--and that probably won't happen for several years.

With this in mind, we asked ourselves how broadband companies could benefit from providing home-networking hardware to consumers free of charge. The first way might be higher average revenue per customer.

People who use a home-networking security system instead of a traditional burglar alarm, for example, may be willing to split the cost savings with the broadband provider by paying monthly service fees that incorporate some, but not all, of the savings. Such people might also be willing to pay for ancillary services, such as a "nanny surveillance" camera, on the same system. Most important, home networking greatly increases the value of the new video, voice and data services a provider might make available, since they would no longer be tied to a single room. Consumers might be willing to pay for that sort of convenience.

A second possible benefit to broadband companies would be a reduction in the rate at which users cancel services--either by giving up broadband entirely or by switching to other providers who supply it along with telephone and cable TV service. As people come to rely on their broadband provider not only for high-speed Internet access but also for services such as computer networking, they will presumably have more reasons to stick with that company; dropout rates--or churn rates--should therefore decline.

Finally, the services available through home networking are likely to attract more customers, increasing the penetration of broadband, generally.

Do the math
Improvements in three measurements--monthly fees, churn rates and penetration rates--would be needed to justify giving subscribers free home-networking equipment. We looked at all three. First, we put ourselves in the shoes of a U.S. cable provider such as AOL Time Warner or Comcast. Whenever these companies undertake a broadband or a digital TV installation, they leave a $250 digital set-top box, for which they themselves pay, in the subscriber's home. As we have seen, incorporating a home-networking capability into those new boxes would cost the cable company an extra $40, with essentially no increase in the cost of installation.

U.S. cable providers could make money including home-networking equipment in their digital set-top boxes at their own expense if the extra features cut the annual churn of cable subscribers by only 1 percentage point, down to 32 percent, from 33 percent, or increased the cable modem penetration rate by a modest 5 percent, to 5.8 million U.S. households from 5.5 million, by the end of 2001.

Alternatively, if home networking delivered extra fees of just $2 a month on average, free installation would make sense even with no change in the churn or penetration rates. Furthermore, any combination of the churn rate, the penetration rate and average revenue might be enough to justify free installation. The message is clear: For cable providers, the time to provide home networking free of charge may have arrived already.

For DSL providers, such as Verizon Communications, BellSouth and Telocity, the economic case, though not quite as strong, is still impressive. Adding a home-networking capability to a DSL modem costs the provider an extra $108. If the added functionality reduced the yearly churn rate of all telephone subscribers--long-distance, local and DSL--by 2 percentage points, down to 22 percent from 24 percent, that extra revenue alone would justify free installation. Even if the churn rate didn't drop at all, an 8 percent increase in DSL penetration--to roughly 2.7 million U.S. households from 2.5 million, as of this coming December--would also make free installation a winning strategy. The same would be true if the average user paid $4 more a month for home-networking services.

All of these number targets, we think, are within reasonable ranges: Most subscribers, for example, pay more than the required $4 a month just for call-waiting, caller ID and other services.

To finance the installation of home-networking gear, the
broadband companies will have to find a way to capture at least some of
the money it saves. The case for free deployment is obviously a bit better for cable companies than for DSL providers. But when existing boxes are replaced by new ones equipped with home-networking gear, the economics on the DSL side improve. New cable boxes with home-networking chips cost providers $290, while new DSL modems with them cost only $170. DSL companies thus have a lower cost hurdle to surmount when they decide whether to replace existing broadband boxes with network-enabled ones.

The home stretch
Right now, the biggest decision that all cable and DSL providers have to make is whether to unveil home-networking equipment quickly or slowly. Industry economics, as we have seen, may support immediate release, but some players may want to bide their time. For instance, if either cable or DSL providers believe that they can distribute network-enabled modems much more quickly than the other can, the speedier group may want to hold off deployment: Why place a bet early if you can afford to wait and see how events unfold?

If the providers of one or the other technology believe that their current boxes are harder to retrofit than those of their rivals, that would argue for moving more rapidly--before too many more non-networked boxes are installed.

The strategic combinations are many, but the point is simple: The decision to move quickly or slowly expresses a company's beliefs about the strengths of the competing technologies and the evolution of consumer tastes.

One thing appears certain: In three to five years, home networks will have moved off the pages of science magazines and into millions of U.S. homes. And millions of people, seeing a home network as a huge opportunity to save money and obtain valuable new services, will embrace the technology with enthusiasm--even if their toasters remain silent.

Authors Jacques R. Bughin, Renee C. Foster, Alan Miles and Luis A. Ubinas contributed.

For more insight, go to the McKinsey Quarterly Web site.

Copyright © 1992-2001 McKinsey & Company, Inc.