During the early 1990s, many observers of the personal-lines property-casualty insurance industry¹ believed that "direct response"—the sale of policies via toll-free telephone numbers and without the help of intermediaries—would storm the industry and topple its leaders. It didn’t happen. Instead, traditional insurers beefed up their technology, slimmed down their operations, and made their traditional channels sufficiently competitive to hold on to the greater part of their business.

Today, most insurance companies, or carriers, continue to distribute policies through agents: either captive agents, who sell the policies of a single company, or independent agents, who distribute policies on behalf of many. A smaller group of carriers sells policies through other channels, including banks and work-site marketing programs, or direct to the customer.

So far, few of the industry’s incumbents make much use of the Internet. Judging from their limited response to it, many large incumbents, whether they sell directly or through agencies, think they can hold back the tide of electronic commerce by using the same incremental defenses that defeated the direct-response threat. No doubt, these incumbents are encouraged by industry data showing that retention levels remain high and that most consumers are "very satisfied" with their carriers. Research shows that on-line sales are small, and market observers predict that these sales are likely to remain below 4 percent of total insurance sales for two more years. Histori-cally, it has taken either price differences of at least 15 percent or poor claims execution to prompt customers to move.

But the complacency of the incumbents is misplaced because the Internet, a disruptive technology, is set to transform the competitive landscape in personal-lines insurance. Already, more than one million individuals visit the sites of the on-line operators QuickenInsurance and InsWeb every month. And almost three-quarters of Internet users plan to resort to the Web during the insurance purchase process at some point in the next 12 months. Given the limited number of options now available on-line, these are remarkable numbers.

While incumbents take their time, the new on-line attackers are busy enhancing every step in the business system (Exhibit 1). The site of InsWeb, a "front-end" marketer (and one of Yahoo!’s "50 Most Incredibly Useful Sites"), has more than 50 participating carriers, including AIG, The Hartford, Nationwide, and Progressive. The site offers auto, home, medical, renters, and term-life insurance, as well as specialty lines such as motorcycle policies, home warranties, and pet health plans. It also features frequently asked questions, glossaries, an archive of articles, and tools that allow customers to calculate the cost of different kinds of coverage. Quotations can be sent via e-mail or telephone within three days or delivered on-line by certain carriers in real time.

QuickenInsurance offers similar services. And what InsWeb and Quicken-Insurance are doing with referrals, InsurQuote and Insurance Holdings of America are doing with instant underwriting. (IHA has kiosks in Sam’s Club stores, a chain of discount warehouse centers, offering instant quotations and coverage.) At the back end, CCC Information Services, CarStation, and Cybersettle are improving the effectiveness of claims procedures.

As these innovations transform the business model, bringing down commissions and improving back-end operations, consumers are likely to benefit. Although prices won’t drop as sharply as they have in, say, stock brokerage (because about 70 percent of every auto insurance premium dollar goes to pay for claims), declines of 10 to 20 percent can be expected. A discount on this order of magnitude could, our research suggests, persuade up to 43 percent of all customers to switch from their current provider (Exhibit 2), even though 85 percent of consumers report being "completely" or "very" satisfied with it. Even allowing for the fact that consumers often overstate their willingness to switch, the shift in market share is likely to be significant.

And the Internet has only just got going. Although it has yet to shift much market share and can still be awkward to use, Internet pure plays will surely improve the experience because they have no other way to compete. They will streamline their application procedures, use broadband to deliver real-time advice on-line, and offer real-time underwriting. Gradually, more and more people will become comfortable with such services.

What the attackers are doing

Attackers are employing a number of models to help them take on the incumbents. The answers to a pair of questions—who owns the customer, and where in the business system is value being improved?—make it possible to map current players on a grid, which shows that most of the action to date has been concentrated at the front end of the business system (Exhibit 3).

InsWeb and QuickenInsurance are two of the most familiar protagonists in that space. The agent-based marketplace or referral model that they have adopted allows consumers to compare insurance quotations from a number of competing carriers (the exact number varies by product). Once consumers have chosen their carrier, they are referred to a local agent, or vice versa, to close the deal.

These services significantly enhance the consumer’s convenience and ability to shop for a lower price. But they do little to improve the economics of the transaction—partly because the companies offering them add a layer of fees while in most cases still requiring an agent to close the deal. Responding to the need to add value and improve the customers’ experience, InsWeb and QuickenInsurance are beginning to play the role of agent for selected carriers, even to the point of closing policies.

Meanwhile, on-line insurance companies that can themselves complete the transaction in real time are emerging. This second, fuller model, whose practitioners include eCoverage and esurance, won’t be successful overnight. To carry it off, companies will first have to achieve superior underwriting and claims-handling skills—no easy task for start-ups.

A third model, that of product bundling, offers insurance for cars or homes at the point of sale. Carriers are rushing to set up partnerships with on-line auto and mortgage Web sites such as Autoweb and iOwn. The shopping service Autobytel is a leading operator in this segment, though sales have been limited to date because of an inability to bind coverage on-line.

The fourth model involves auction sites. Ebix, for example, invites consumers to name a target price for auto, health, home, or life insurance; agents and brokers then make bids on the basis of the details that the as-yet unidentified consumer has submitted about the person or item to be covered. As soon as the consumer accepts a bid, contact information is sent to the successful broker, which then pays ebix a fee. At present, however, regulations of US states forbid carriers to drop their prices without refiling their rates for risk classes—an onerous and time-consuming process.

Consumers who consolidate their financial relationships would be the basis of the fifth insurance model—that of the financial supermarket—which for the time being is hypothetical for personal-lines property-casualty insurance. This model would emerge if financial-services providers such as Charles Schwab bundled auto or home insurance products with their current offerings. The extent of consumer interest is unclear, however, and supermarket owners would probably be reluctant to risk their brands on a bad claims experience.

A sixth model could develop if intermediaries were to start using the Internet to aggregate consumers into finely drawn risk classes that they would then auction to carriers at regular intervals. Such a step, which requires regulatory approval, would expose carriers to further pricing pressure.

Each one of these models is attractive to on-line consumers, according to the findings of a recent McKinsey survey. When cost reductions likely to be generated by the various models were attached to descriptions of the models, 41 percent of consumers said they were "very" or "somewhat" likely to switch to agent-based marketplaces, the least attractive model, and 57 percent to consumer auctions, the most attractive (Exhibit 4).

In addition to these business-to-consumer (B2C) models, various business-to-business (B2B) players have emerged to take costs out of the back offices of carriers.

What should incumbents do?

Some carriers have already begun to act. State Farm has formed alliances with Autoweb and others, Allstate has introduced an on-line multichannel offer, and AIG has launched AIG Direct for on-line sales. In addition, Progressive has a comprehensive on-line capability that leads the field.

But among the almost 30 leading incumbents in property and casualty (P&C) and life insurance we studied, we found little evidence that the hard choices and the investments needed to develop an Internet-age business model have been made. A few companies are permitting their customers to use the Internet to undertake basic functions such as locating agents and reporting and tracking claims. A few more are arming their agents with additional on-line capabilities that let them track commissions or trace a consumer’s payment history—capabilities that improve efficiency without threatening the incumbent’s distribution partners. Most incumbents, though, appear to hope that the high level of customer satisfaction still being expressed for their traditional models will allow them to evade the challenge of the Internet and the difficult business of dealing with the issues of cannibalization and channel conflict (see "Click and save," in the current issue).

In light of these developments, the Internet is likely to generate two stages of evolution. The first, lasting three to five years, would see traditional insurers losing share to new attackers, to specialist intermediaries, and possibly to diversified financial-services companies seeking to extend their brands into new product lines. The second stage would see a wave of Internet-related consolidation as falling returns and the rising cost of acquiring customers drive the weakest incumbents out of business or into one another’s arms.

In view of these likely developments, what should incumbents do now? Companies that have yet to get moving need to mobilize four kinds of Web strategy: a Web enablement strategy, a portal strategy, an investment strategy, and an attacker strategy.

Prepare your company’s Web strategy

Companies will be forced to exploit the capacity of the Internet to improve efficiency or it will merely be an additional cost center

To be successful, companies will have to sell policies over the Internet. As carriers move toward that goal, they will have to exploit the Internet’s capacity to improve the efficiency of every stage of the traditional business—otherwise, the Web will be just an additional channel and cost center.

Carriers also need to start integrating their on- and off-line customer-management systems, for customers won’t accept a company’s inability to interact seamlessly with them through whatever medium they find most convenient. Above all, companies must begin to manage their agents differently. The best agents will find that the Internet allows them to devote to higher-value selling some of the time they had spent on servicing. Such a shift will widen the performance gaps among agents. Rather than fighting this trend, companies should think about moving less productive agents into roles involving a greater degree of servicing and support. Many will earn lower commissions, though from a larger set of customers. Some, inevitably, will find themselves out of a job.

Decide on a portal strategy

How should an insurance company draw customers to its site? Too many companies spend a lot of marketing money placing banner advertisements on mass-market portals. Although these offer the broadest access, they are the least focused and may not be very effective at drawing traffic to insurance sites. Instead, companies should think about "vertical" industry portals and "segment" portals. Vertical portals are sites devoted to a single industry or function: for example, insurance content sites, on-line product marketplaces (see "The real business of B2B," in the current issue), sites devoted to personal financial services, and sites that attract consumers ready to purchase such products as health or auto insurance.

Segment portals are subsites of the mass-market portals; an apt example for insurance companies is Yahoo! Finance. Such portals offer focused searching capabilities to Internet users interested only in a single range of products. Although carriers that form strategic partnerships may be successful with financial-services portals, less well-positioned companies could find them ineffective—they might not attract enough insurance shoppers to justify the cost of advertising on the site. Instead of paying the standard rate for advertising space, carriers should consider paying a higher rate for deals concluded after a click-through. (Regulations may, however, require a portal that wishes to collect such payments to be licensed.) If the portal had a strong desire to be associated with the carrier’s brand, the carrier might be able to reverse the charges.

Devise an investment strategy

For three reasons, established carriers should think about investing in Internet insurance businesses. First, the terms may be favorable, since start-ups will probably seek a recognized carrier’s expertise, resources, and stamp of approval for their new ventures. Second, getting close to an Internet insurance company is a low-risk way for an established company to test the validity of a business model it may wish to move toward itself. Third, large incumbents can use their influence as investors to shape the personal-lines industry in ways advantageous to them; a carrier could, for example, induce a start-up to make its technology platform compatible with the carrier’s business model.

Carriers need not invest actual money to reap the rewards of an investment policy. Although financial capital for start-ups is relatively cheap and plentiful, brand and operational skills are not.

Consider a full-scale attacker strategy

In the short and medium term, incumbents should take the measures recommended above. But in the long term, exploiting the opportunities the Inter-net offers will probably mean acting like an attacker. Incumbents ready for that course can follow either of two paths. One is to turn a skill in which the incumbent is strong into a service business; for instance, a carrier with a strong core skill such as back-office policy administration could craft a subsidiary providing that service to other carriers. The same could be done in such areas as sales force management, call-center support, vendor management, and claims handling. A carrier might even want to commercialize an agency-management system it had developed, though its technology would have to be superior.

The other path calls upon companies to attack their own core businesses. An agent-based carrier, for example, could launch a direct insurance play on the Internet under a different brand. Such strategies, of course, risk cannibalization and channel conflict—powerful deterrents to action, since no company wants to offer on the Web something it can sell off-line for more money, and disgruntled agents will assume, generally correctly, that sales through a direct channel will siphon off sales from their own. But companies that hold back from cannibalizing themselves or from inviting channel conflict could be consumed by competitors. At any rate, the simple act of pondering different attacker strategies might help to identify the most significant threats to the core business.

Should the whole industry migrate more and more to the Web, incumbents will of course want to be part of the process. Those that lack the stomach for cannibalization can at least start behaving like attackers in areas where channel conflict isn’t an issue—such as servicing, vendor management, and dispute resolution. While these B2B forms of e-commerce might be less glamorous than B2C selling, they are just as likely to yield significant improvements in economics, and they will be useful preparation for a company’s eventual wholehearted embrace of the Internet.