Feb 20th 1999: The gathering momentum towards free Internet access is set to boost online application services, but only as a short-term tactic. Long-term, free access bundles will limit user choice and hamper the development of online applications.

Access bundles emerge

Even as free Internet access has begun to blossom in the UK, such moves have been trumped by last week's giveaway of the PC as well by US venture Free-PC.

Some 500,000 prospective users reportedly applied to join the Free-PC scheme within days of its launch on Feb 8th. Within the next few months, the 10,000 successful applicants will receive a brand new Compaq Presario PC complete with Internet access and email, completely free of charge.

The scheme is funded primarily out of revenues for advertising stored on the computer's hard disk and constantly displayed around the working area of the PC's screen. The ads will be targetted based on personal information supplied by the applicants in a detailed questionnaire that they must submit to qualify for the scheme.

Free-PC is the brainchild of Bill Gross, CEO of Pasadena CA-based idealab!, the investment and development operation behind 20 Internet startups including GoTo.com, eToys and Peoplelink. Free-PC also has a $10m investment from media group USA Network, recently in the news after agreeing a merger of its Ticketmaster Online-CitySearch subsidiary with Internet portal Lycos. Free Internet access and email is provided by another idealab! startup, NetZero, an advertising-funded, free-access ISP.

The Free-PC announcement stole limelight from a scheme launched in France two weeks before. Leading supermarket chain Géant, part of Paris-based Groupe Casino, on Jan 27th began selling 20,000 units of an Internet-ready PC priced at FF1,990, some $335. 5,500 were reportedly sold the first day, even though buyers also had to sign up for a two-year contract with Internet access provider Infonie, adding an extra FF149 per month, or $600 over the life of the contract. The PC was available without the Infonie contract for FF3,990 ($672).

Géant's scheme is firmly in the mould of the cellphone business model, in which the cost of the access device is subsidised out of the subscriber's monthly payments. Free-PC combines this with the advertising-based funding of commercial TV to create a proposition that requires no outlay at all by the consumer. Even in this model, of course, the consumer still pays, this time indirectly, out of the proceeds of goods and services purchased from advertisers.

Free-PC is only viable because it offers advertisers highly-targetted advertising to a pre-selected, captive audience, and it is unlikely this model could work across more than a limited segment of the population. But for its backers it has proved a highly effective means of gaining profile for the underlying concept of at least partly subsidising Internet access costs with advertising.

Free access economics

For the UK's free Internet access providers, compensating revenues have come from a more predictable source. Until the advent of Freeserve last October, European Internet users were at a disadvantage compared to their peers in the USA since they had to pay local call charges to the telephone company for their time online as well as paying a monthly fee to the ISP.

Freeserve, a venture by Hemel Hempstead, UK-based store chain Dixons Group, created a mirror image of the US setup by foregoing the monthly ISP fee and charging only for the phone call. The model works because of an artificial quirk in the regulation of the UK telecoms market. British Telecom (BT) has a monopoly on local call access in the UK, but must pay a termination fee to other telcos on certain calls. In the case of Freeserve, BT must hand over more than two-thirds of its call revenue to Energis, the telco whose ISP division, Planet Online, operates Freeserve on behalf of Dixons. Given sufficient economies of scale, that income alone is enough to fund the service.

Dixons has said Freeserve is on track to break even this year, despite significant extra infrastructure investments by Planet Online to keep pace with demand. Freeserve already has over 1m registered users, and a policy of discontinuing accounts after a month's inactivity keeps them fresh.

This month, rivals have acknowledged the success of the model by introducing similar schemes. Stores group Tesco was first off the block, with a free service operated by British Telecom. BT also dropped online charges on its pay-as-you-go service, BT Click. Another leading ISP, Virgin.Net (part of Richard Branson's Virgin Group) this week announced it will drop access charges from April 1st.

Although the British model is dependent on the regulatory regime remaining in force, it is a powerful demonstration of the benefits of bundling access and call charges into a single fee. Telcos should embrace this model and lobby for it to be replicated in other countries. It takes advantage of their existing billing systems and prepares the ground for usage-based billing when infrastructures move across to IP-based systems.

Enter applications

The huge demand unleashed for all three examples also demonstrates that, whether in the UK, France or the USA, a large body of users are held back from Internet access by a cost of entry that is perceived as too high. Anyone who removes that barrier is instantly rewarded by a massive surge in new customer sign-ups.

The opportunity this creates for application service providers is clear. In all of these examples, the free offer is really only an illusion created by substituting one form of expenditure for another. Free-PC will only be viable if users end up spending more with advertisers than they would have done in the absence of the service. Freeserve still collects a fee, but in a way that is transparent to the user. Géant merely spreads the payments across a two-year contract.

For an ASP, it is a simple matter to create a similar proposition around an application or an application bundle in which the Internet access, and perhaps even the access device itself, appear to be free. The model might even adapt what is already common practice among consumer ASPs, of providing a base level free service which users inevitably upgrade to take advantage of more sophisticated features as they find their feet with the package.

Existing free access providers may have cause to partner with ASPs and rental software vendors in any case to bolster their incomes. Free-PC, and Freeserve to a lesser extent, rely on advertising income for their profits, but that source of income is dependent both on wider economic conditions and on users actually seeing the ads. It will not be long before hackers work out how to switch off the ads on a Free-PC screen, just as they very early on discovered how to set up a Freeserve account without ever installing the setup software or visiting the Freeserve portal more than once. The advantage of offering applications as part of the service is that it provides a reason for users to keep coming back to the portal, thus sustaining the effectiveness of advertising.

In the business market, the higher value of applications will make it easier for ASPs to devise a costing model that bundles the access device. Thin client appliances such as Windows terminals lend themselves particularly well to a model with an entry-level price of zero. Here again, there is scope for increasing revenue by offering users extra applications and features which they simply switch on. Its success depends on making the applications both irresistible to the user and easy to bill.

The short-term appeal of such models for expanding the reach of the Internet and of application services is evident, particularly to those with a telco background. The concept of giving away a low-cost access device to eanble users to take advantage of services that are billed by the minute or on activation is one that recalls the golden age of the telco.

The long-term danger of this becoming the dominant model for Internet and application service providers is that it is a monopoly model in which users are tied into restrictive contracts by providers desperate to recover massive upfront investments.

In telecoms, users are still learning the hard way that they achieve the most cost-effective pricing by buying access devices and services separately in a competitive market. They should not be deluded by the short-term appeal of bundled access deals into giving up that freedom in the Internet market. Nor should the ASP industry encourage them into making that error. It may earn welcome profits in the short term, but the history of telecoms shows that the investment required to sustain bundled access has a dampening effect on creativity and innovation in the long-term. That would limit the ultimate prosperity of the industry as a whole.