Paying for Internet Goods and Services

By Peeter Deutsch
Bunyip Information Systems, Inc.

There's been some talk lately on such lists as com-priv concerning how we're going to pay for all those spiffy new Internet services that everyone now agrees are on the way.

For those who don't know me, I am Peter Deutsch, one of the developers of archie and now president of a startup company developing and deploying Internet goods and services. Finding a way to pay for such services is certainly a subject dear to my heart right now, since I've grown so fond of eating, and my kids so like sleeping under a real roof.

How to pay for Internet services?

Actually, there are a number of funding models straight out of economics 101 that are being examined by those of us trying to make a living developing and deploying new Internet tools and services. These include:

I'm sure there are plenty more, but this is enough to get us started.

User pays

User pays is simple, although it can have a number of variations. For example, you can charge per access or charge a subscription or flat fee, or some variation of these. The important point is that this model is probably most appropriate where the users want the information and nobody else cares if they get it (e.g., Usenet reading service, dial-in accounts, etc).

There is also room for the public radio or public television model here, with users sending donations in to their favorite tool or service providers. To my knowledge, this particular angle has been singularly non-existent on the Internet. We've received only one or two offers of support over the past two years, and not a single penny in donations. I think this contributed to McGill's [McGill University in Montreal] perception that what we were doing was not as popular as we claimed it was, since they never saw any prospect of the users helping to defray its costs. I had thought of pursuing donations from end users, trying for over a year to solicit hardware donations, but it just didn't seem like an idea whose time had come to the net yet.

Service provider pays

Service provider pays could be used by Internet services providers to provide value-added services ensuring market differentiation. When I first tried to find outside support for archie, I started with this model. When seeking support, I argued that service providers operating an archie server distinguish themselves on the network, making a considerable name for themselves and could thus fund it out of the advertising budget. We also have cases where service providers (e.g., Australia, Finland) decided that archie would cut down on their International link traffic and so made economic sense to operate.

Despite the number of sites running archie, I was less successful in this approach than I thought I would be, although in the past few months it seems that the attitude toward services is starting to change. Still, in my experience if you take this approach you're going to need substantial financial arguments. From our experience, saying that the service is useful and should be supported on philanthropic grounds is not enough in most cases.

Information provider pays

Information provider pays seems to hold a lot of promise as the Internet grows. There are people out there who want you to know about their goods or services and will pay someone to get that message to you. The trick is putting together a complete service, which is a lot more work than it appears. Still, I think there is a lot to be done in this area in the future.

Government pays

Government pays has obvious problems in an international Internet environment, although there is the soon to be announced NIS solicitation to show how some services could be funded through taxes. Although I welcome such support for some basic infrastructure, I personally would prefer to see a free market of goods and services since I have a great deal of faith in the power of the market to allocate resources where they are most wanted or needed.

There is a role for governments to fund research and startup services in a new market such as the Internet. The trick is not to get too dependent upon such support, since with money goes power and influence.

Nobody pays

Finally, there's our old friend "Nobody Pays," which appears to have served so well to date. I happen to think this model has serious problems, and caution anyone who thinks the Internet is or should be free to ask yourself what we lose with the "all-volunteer" approach. Certainly it has slowed down the future development of archie and its follow-ons, since those of us working on it found ourselves having to choose between our jobs and our work on the Internet. I think it has a role in deploying interesting projects but is singularly unsuccessful in deploying new services. The distinction is an important one, since it implies a level of support and quality that we currently just don't see on the Internet in most cases.

In our case [Bunyip Information Systems], we've chosen to try the capitalist road, since we think this will be healthier for both archie and ourselves in the long run, but we did have to choose. For those who argue that we have an example of volunteer services I need only point out that McGill is no longer even operating an archie and virtually all of those operating servers have agreed to pay us to provide support for future releases. I hope this implies that the Internet community is starting to realize that there really is no such thing as a free lunch, or a free service.

Although there is a slim chance that McGill may again run an archie, the fate of this particular would-be service here in Canada illustrates once again what happens when a volunteer project becomes too successful. You really do give up something when you don't pay.

While we don't plan or want to charge the end user for our current service, the archie developers have definitely "gone commercial" and are now charging for access and support for the archie system software. We also plan a range of future services based upon this technology and at least some of these will be using various combinations of the funding models I described above. Of course, we predate WAIS, Gopher and WWW on the net and have some claim to being an "innovative" service.

For what it's worth, I understand that Brewster Kahle, developer of WAIS has also formed a company. I have also heard rumours from some of the other people involved in the various projects of potential companies, as well. (Editor's note: There is now a company called WAIS, Inc.)

Yes, there are a lot of freebie projects out there, and a lot of good volunteer work has been done, but I think if we are all to make the transition from interesting projects to interesting services we're going to have to start seeing money change hands. I don't find this as offensive as some people, but those who believe that it can be done for free are encouraged to do so. We need all the services we can get and if someone is feeling philanthropic then go for it. Those of us with mortgages will be taking a more pragmatic approach.

Peeter Deutsch
President, Bunyip Information Systems, Inc.

Taken from The Link Letter, Vol. 6 No. 1, April 1993.