• Back to Knowledge Base

I admit I'm making these numbers up (because I'm not a nuclear physicist), but bear with me. Let's suppose that "critical mass" for plutonium is 10 kilograms. If I bring 10 kilograms of plutonium close enough together (but not too close), I can create 10 megawatts of power in a reactor. Bring 20 kilograms together, and I can create 20 MW of power. But bring 5 kilograms together, and I get nothing. Why? Because I can't create a reaction that sustains itself. There are not enough neutrons available to stimulate enough fission of more atoms to provide more neutrons.

Any time a website contains any feature that is driven by the viewers of the site, the same principles apply almost as reliably as the laws of physics. The most common example is a discussion or help forum. If there are not enough viewers who post content, there will not be enough viewers to reply to the content, since by nature only a fraction of the people viewing or using the site will have a relevant answer to my question. And if nobody has a relevant answer for my question, I (who also have an answer to someone else's question) will not use or view the site.

This may seem obvious but it really needs to be analyzed closely when we realize that the site owner's goal (and the goal of the Internet) is not to obtain "critical mass" but to make a profit.

If the business model of the site is based on banner advertisements or some similar sponsorship model, the issue is simply one of server load and storage, and computing (to process the searches and display the content). The more traffic the better, as these numbers boost the cost chargeable to sponsors.

However, if the business model is based upon subscription for the content of the site, there are now two fundamental principles working against each other. The subscription model is by definition a barrier that prevents traffic to your site; if you want to get subscriptions from those willing to pay, by definition you are excluding those who aren't. If you believe there is a critical mass to your site's content, the trick is to draw power (profit) from the reaction without stopping it.

It is my firm belief that sponsorship business models on the Internet only have a limited potential, and that for the quality of information that still requires human oversight, a subscription based model will ultimately be the answer. Desktop personal computers and software have really matured because the end user deemed it valuable enough to pay for it. This in turn is an incentive to programmers (whose capital investment is limited to some type of compiler-an oversimplification I know) to improve the quality of that computing even more. Currently we have the Internet model somewhat in reverse, where the programmers and providers pay for content creation, the end user receives it for free, and this is paid for by sponsorship. There's nothing wrong with the free model, I just simply don't believe it can sustain the quality of content that will drive enough traffic. In fact, I think you will find (low cost and usage-based) subscription become more common on the Internet. One of the areas that should take shape in this way is search engines, who may still use sponsorship but may also charge for the more technical information which is difficult to assign a price (but valuable to the user because of its difficulty to locate).

A practical example, a recent web project I worked on, is a State Association of Manufacturers that sought to maintain an interactive site that would enhance the value of its membership. The organization was already well established, but whether intended or not, the purpose of the site should be to provide a further incentive to join the organization, and to increase the perceived value of membership.

That brings up the first point-when your company designs a website, the first step to a successful website, whatever the model for the site, is to look at it as a business model. Even if the site does not generate cash of any sort, there must be some perceived material benefit to the company offering the site that outweighs its expense.

For this association, the business model was originally this: companies could become members and list their company information, along with key managers, presidents, CEO's and the like. In addition they could categorize themselves according to manufacturing process. There is no doubt that this type of database has intrinsic value. In addition, any manufacturer that was approved and a member would be able to post requests for manufacturing, which would be stored in the database also, and visible to members. When posts were made, all the members would be notified by email. In addition, upon reviewing the posts, any of the other members could reply to the post.

The "privilege" of posting a request for manufacturing was given to the members, and the "privilege" of replying was also given to the members.

I looked at the business model of this approach and felt it had flaws in it. In the simplest terms, manufacturers were posting a request that said, "I want to pay you (a manufacturer) money to do something for me." Although this is acceptable, it does not fall under a company's primary goal, which is to have someone else pay it to perform the service that it specializes in. For example, and aircraft plant may want to post a request for a new fire-proofing system, but its primary mission as a company is to answer requests for the manufacture of airplanes.

Under the model I proposed, only manufacturers could be listed in the database (paying members or not). This was necessary because the value of that data outweighed the value (and results) of charging every manufacturer to be listed. As a change to the model, anyone could make a post and could specify what category of manufacturing was needed (according to the NAICS coding standard). When posts were made, emails would be sent out to those companies with a PAID membership status, and who provided that specific service to manufacturers. For example, machining requests would go out to all machinists.

Since the association made money (already) from the dues of its members, we felt that fundamental business principles and interests of all were better served. Free membership by all qualified manufacturers was necessary to establish quality content, which would in turn drive traffic from the public (and other manufacturers and non-manufacturers) seeking those services. These factors had to first be in place before there would be an incentive for a company to be a paying member-allowing them to be instantly notified of requests that meant business to them. This, in turn, increased the perceived value of membership in the association. Although this is obviously not as complex as nuclear power, it is something that needs to be thought through to develop a successful site.

The final point is that, unlike nuclear reactors, you're not likely to see a "meltdown" or explosion in profits on the Net. There are limiting factors involved (quite a few in fact), and to get that power from the system you must maximize the efficiency of your operations to a surgical level. A few general concepts and I'll close:

1. Any website must ultimately be looked at as a business model.
2. If possible, sit down first and evaluate whether it is possible to create an end-user-pays model, or whether it will require a hybrid (end-user-pays-some + sponsorship) approach, or whether only a sponsorship model will work.
3. Identify any valuable resources you have based on the users themselves, and seek to maximize the value of this resource, and to minimize the cost of acquiring and maintaining it.
4. Use the highest quality means to present that data and maximize its usability.
5. Seek to create a site that answers fundamental business needs of all parties who will be using it.

I hope this helps your decisions and understandings about the development of your site, before you hire the database consultant or the web designer or graphic artist; and to quote Yogi Berra, "Nobody goes there any more, it's too crowded"-on the Web, at least, he was wrong.

--by Sam Fullman
  2/28/2002

TOP | HOME