Matthew James Gering [SMTP:email@example.com] writes:
At the network level, Content Providers bear the cost of providing content
on the Internet. They must find another way to recoup this cost, and/or
pass it on to the consumers of that content. This should *NOT*, and never
be a networking issue.
Why not? What else is the network there to do, but deliver our content? What are ISP subscribers paying $19.95 / month for, but to access our content? I doubt more than 5% of their connection time is spent inside an ISPs network, and less than 15% is consumed with electronic mail access. The remainder is "content consumer mode" -- and will become increasingly so with WebTV and other "passive" devices moving into the market. The net's not just nerds using IRC and Muds anymore.
As diagrammed (with Internet / CableTV models referenced)
The CableTV model does not apply at all to the Internet, it is entirely
different technology. CableTV you have finite bandwidth, which is
broadcast. CableCo must choose what content to carry, and users get all
that content. In this essence the CableCo is the content reseller. The
Internet is not broadcast, ISP's are not content resellers, they provide
network access, the user can choose the content, and can establish a
relationship directly with the content provider.
I think this misses the point, especially given the advent of push content providers like Pointcast. Internet /does/ have finite bandwidth, it /does/ support broadcast (push), and like it or not, ISPs will likely be forced into a role similar to that of cable tv provider in assuming subscription expenses.
What happens when a MTV, CNN or other content provider tells ISPs that they will need to pay $0.15 / mo. per subscriber to have domain access to mtv.com? Content providers of this nature are not interested in billing 10 million subscribers in micropayment increments. ISPs already have a relationship with these consumers, and are benefiting from the access to such content (i.e. what else are people doing online?). These players like the current model, as it works in the other content delivery medium. They're not geared up to issue 30 million invoices a month for $1.00, and they see ISPs as beneficiaries of their service.
Additionally, subscription models (from content provider to consumer) have underperformed. How likely are you to pay for your access to cnn.com or pointcast.com? It may be worth $1.00, but are you going to put that on your Visa every month? Micropayment software presently is too much work. Online advertising revenues are also not up to expectation, given the markets wariness of the quality of the medium. Bulk subscription charges passed to ISPs are an inviting method given their existing billing relationship and subsequent benefit from the content provider's labor.
Furthermore, expect sweetheart deals between content providers and NSPs. MCI can deliver X million heads and throw in some promotional perks to content providers. They'll play ball with content providers in this model as it adds barriers to entry to the smaller ISP. Peering problems are an indication of rougher water ahead; not only will small/mid sized customers be blocked at a routing level (unless they pay transit to a NSP), they'll be blocked at an application level and forced to pay higher subscription
What's the solution? Unless small to mid-sized ISPs leverage their networks through some sort of affiliation (something the CIX should have facilitated, but was unable to do so), the future is bleak. Litigation and regulatory intervention take too long and favor deeper pockets -- a game ISPs will likely loose at. An ISP coalition, "ISP-ORG" could assemble hundreds of thousands of subscribers, represent many AS's that have to be contended with at peer points, and provide a subscription base to be dealt with. It could develop a multilateral peering contract, represent the organization in obtaining reduced rates where transit must be bought, and negotiate with content providers for discounted service (potentially free, given the promotional ability of the large base).