Peer pressure

ISP peering issues could soon impact 
corporate e-business costs.

The issue of ISP peering is again raising its ugly head. And its reconsideration could soon affect how companies implement and pay for their Internet connectivity.

For the uninitiated, the term “peering” refers to the interconnection between Internet service providers (ISPs). Peering was a hot topic in the early days of Internet fever, because competing ISPs originally didn’t exhibit much enthusiasm when it came to interconnecting their networks. After all, who wants to help a competitor’s customers get connected?

Once it became clear, however, that everyone would win if the Internet provided universal connectivity, ISPs began to get with the program. They began to provision peering links, generally splitting the cost of strategically placed T-1 lines. As their Net businesses grew and end-to-end performance became increasingly critical to attract and retain paying customers, provisioning of peering links continued apace with little or no urging from corporate or consumer markets. Peering as a topic of debate therefore quickly passed from the scene.

ECONOMICS OF PEERING

A variety of circumstances are causing the industry to take a second look at the economics of peering. One of these circumstances is the increasing difficulty ISPs are having with making money. When everyone was gung-ho about an exploding digital economy, it was more important for ISPs to grab market share than it was for them to worry about near-term profitability. Today, on the other hand, the cooling sentiments of capital markets are making profitability the number one goal of telecom/data networking companies. They need to make money. Spending money on peering facilities just to help grow the market no longer makes good business sense. Peering links have to clearly and concretely pay for themselves.

Exacerbating this problem is the continually escalating volume of Internet traffic. Back when peering links only had to support a few million static Web-page hits a day, the total amount of money invested in peering infrastructure was relatively low. By contrast, the amount of traffic flowing over peering links today is staggering—often more than ISPs’ own backbones once handled. That’s a lot of money to spend with no clear economic return.

The problem isn’t just peering infrastructure. When an ISP receives traffic from another ISP, it may have to “backhaul” it to another ISP—adding to the total cost. That backhauling may even take place over an expensive transoceanic link, adding even more uncompensated costs into the mix. As the Net evolves to incorporate high quality-of-service traffic types, such as streaming multimedia and voice-over-IP, the infrastructure burdens of supporting peered traffic from competing ISPs will grow even more. Peering is thus becoming a serious business issue that ISPs must address if they’re going to achieve any kind of reasonable profitability.

YOU GET WHAT YOU PAY FOR

ISPs are likely to address their peering issues in three ways. First, they’re going to monitor their peering links more closely to ensure that their peering partners pay their fair share of peering infrastructure costs. Rather than simply splitting the cost for links 50-50 and hoping that things balance out in the long run, they’ll probably start tracking peered traffic to determine (a) what proportion of each monitored link is being used by each partner, and (b) whether the traffic flowing into their networks is also resulting in inordinate backhaul costs.

Second, they will begin to look at peering as a for-profit service. That means that they will proactively seek out peering relationships that will increase revenue without creating undue additional transport costs. Conversely, they will start to shy away from—or drive a harder bargain before implementing—peering relationships that don’t offer much profit potential.

Third, ISPs are going to start charging different rates for traffic that stays on their own networks from end to end and traffic that has to be peered. They’ll do this to make sure that they realize incremental revenue from customers to pay for their peering investments.

This third strategy is obviously the one that corporate communications managers should be concerned about. Several ISPs have already begun offering discounts for “on-net” traffic. This is simply a clever marketing spin to put on the fact that they’re charging a premium for “off-net” traffic.

While corporate customers may be miffed about this, they probably shouldn’t be. You get what you pay for, and good peering is worth money. It improves the performance of your websites and helps you implement better e-business connectivity to your diverse trading partners. Rather than taking an adversarial position, buyers should probably applaud more realistic Net pricing schemes and remember that bankrupt ISPs rarely provide very good service. E-business managers may even consider making changes in how they support their e-business relationships to bolster their on-net traffic in proportion to their off-net traffic. This, of course, is what ISPs have been angling for all along.

But the fact is that the heady days of something-for-nothing e-business are over. Financial markets are putting more pressure on everyone—from dot-coms and ISPs to mainstream corporations—to refocus on the bottom line. In the long run, that’s a good thing, because it will lead to a more stable e-business environment and more reliable networking services.

To comment on this month’s column, contact me at ll@exit109.com.