DSL-to-frame: 
an object lesson in industry economics

Complex market forces will determine if an idea that makes sense on paper makes it in the real world.

On paper, the concept seems like a good idea: use the economical, high-speed “last mile” access that DSL (digital subscriber line) provides to achieve higher Frame Relay throughput rates to remote sites. In fact, it may seem like such an inherently obvious wide-area-networking strategy that you may be surprised you didn’t think of it yourself—or that no one has already tried to sell it to you.

But the trip from the drawing board to the CO (central office) can be a long and treacherous one. Here’s why.

First of all, the folks who make DSLAMs (digital subscriber line access multiplexer) have to decide that they want to invest in developing Frame Relay interfaces. To do that, they need to know that there is enough of a market for DSL-to-Frame-Relay connectivity to deliver a reasonable return on their development and manufacturing costs. To help them make this determination, they’ll go to one or more market research firms. The research firm’s job is to provide reasonably accurate projections about a market that doesn’t presently exist. And you thought your job was tough.

Said research firm will look at the current number of Frame Relay ports and the current rate of growth in that number. A top quantitative telecom research firm like Vertical Systems will probably tell you that there are about 800,000 Frame Relay ports domestically and that port growth is slowing—although it’s still a healthy number, around 200,000.

Now, in order to calculate how many Frame Relay ports are candidates for DSL access, you have to distinguish between existing ports and new ones. New ports are better candidates for higher-speed access, since a buyer faced with paying $200 per month for a switched 56-kbps link or 416-kbps SDSL (symmetrical digital subscriber line) service is likely to choose the higher capacity, even if the immediate need isn’t present. 
A buyer with existing 56-kbps ports, on the other hand, will only make a change if there is a pressing need. Plus, in many cases, 56-kbps users have simply learned to live with the slower port speed. They have engineered their applications to work within those constraints and have a lot of other more pressing issues on their plates than replacement of their DSL CPE (customer premise equipment) gear.

E UNUM, PLURIBUS

Speaking of CPE equipment, manufacturers of DSL CPE hardware have to make the same decisions about market size vs. development/production costs. But their decision is further complicated by the fragmentation of the DSLAM market. In other words, they not only have to believe that the overall DSL-to-frame market is worth pursuing—they have to decide that it’s worthwhile to build the necessary integration for each individual DSLAM vendor’s DSL-to-frame architecture.

To make matters worse, CPE vendors have to compete for DSL-to-frame market share with multiple other DSL CPE players. In MBA-speak, the DSL-to-frame market is fragmented by the proprietary nature of DSLAM platforms. To use some rough numbers, there are two or three main Frame Relay providers, half a dozen leading DSLAM players, and over a dozen DSL CPE vendors. 

Do the math, and you will see that DSL CPE vendors will have to pay for their development, manufacturing, and marketing costs with a piece of a piece of a piece of an uncertain market. On top of all of this, the availability of DSL itself is limited by distance issues that may or may not be overcome over the next few years.

For DSL-to-frame to succeed in the real world, DSL equipment vendors also have to solve more than just transport issues. They have to deal with carriers’ management concerns as well. Carriers are accustomed to managing Frame Relay circuits on an end-to-end basis. How will they get the data they need from third-party DSL equipment? What software development costs will they be willing to bear to support multiple vendors’ hardware? Or would they rather just support one or two strategic partners and leave the other DSL players out in the cold? 

Of course, despite all these negative market factors, many of you reading this column undoubtedly find DSL-based access to Frame Relay very appealing. It may, in fact, offer an ideal solution for your projected networking requirements. The problem is that, unless you take some action on your own behalf, it will never see the light of day.

What can you do if you want to see DSL-to-frame become a reality? First, you can start banging on your Frame Relay sales reps. Loud, persistent customers can do a lot to offset clinically depressing market research. You can also use vehicles, such as user groups and online forums, to stimulate discussion of the issue. It wouldn’t hurt to send me an e-mail either. Vendors use the press all the time to push their market agendas. Corporate users have to do the same if they’re to get the solutions they need.

Do you want to see DSL-to-frame succeed? Or do you think it’s irrelevant to your long-term technology platform? Let us know—or forever hold your switched 56!