I look to the UK for guidance, the cable industry over there pulled twisted copper in the same cladding as the coax to offer apples to apples local phone service. Now the British public's distaste for BT is legendary and the cable competitors have taken about 40% share of local service in subscriber households, while offering ~15% discount.
US telcos I have talked to on this issue expect to keep the lion's share of their telephone business to customers who subscribe to cable modem service. Apparently they believe the target customer for cable modem is the type of user who will continue to maintain multiple POTS lines for other purposes. While this is in keeping with the UK experience, I think it is a little wishful thinking, as it would be ATT branded service. So I would expect T to get ~60% of its cable modem subscribers to take phone service over time.
The greater question is how many US households will adopt cable modems. About 35% of US households have PCs, growing 4-5% a year. Most are not particularly sophisticated users. Many will not have two way capable cable systems and/or do not subscribe to cable. T will have to resolve the shortage of capable installation personnel that plagues the cable industry. 5 year forecasts of cable modem subscribers range from 5 to 20 million, or 4-16% of households. If you pick a mid-point, thats about 10%, multiply by the 60% telephony take rate, you get 6% roughly.
I project second line POTS growth (about 12 million in 1998 and still accelerating) will outstrip the growth of replacement alternatives for 4-5 years (including wireless replacement of fixed line). Add in line growth for business and natural population increases - we'll see a slow deceleration in POTS line growth rates starting in 2000 in US IMHO.
The challenge for LU will be to get their share of the alternative access market, and take share overseas where 2nd line growth (and 1st line growth) will be eyepopping. Just a note: in Europe, where they still have variable charges for local calling, ISPs will have one dial up port for every 25 subscribers (more like 10 to 1 or less in US). These charges are coming down and will likely go away - you remember what happened when AOL went to flat rate pricing? Great for RAS, Great for internet backbone eqt. (CSCO) Also great for POTS line growth. |
