New York Dead Zones Are Just a Fact of Life

For the 20 million people living in the greater New York area, spotty cellular service is a constant source of frustration. To document the extent of the problem, The Wall Street Journal examined data on dropped and unsuccessful calls compiled by the Nielsen Co., which sends out equipment-filled vans to make 140,000 test calls a year across the five boroughs, Long Island, northern New Jersey and southern Connecticut.

The results raise a pressing question: Why is the most populous metropolitan area in the U.S. plagued with dead zones?

Carriers deserve only part of the blame, the Wall Street Journal says.  The city is one of their highest-profile markets, and even as they cut capital spending nationwide during the recession, they kept budgets high in New York, even if that spending didn't always keep up with the pace of smartphone sales.

Ask the people who build and maintain the city's networks, and they'll give you a dozen alternative excuses: Too many people. Too many buildings. Too much reflective glass. Too much water. Each plays a role, the Wall Street Journal says.

It all adds up to wireless dead zones dotting the city and its surrounding suburbs.

Take the West Side Highway, a notorious trouble spot where Nielsen recorded eight failed calls up and down Manhattan. There, phones get a signal only on one side of the road, from cell towers high atop office buildings. (The Hudson River is on the other side.) So when there's a hiccup with a connection to the cell tower on the Manhattan side of the river, there's no other tower to back it up, and the call drops.

Dropped calls also happen because of quirks in the way carriers have set up their networks. For example, AT&T Inc. routes calls south of 59th Street in Manhattan to a switch downtown. North of 59th, calls go to a facility in Westchester. So when an AT&T customer crosses 59th, calls can get dropped as the network reshuffles from one switch to the other. Nielsen recorded three fails on or near that dividing line. AT&T declined to discuss coverage at 59th in further detail.

Sometimes, there just aren't enough cell sites to handle the load.  But it always is difficult to get authority to put up new towers, anyplace. In New York, it is harder because of the fragmented ownership of buildings. Outside New York, it often is possible to negotiate once with an owner of hundreds of buildings.

The upshot? Service issues are likely to continue to be an irritant.

Microsoft OCS Starts to Disrupt

Microsoft's Office Communications Server 2007 is going to disrupt market share in today's business phone system market. It also is going to take share and rearrange markets in other areas you might not expect, such as the test and measurement space. Huh? Isn't voice quality testing, on both qualitative (subjective) and quantitative (mean opinion scores, for example) dimensions, something that specialized test and measurement firms do? Well, yes.

But Microsoft also is bringing to market its own "quality of experience" server that automatically tracks end user voice quality no matter where a call is placed--from inside the enteprise or at a hotel. No network test probes are required.

That's just one more example of how incumbents are finding their core businesses under threat of rearrangement from upstarts with deep pockets, strategic motivation, different visions and deep expertise in software and networking.

Voice Quality is Getting Worse: What Would You Expect?

Those of use who grew up with one phone company got spoiled by the reliability and quality of its communications network (despite "customer service" so bad it became an oxymoron)," says technology journalist Mark Stephens, whose pen name is Robert X. Cringely. "Those of us trying to save a few bucks by piggy-backing voice services on the Internet are starting to get what we've paid for."

Skype itself now is experiencing an outage that might take 12 to 24 hours to fix (Aug. 16).

There's a larger trend at work here, and it happens in virtually all formerly highly-regulated businesses when deregulation and new technology hit. Remember when airlines were highly regulated, and could not compete on the basis of price? How did they compete? Amenities and other non-price differentiators. Of course, prices were high and not that many people flew.

Deregulation hits and all of a sudden price becomes a key competitive weapon. Of course, when people start paying lots less, something has to give. Like amenities. But more people fly now.

So here's the problem communications service provider executives face: they can't afford to run "gold plated networks" for the same reason airlines cannot. Obsessive concern about voice quality and service availability are one thing in a highly regulated environment. Such concern is quite something else in a highly competitive marketplace where customers in fact choose to pay money for service that is quite a bit less intensive than it once was.

In a nutshell, the business problem is that operators cannot afford to maintain the same obsessive levels of quality when customers demonstrably don't care. Mobile communications is the best example. Everybody uses mobile service. And everybody knows it simply is not as reliable as wired phone service. Nor is the audio quality as good. But it's a wild success, anyway.

If people will not pay you to maintain a higher quality of service, can you afford to do it? That's the problem the global communications business faces. People are voting with their pocketbooks: buying services with lesser quality on some metrics because the overal utility of mobility is so high.

In other cases, such as over the top VoIP, they are voting with their wallets to buy cheaper services with less reliable service.

Get used to it. In virtually every deregulated, formerly monopolistic industry, overall quality will drop. Of course, there's another trend as well. New, higher cost alternatives will develop. Because some people need high quality enough to pay for it.

IP communications are very valuable. They are very useful. But they are not as robust as the old public switched network, if only because of things like latency. The services can be made more rugged, of course. It just costs money.