Are Telcos Losing Revenue Because of Low-Quality Voice?

Telcos are losing some voice revenue because existing voice quality is poor, says Andrew Odlyzko, University of Minnesota Digital Technology Center director. Despite growing volumes, “toll quality voice is lousy quality, especially for women’s voices," he says. Likewise, "wireless voice is horrible; it has been selected to be the lowest quality one can have that people wouldn’t reject.”

There is room for higher-quality voice as a premium service, Odlyzko maintains. As others have argued, Skype provides an example of quality "better than toll." 

The problem is that many examples of higher quality voice are hard to separate from other variables that also might illustrate what the specific value proposition is. Skype, with good microphones and on a good broadband connection, can sound better than toll. But such connections also provide other sources of value, such as the ability to talk internationally for "no incremental cost."

What cannot easily be tested is the thesis that, for this particular application, even lower quality would not reduce usage, or destroy the value of the application. In fact, if voice volumes, using Skype-style mechanisms as well as mobile and fixed line long distance, continue to grow, it is hard to argue clearly that there is an unmet consumer need. There might be. There should be. But so far, no clear test has been devised.  

“There are opportunities for higher quality voice,” he says. One suspects he is right, though it is not so clear, beyond business videoconferencing, where one can point to examples. 

Odlyzko is an original thinker, so some of his views run counter to conventional wisdom. He argues, for example, that wireless providers have erred in focusing on mobile data, where they might have used their 3G networks to provide better voice. 

I'm not sure I agree with that conclusion. And, perhaps, despite his views, neither does Odlyzko. He pointed to at least some forms of wireless data as services that provide higher value and margins. That clearly is the case for text messaging. It appears to the case for mobile Web access. 

So far, the service provider effort has been to provide IP-based equivalents of time division multiplex voice, with equivalent stability and quality, not greater quality. It might be interesting to see whether such a clean test could be market tested. 

Not many service provider executives appear to believe there is a significant market for higher-than-toll-quality voice in the consumer market, though. Users might well like it, but service providers likely are skeptical about whether consumers would pay a premium. Business users might be a different case, but not many clear examples aside from room conferencing systems can be noted. 

Service providers might be losing some revenue because of poor voice quality. They also are losing some customers for other reasons, such as a secular decline in demand for landline voice. It is unknown whether a revenue opportunity exists in the mass market for higher-quality voice. There should be an opportunity in business markets. But the thesis largely remains untested. 

That isn't to say some providers have failed to offer better-performing codecs, for example. That should provider higher-quality voice. The issue is whether consumers are willing to pay a premium for it. They might be. The problem is arranging a clear market test where some other obvious variable does not cloud the test.