Showing posts with label Gizmo. Show all posts
Showing posts with label Gizmo. Show all posts

Tuesday, January 1, 2008

Thriving Even When a Market "Vanishes"

Every trusim must be qualified. Consider the gradual folding of the "long distance" calling business into a "triple play" or "mobile calling" bundle. One might correctly note that long distance increasingly is a feature of some other product that a customer buys.

Some of us have said over the years that long distance has ceased to be a stand-alone business, pointing to the physical disappearance of firms such as at&t and MCI, which were built around long distance calling. But that general statement must be qualified.

It might be true at a high level that long distance has changed, and cannot support a firm such as the old AT&T. But that doesn't mean every stand-alone long distance business disappears. Calling card revenue still seems to be growing, for example.

And one has to point to Skype, Gizmo, Jajah, Rebtel and others as examples of companies making a living on long distance calling.

In the same way, some observers have argued that VoIP increasingly is becoming something sold as part of a triple play bundle, or as the technology underpinning for analog voice. The salient example is Vonage's stand-alone VoIP service contrasted with cable operator digital voice.

Again though, one must say the general rule does not rule out the continued ability of some entities to grow their "stand alone" businesses, even in the face of the existence of the larger trend.

At the same time, voice calling and voice features and services are emerging as an attribute of experiences for which there is some alternative revenue or business model.

Stand alone VoIP will be difficult, in many cases. It will not be impossible or unprofitable. The cost structure of such businesses will have to be optimized. But even the pedestrian calling card has continued to offer some firms an opportunity even as ubiquitous mobile phone usage has become a dominant trend.

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...