Will Higher Bandwidth and Data Consumption be Matched by Revenue?
Internet access prices always fall, over time, history suggests, just as the cost of computing or storage falls, over time. Since 1998, to note just one example, internet transit prices have fallen by four orders of magnitude. In the retail internet access business, it also is possible to predict that prices are poised to fall further. That has been true, globally, since at least 2010, according to International Telecommunications Union figures.
The issue: what are the implications for the service provider business model. In the voice business, lower prices lead to higher usage, so lower price-per-unit was offset, to a large extent, by higher consumption.
That also is true of internet access services. As prices fall, consumption increases. But there arguably is a difference. Supplying voice services required some tweaks and some investment, but not wholesale upgrading of the networks on a continual basis. And that is precisely what internet access historically has required.
So the fundamental problem is not that unit prices fall, or that demand fails to increase, but that the supply increases faster than revenue at the new levels of consumption, while investment likewise has to grow substantially to support the new levels of consumption.
In simple terms, usage grows faster than revenue, while investment requirements also remain high. That is a crucial problem for legacy service providers with legacy cost structures, less a problem for new providers without such cost burdens.
Even Google Fiber seems not to have had as much success as expected. Cable TV operators have fared far better.
Virtually all studies of U.S. residential internet access services show declining “cost per megabit per second” trends, nearly matching the pricing and performance trends one would expect from any business driven by Moore’s Law.
Such radical revenue-per-unit declines arguably must be accompanied by other moves to reduce the cost of supplying bandwidth.