The Internet of Things, in all its forms–has a total potential economic impact of $3.9 trillion to $11.1 trillion a year by 2025, according to an analysis by McKinsey Global Institute. The biggest segments might well be industrial automation, smart cities and health apps.

At the top end, that level of value—including the consumer surplus—would be equivalent to about 11 percent of the world economy.

With the caveat that today’s access providers likely always will earn a majority of their revenue providing access services of various types (mobile, fixed, Internet access, Wi-Fi hotspot access,

Ethernet and other high-bandwidth connections), the key to future revenue is likely going to come from providing managed apps and services, as always has been the case.

That statement should be unremarkable, except for the fundamental change in business model since the advent of the Internet and IP communications.

In the past, one might have argued that most revenue came from selling apps (voice and messaging), and relatively less from “data access” connections (dumb pipe T-1, DS3 and so forth).

Even data access services such as frame relay, MPLS and ATM essentially were managed services, or “apps.”

These days, as apps are structurally separated from access, legacy communications providers are earning less revenue from fully-owned apps and lots more from data access (dumb pipe).

So the issue is how the revenue model might evolve over the next decade or two. Broadly speaking, it is clear that dumb pipe access and apps now are separate parts of the revenue stream. So the big question is how access providers can create viable and useful roles for themselves in the app part of the business.

By some estimates, there already are between six and 14 billion autonomous connected devices or “Things” that are connected via some form of communication mechanism, not including smartphones, tablets, computers and similar consumer devices. By way of comparison, there might be seven billion connected consumer devices in use today, including phones, PCs, tablets, TVs, game players and so forth.

As you might guess, the further out one goes, the more current estimates diverge. The number of connected devices by 2020 ranges between 18 billion and 50 billion devices. Some of us would argue the issue is not the number of devices in use, but the timeframe.

We will probably see less deployment up to 2020, then much more development after 2030.