Will 5G Be a General Purpose Technology?

Some believe that 5G could become a general purpose technology (GPT), as was the printing press, internal combustion engine or electricity. The implication is that 5G might then prove to be a foundational input for many other innovations, industries and use cases.

So one might think that vastly increases mobile industry (and ecosystem) revenue potential. Maybe not.

Other generally-accepted general purpose technologies have arguably created lots of growth elsewhere in the economy. That is part of the definition of a GPT: it is an innovation that spurs other innovations elsewhere.

But that might not mean that the industry around the GPT itself necessarily reaps most of that growth. Electricity always is considered a GPT. The U.S. electrical industry generated perhaps $366 billion in revenue in 2016. Those figures do not include the value of output of suppliers to the electricity companies, which would boost total revenue a bit.

Total U.S. economic output in 2016 was about $18.6 trillion. So the electricity business was less than two percent of total GDP, as important as electricity is to nearly every industry.

In 2016 U.S. service provider revenues were in the $350 billion range (including video subscription services). Adding in the value of industry hardware, software, cloud services, advertising and so forth might boost total revenue into the $720 billion range. If you wanted to add in information technology services, consulting and integration services, one could reach a total of perhaps $1.2 trillion in annual revenue.

But that arguably is stretching matters. Core telecom service provider revenue (in the $350 billion range) is about the same size industry as is the electrical industry.

In other words, telecom arguably already earns about as much revenue as electricity providers do, and nobody would disagree that electricity is a GPT.

The point is, how much more revenue could telecom, and especially 5G, earn itself, were 5G to emerge as a GPT? Yes, GPTs spawn huge amounts of other economic activity. But that does not necessarily mean that the GPT itself amounts to more than a low single digits percentage of the overall economy.

In 2016, the U.S. automobile industry had perhaps $280 billion in annual revenue. If one takes auto sales as a proxy for the internal combustion engine GPT, that again is less than two percent of total GDP.

So the point is that, even if 5G does emerge as a GPT (and that is perhaps questionable), most of the revenue creation for the connectivity industry itself might be confined to two percent of total GDP (about the same percentage of GDP as the whole telecom industry already represents).

On the other hand, additional revenue growth could be captured by firms supplying the GPT if they invest in the “up the stack” industries using the GPT as an input. That already has been the case as connectivity providers have moved into the video subscription business and then the content creation business. And a shift into advertising also is happening. Something of that sort also has happened with mobile financial services.

So perhaps we not get too excited about whether 5G is a potential GPT. Whether it winds up that, or not, probably does not change strategic options for telcos and mobile operators.