Edge Computing, Private Networks, APIs Won't Drive Net Revenue Growth for Telcos by 2030

Respondents from eight telcos believe their revenue prospects are highest in such new areas as edge computing, private networks and wholesale access, such as allowing third parties access to network features. 

While logical beliefs, and while revenue might be a positive number in all these cases, service providers, on the whole, are likely to find disappointment in these sources as a way of moving the revenue needle. 

Put simply, there is not enough revenue in those areas to offset the huge reliance on traditional subscriptions. 

source: STL Partners

Using estimated revenue from a variety of sources, including IDC, International Telecommunication Union, GSMA and ABI Research, for example, global service revenue might grow an additional $250 billion from 2023 to 2030. 

Revenue from new sources including edge computing, private networks and other horizontal services might add about $300 billion. But revenue from the mainstay subscription revenues might decline $250 billion as well. 

Revenue Source	2023 ($ billions)	2030 ($ billions)
Mobile subscription revenue	1,200	1,400
Fixed network subscription revenue	300	250
New revenue sources	100	400

 In other words, new revenue sources will be welcome, but might only offset declines in the legacy drivers of revenue. 

Revenue Source	2023	2030
Mobile subscription revenue	$1.5 trillion	$1.2 trillion
Fixed network subscription revenue	$0.5 trillion	$0.4 trillion
Edge computing	$0.3 billion	$3.2 billion
Private networks	$0.2 billion	$2.72 billion
IoT connections	$0.1 billion	$1.42 billion
Total

Of course, if one were to compile a list of forecasts for new revenue put together by suppliers, one would see a different set of numbers. New sources such as private networks might add $2.5 trillion in new revenue, for example, by 2030. Does that seem credible for an industry that only grows about two percent a year, and generates about $2 trillion a year in annual revenues?

Likewise, edge computing is thought by some proponents to represent $2.9 trillion in new revenue, in an industry generating only about $2 trillion in total revenue in 2023. 

Revenue Source $ Trillion	2023	2030
Mobile subscription revenue	1.50	1.20
Fixed network subscription revenue	0.50	0.40
Edge computing	0.30	3.20
Private networks	0.20	2.70
IoT connections	0.10	1.40
Total	2.60	8.90

The point is that revenue growth of such magnitudes represents a cumulative annual growth rate in excess of 23 percent. That is an order of magnitude higher than analysts expect the global service provider business to grow, overall. 

The only real issue is how far short the predictions of new revenue from edge computing, IoT, private networks and all other sources will fall short of the aggressive predictions. 

Even if "Fiber is Always the Answer," It Does Not Change Buyer Behavior

Incumbent internet service providers claim they need financial support from a few major hyperscale app providers as they arguably cannot sustain their home broadband networks without such support. 

Governments, on the other hand, have pursued all sorts of policies to ensure that home broadband service is affordable for citizens, resulting in “low prices” for home broadband, in many countries. But low prices are a disincentive for investment, and governments want that as well. 

So some of those governments now seem receptive to the idea that perhaps they have pushed for “low prices” at the expense of sustainability of their ISP supplier bases and incentives for investment. 

Which is a bit of a seemingly-enduring paradox. To a large extent, incentives for investment, which require relatively higher prices, but lead to faster speeds, clash with the desire for low retail costs, which are a clear disincentive to invest and tend to result in lower speeds.

But even a shift to “all fiber” access does not necessarily solve the “low price, high speed” dichotomy, though that can be the case. 

Even where fiber-to-home networks are widespread, and considered the “best” home broadband platform, they sometimes face facilities-based competition from hybrid fiber coax networks, for example. 

In many such cases, only 40 percent to perhaps 50 percent of customers buy those services from the FTTH provider. In markets where there is no significant facilities-based competition, and there are “best, better, good” service plans, we see the same pattern.

Country	Network	Best Plan Price	Best Plan Speed	Good Plan Price	Good Plan Speed	Better Plan Price	Better Plan Speed	Take Rate for Best Plan	Take Rate for Good Plan	Take Rate for Better Plan
United Kingdom	Openreach	$50/month	1000 Mbps	$35/month	500 Mbps	$45/month	750 Mbps	20%	60%	20%
New Zealand	Chorus	$75/month	1000 Mbps	$50/month	500 Mbps	$60/month	750 Mbps	15%	65%	20%
Singapore	Singtel	$80/month	1000 Mbps	$50/month	500 Mbps	$65/month	750 Mbps	10%	70%	20%
Australia	Telstra	$100/month	1000 Mbps	$60/month	500 Mbps	$75/month	750 Mbps	15%	60%	25%

As with any other consumer product, not every customer buys the “premium” product version. So is “fiber always the answer?” Yes, in a long-term sense, for fixed networks, as a physical media choice. 

But even when a single FTTH wholesale network operates, customers still seem to choose “good enough” service plans, not the “best” and not the “value” tier, either.

Customers Buy What They Want, When They Want It

Some “debates” never seem to go away. Some decades ago, an engineering vice president quipped that “fiber is the future,”  followed after a short pause by “and always will be.” 

The humorous point was that customers make concrete choices under concrete circumstances that often defy our notions of “what is better” or “what they should buy.” 

The success of 5G fixed wireless and cable hybrid fiber coax alternatives to digital subscriber line probably comes as no surprise to anybody. But the actual choices consumers make, when fiber to home, hybrid fiber coax and other alternatives are available might confound technologists who believe “fiber is always the answer.”

In fact, consumers make choices that suggest that is not automatically true. Most telcos (perhaps all) have found that after marketing for several years against hybrid fiber coax, they tend to get about 40 percent adoption (four homes out of 10 will buy). 

So something other than “performance” is at work.  Rather obviously, a substantial number of customers believe the product version that makes most sense is a “good enough” product for a “reasonable price” For perhaps half the market, that means whatever commercial service offering speeds “in the middle” and also “prices in the middle” of all offers, low to high. 

Data from OpenVault consistently suggests this is the case. 

source: OpenVault 

Physical media matters, since not every platform has the ability to keep scaling bandwidth into scores-of--gigabit-per-second ranges, either symmetrical or not. 

But consumer demand also matters, as even when the difference between the fastest tier and the slowest tier is two orders of magnitude, most customers will likely keep buying speeds in the middle, about one order of magnitude below the “fastest” tier of service, and one order of magnitude above the lowest tier of service. 

Internet service providers often need to offer speed that is “good enough,” at prices that also are viewed as reasonable, to remain viable for half the market. But it might always be the case that up to 25 percent of the market will buy even a “slower speed” service if the pricing is right.

Even as speed requirements continue to grow, demand for “slower but affordable” and
“Good enough at a reasonable price” should both remain viable segments of the home broadband market.

Is Internet Really Carbon Intensive?

Mobile and fixed networks have made important strides to reduce carbon impact, leaving computing activities, a study by Telefonica suggests. As always, methodology matters. Telefonica says it has included all embedded carbon footprint into its analysis. “All the life cycle inventory for materials and processes have been taken from the Ecoinvent 3.8 cut-off version database (published on 2021), says Telefonica. 

Ecoinvent is one of the most well-known LCA database worldwide used in more than 40 countries and is said to include both direct and indirect emissions, which means that it accounts for the emissions associated with the production of the materials used in a product, as well as the emissions associated with the transportation and use of the product.

source: Telefonica

Other studies suggest, in fact, that networks themselves have gotten so much more efficient that carbon emissions increasingly are dominated by end user equipment and behavior. 

Source	Footprint (gCO₂/GB)	Source of estimate
5G network	0.1-0.2	IDATE DigiWorld: https://en.idate.org/content/uploads//2022/02/White-Paper_Fiber-for-a-sustainable-future.pdf
Fiber to the home (FTTH) network	0.05-0.1	IDATE DigiWorld: https://en.idate.org/content/uploads//2022/02/White-Paper_Fiber-for-a-sustainable-future.pdf
Copper access network	0.2-0.4	IDATE DigiWorld: https://en.idate.org/content/uploads//2022/02/White-Paper_Fiber-for-a-sustainable-future.pdf
Customer gear (routers, modems, etc.)	0.01-0.02	IDATE DigiWorld: https://en.idate.org/content/uploads//2022/02/White-Paper_Fiber-for-a-sustainable-future.pdf
Computing activities (servers, data centers)	0.1-0.2	IDATE DigiWorld: https://en.idate.org/content/uploads//2022/02/White-Paper_Fiber-for-a-sustainable-future.pdf
End user equipment (computers, TVs, game players, etc.)	0.6-0.8	IDATE DigiWorld: https://en.idate.org/content/uploads//2022/02/White-Paper_Fiber-for-a-sustainable-future.pdf

It is easy to capture footprint from hyperscale data centers, which might suggest that part of the value chain contributes “the most” to carbon footprint. 

Many estimate that use of the internet represents about 3.7 percent of global carbon footprint. Compare that to the impact of a few other industries.

Industry	Percentage of global CO₂ emissions
Energy	25%
Agriculture	24%
Industry	21%
Transportation	14%
Buildings	6%
Other	10%

As always, we need to keep degree of improvement versus cost in mind (cost versus benefit). As much as 70 percent of total global carbon emissions come directly from producing energy, agriculture and industrial processes. Add transportation and buildings and you have 90 percent of emissions footprint.