Thursday, July 25, 2024

AI--Like the Internet Before It--Will Cause Net Job Loss

Perhaps you have not lived through a disruption such as digital media, social media, the internet, targeted versus linear advertising or generative AI before. But we call it a “disruption” for a reason: revenue models; profit margins; growth trajectories and jobs all are affected. 


And it is unreasonable to expect that GenAI--a tool for creating content--will not disrupt business models, revenue streams, profit margins and jobs, like it or not. The internet, for example, disrupted advertising-driven industries. 


The shift to digital platforms dramatically reshaped the value of many advertising venues that had used a linear format. Compared to legacy media, online media was able to feature targeting based on actual end user behavior. 


Based on that understanding, content could be personalized, raising the possibility of engagement and better matching expected purchase intentions. 


Digital advertising allowed for more accurate measurement of campaign performance and return on investment. 


The ability to reach potential audiences--especially for specialty products--increased as a result of global reach.


Real-time adjustments: Unlike traditional media, internet advertising allowed for real-time campaign adjustments based on performance data.


Year

Total Ad Spend (Billions USD)

Digital Ad Spend (%)

Legacy Media Ad Spend (%)

2000

~$250

~5%

~95%

2010

~$400

~20%

~80%

2020

~$600

~50%

~50%

2024

$667.6

68.9%

31.1%

2027

$870.9 (forecast)

73.8% (forecast)

26.2% (forecast)


Generative AI, on the other hand, is more likely to disrupt the production of content rather than its monetization. Since 2000, we have seen mostly net negative changes in jobs in various industries, based on whether they are emerging (new fields such as AI or e-commerce, for example) or declining (replacement by internet mechanisms). 


Year

Job Sector

Jobs Gained (Millions)

Jobs Lost (Millions)

Net Change (Millions)

2000

E-commerce & Online Retail

0.1

-

0.1

2008

Manufacturing & Production

-

0.2

-0.7

2009

-

-0.8

0.1

-0.9

2012

Sharing Economy (e.g. Airbnb, Uber)

0.6

-

0.6

2016

Artificial Intelligence & Machine Learning

0.2

-

0.2

2020

E-commerce & Online Retail

2.8

0.8

-2.0

2021

Remote Work & Freelance Platforms

1.9

0.9

-1


Most industries have seen net job losses in most years since 2000, it might be worth noting. It might not be unreasonable to suggest AI in general will have a similar impact, over time. Though other value might be obtained, productivity improvements are likely going to have negative pressure on jobs. 


Year

Job Sector

Jobs Gained (Millions)

Jobs Lost (Millions)

2000

E-commerce, IT & Tech Support

0.1

Negligible

2003

Digital Marketing, Social Media

0.5

Printing & Publishing (0.2)

2006

Online Education, Web Development

1

Travel Agencies (0.3), Retail Sales (0.2)

2009

Cloud Computing, Cybersecurity

1.5

Manufacturing (0.8), Financial Services (0.4)

2012

Big Data, App Development

2

Administrative Services (1.0)

2015

Social Media Management, Online Customer Service

2.5

Telecommunications (0.5), Cable TV (0.3)

2018

Artificial Intelligence, Blockchain

3

Retail Sales (1.2), Transportation (0.8)

2021

E-commerce Fulfillment, Data Science

4

Cashiers (1.5), Factory Workers (1.0)

2023

Renewable Energy (due to climate change awareness)

0.8

Fossil Fuel Industries (0.6)


Wednesday, July 24, 2024

Telcos are Generally "Not Good" at Predicting Sources of New Product Revenue

The mobile and telecom industries have rarely been good at predicting the actual "new product" value of each next-generation mobile network. The switch to 2G, the first digital platform, was seen as increasing the efficiency of spectrum use; improving voice reliability; enhancing security and increasing capacity.


The use of Signaling System 7 also allowed sending of control data “out of band” (separately from the voice channel) to manage the network. Aside from security advantages, that also provided faster call set up and tear down. 


SS7 supports a wide range of then-advanced telecommunication services, such as call forwarding, caller ID, call waiting, and short message service (SMS or text messages). It is reasonable to say that telco executives primarily expected operating efficiencies from SS7, as well as the ability to embed voice handling features in the network. 


To be sure, some “new products and capabilities” were anticipated for 3G and 4G, ranging from mobile internet access in 3G to video entertainment and video conferencing (3G and 4G). But many more-exotic use cases failed to develop. In fact, many would attribute the one-time leadership of smartphones by BlackBerry devices and Research in Motion as an example of the role of email value. 


In fact, some might argue that mobile e-mail was a more salient outcome of 3G than “web access,” which remained painfully slow on 3G networks. Likewise, some might argue that “turn by turn” navigation apps were a clearer “new use case” for 4G, early on, than video conferencing or video entertainment. 


Later 4G development of ride sharing apps is an example of a new use case not envisioned by 4G architects. 


The point is that we are not very good at predicting what new use cases, revenue drivers and value users will see in each next-generation network, beyond an order of magnitude increase in bandwidth and a similar improvement in latency performance (which most users will not be able to identify). 


Likewise, virtually nobody predicted that fixed wireless would be the first “at scale” new revenue opportunity 5G enabled. 


Despite the fact that fixed wireless was hardly ever mentioned as among the new features and capabilities 5G would bring, it has, so far, proven to be the best example of a new use case that generates significant new revenue for mobile service providers selling it. Over the past few years fixed wireless has generated virtually all the net new home broadband connections in the U.S. market, for example. 


Some might point to consumer dissatisfaction with cable-provided home broadband as one reason for the growing adoption of fixed wireless, where the value proposition might well turn on lower prices rather than an increase in speeds, in some cases, while being based on higher speeds and lower prices in other cases. 


The former might be the driver in urban areas where fixed wireless provides lower prices; the latter the driver in rural areas, where fixed wireless might actually be faster than DSL but also less costly than cable. 


The point is that the value proposition in urban areas is “good enough” speed and lower price; while in rural areas the value proposition might be “faster speed, lower price.”


J.D. Power surveys actually suggest that fixed wireless consumer satisfaction are comparable to optical fiber and cable home broadband, and actually higher than satisfaction with fiber-to-home services. By definition, that means the value proposition for fixed wireless is based on price and speed, as FTTH always is much faster. 


2023 U.S. Residential Internet Service Provider Satisfaction Study

J.D. Power

Fixed wireless, especially 5G FWA, leads in customer satisfaction scores compared to fiber optic and cable. FWA outperforms fiberoptic by over 20 points on a 1,000-point satisfaction scale.

October 2022 - August 2023

2023 HSI Customer Satisfaction Survey

HighSpeedInternet.com

Fixed wireless customers gave the highest overall satisfaction ratings, particularly in pricing and customer service.

September 8, 2023

CableTV Survey

CableTV

Fixed wireless and fiber providers scored highest in overall satisfaction, with cable providers lagging, especially in price satisfaction.

2023

S&P Global Market Intelligence

S&P Global Market Intelligence

Fixed wireless is particularly popular and satisfying among rural customers, where it often represents the most reliable option.

2023


Price seems important. “T-Mobile’s current FWA (fixed wireless access) plan retails for $50/month, but that falls to $30/month for customers subscribing to its Magenta MAX mobile plan,” analysts at Ookla say. “Verizon prices at a slight premium to T-Mobile, with its FWA service currently retailing for $60/month, but falling to $35/month with select 5G mobile plans.”


Speed might be less important in urban areas, but perhaps more important in rural areas. The median download speed across the United States for all fixed providers combined in the third quarter of  2023 was 207.42 Mbps, Ookla says. The median speeds for Verizon and T-Mobile fixed wireless was 122 Mbps, Ookla notes. 


  

source: Ookla 


And though fixed wireless has traditionally been viewed as an attractive platform in rural areas, 5G home broadband gains are driven by consumers in urban markets. Both T-Mobile and Verizon are getting 80 percent of their gross additions in urban locations, Ookla says. 


Among the key takeaways from 5G home broadband is that the value proposition--as always--is a mix of drivers, including both speed and price. Consumers seem willing to accept less of the former to get more of the latter. 


If the success of 5G home broadband shows anything, it is that the consumer estimation of home broadband value can change when a new value proposition is available (faster speed and lower price).


According to Nerdwallet, the most-popular home broadband service plans in 2023 cost $41.31 per month for 104 Mbps in the downstream direction. According to OpenVault, in the first quarter of 2024 about 33 percent of U.S. home broadband customers purchased service plans operating between 200 Mbps and 400 Mbps. Some 11 percent purchased service plans operating between 100 Mbps and 200 Mbps. 


The download speed provided in this top-selling category was 104 Mbps.


Telcos Poised for Historic Home Broadband Market Share Gains

At least some observers believe the traditional U.S. home broadband market share pattern could be quite different by 2030, though there might be some disagreement about shares held by various ISPs (cable, telco, independent providers). Historically, U.S. cable operators have held as much as 70-percent share of the home broadband market. But that seems unlikely--perhaps impossible--to sustain.


Perhaps the single biggest change could be a shift of share in favor of telcos, using both the fiber-to-home and wireless platforms. Many observers would not be surprised to see a major telco-cable share of about 40 percent each, with others holding the remaining share.


That 20-percent share would be held by ndependent ISPs and smaller telcos.


Traditionally, U.S. cable operators have held about 65 percent to 70 percent share of the market. By some estimates, telcos collectively have held 20 percent to 30 percent share, with satellite, fixed wireless and independent providers having single-digit shares. 


Most observers expect change, the only disagreement being the degree of share to be lost by cable operators and the share to be gained by telcos and other providers. 

source: GlobalData 


Also, some forecasts that look at technology (optical fiber, hybrid fiber coax, fixed wireless, satellite) often obscure the fact that telcos use both optical fiber, fixed wireless and digital subscriber line technologies. 


So a forecast of 20 percent to 25 percent “telco” market share in 2030 might not include all telco FTTH share or fixed wireless (largely provided by the big telcos T-Mobile, Verizon and AT&T. In principle, telcos might have 35 percent to 45 percent share in 2030, based only on use of fixed wireless, DSL and FTTH shares held by Verizon, T-Mobile and AT&T. Other telcos also will have some share (Lumen Technologies, Frontier, others). 


Provider Category

Estimated Market Share in 2023

Expected Market Share in 2030

Cable Companies (Comcast, Charter, Cox, etc.)

45-50%

35-40%

Major telcos (AT&T, Verizon)

25-30%

20-25%

Satellite Providers (Dish, Viasat)

5-10%

3-5%

Fiber Optic Providers (Google Fiber, Frontier, municipal providers)

5-10%

25-30%

Fixed Wireless Access (FWA) (T-Mobile, Verizon, AT&T)

5-10%

15-20%


The point is that it remains likely that telcos and cable operators will continue to hold most of the share (perhaps 70 percent to 85 percent or more), with independent providers gaining share. 


As capital intensive as fiber to the home remains, independent and smaller providers cannot hope to reach significant portions of the U.S. home broadband market very quickly, especially given the fact that, in most markets, up to 96 percent of existing share is held by the incumbent telco and cable operator. 


Even if one discounts the local telco’s ability to respond to a new fiber overbuilder market entry, the new FTTH provider will often face an incumbent cable operator able to offer gigabit-per-second downstream speeds. 


Provider Type

2023 Market Share

Discussion

Cable

~70%

Among cable operators and wireless companies controlling 96% of the U.S. home broadband market, MSOs (cable companies) accounted for nearly 70% market share

Telco

~26%

Estimated based on cable's 70% share of the 96% controlled by cable and telcos

Other

~4%

Includes independent ISPs, fixed wireless, and other providers


Why Marginal Cost of Content Creation is Generative AI's Superpower

Virtually every observer might agree that artificial intelligence will automate laborious tasks and therefore increase process efficiency. A...