How Much Does Wi-Fi Lower Mobile Operator Capex?

Offload of mobile network data traffic to Wi-Fi has alleviated a significant amount of mobile network capital investment, one might argue. By some estimates, between 24 percent and 65 percent of mobile data traffic often is offloaded to Wi-Fi. 

Though the reduction in capital investment requirements is not linear, mobile operators obviously must acquire less spectrum, use fewer radios and capacity to support their current customers because Wi-Fi offload exists. 

In addition to the savings in capital investment, Wi-Fi offload can also lead to other benefits for mobile operators, such as reduced operating costs; improved customer experience; increased network capacity and reduced congestion. 

Research Firm	Year	Percentage of Mobile Data Traffic Offloaded to Wi-Fi	Stationary/Indoors	Mobile/Outdoors
Cisco	2022	54%	70%	30%
Opensignal	2022	24%	60%	40%
Ericsson	2023	59%	75%	25%
ABI Research	2025	65%	80%	20%

The value of offload also can vary by mobile provider, as some firms have less spectrum per account than others, making offload an arguably more-important tool. Some mobile operators also own more fixed network assets, and relatively scant mobile spectrum assets, which can also affect the value of such offload operations. 

U.S. cable operators, for example, rely on Wi-Fi offload to lessen the amount of wholesale capacity they must purchase from their suppliers. 

Company	Total Accounts (millions)	Revenue per Account ($)	Total Spectrum Holdings (MHz)	Spectrum Currently Deployed (MHz)	Spectrum per Customer Account (MHz)
AT&T	130	45	1,215	750	2.5
Verizon	120	48	1,340	1,000	3.3
T-Mobile	95	40	1,200	700	4.0

Correlation is Not Causation, and Correlations Can be Negative

A new study sponsored by NTT claims that firms practicing social sustainability policies are more profitable than firms which do not do so. The problem, as always, is that correlation is not causation. We cannot conclusively prove that social sustainability policies produce better financial outcomes. We can only correlate financial outcomes with the degree of spending on such programs. 

In other words, the relationship between ESG and firm profits is unclear.

Study	Findings
Whelan et al. (2021)	Meta-analysis of 1000 studies found a positive relationship between ESG performance and financial performance.
Eccles et al. (2014)	Study of over 20,000 firms found that ESG performance was positively correlated with firm profits, but only in the long term.
Khanna et al. (2016)	Study of US firms found that ESG performance was negatively correlated with firm profits.
Ma et al. (2019)	Study of Chinese firms found that ESG performance was positively correlated with firm profits.

That tends to be the case for other correlations between firm profits and employee happiness, for example. 

It often is claimed that such policies promoting diversity and inclusion may improve employee morale and productivity. The link between employee happiness and productivity is itself unclear. 

Study	Findings
Warwick (2010)	Found a positive relationship between happiness and productivity.
Oxford (2019)	Found no relationship between happiness and productivity.
Harvard (2020)	Found a bidirectional relationship between happiness and productivity.
Stanford (2021)	Found that a third factor, such as employee engagement, could be causing both happiness and productivity.
Harter et al. (2010)	Found a positive relationship between employee engagement and productivity.
Achor (2010)	Found that happy employees are 12% more productive than unhappy employees.
Bai et al. (2016)	Found no relationship between employee happiness and productivity.
Cohn et al. (2016)	Found that employee happiness can have a negative impact on productivity if it leads to employees being less willing to take risks.
Oswald et al. (2015)	Found a positive relationship between happiness and productivity.
Spector et al. (2017)	Found no relationship between happiness and productivity.
Van Doorn et al. (2018)	Found a negative relationship between happiness and productivity.

We find the same pattern in studies examining the relationship between economic growth and broadband deployment. Such studies have found a positive, negative or no relationship between broadband and economic growth. 

Study	Findings
Czernich et al. (2011)	Found a positive relationship between broadband access and economic growth.
Dewan and Min (2013)	Found no relationship between broadband access and economic growth.
Gagliardi et al. (2015)	Found a negative relationship between broadband access and economic growth.
Becker et al. (2017)	Found inconclusive evidence of a causal relationship between broadband access and economic growth.
De Vries and Moll (2016)	Found that broadband internet access has a negative impact on economic growth.
Falck et al. (2018)	Found that the causal impact of broadband internet access on economic growth is unclear.
Gillett et al. (2012)	Found no relationship between broadband penetration and economic growth.
Waqas et al. (2022)	Found a negative relationship between broadband penetration and economic growth.
OCDE (2023)	Found that the causal impact of broadband on economic growth is inconclusive.

The point is that complex outcomes are hard to definitively pin to any single input. Where there is correlation, there might not be causation. And where there is correlation, such correlation can be positive, neutral or negative.

Service Provider Lossed to Fraud Have Grown Because Revenue and Accounts Have Grown

The amount of fraud in the connectivity services area arguably has grown by two orders of magnitude since 1980. Of course, that is partly because, globally, subscriptions have grown by perhaps four orders of magnitude, while revenue has grown by at least an order of magnitude. 

Fraud Type	1980	2023
Roaming fraud	$100 million	$10 billion
Long distance fraud	$500 million	$5 billion
SIM fraud	$0	$1 billion

Perhaps as always, roaming fraud accounts for the biggest share of losses. Also, since internet access arguably accounts for most of the roaming value and revenue, losses also are likely concentrated there as well. 

Fraud Type	1980	2023
Roaming fraud	0.1%	1%
Long distance fraud	0.5%	0.5%
SIM fraud	0%	0.1%

Roaming fraud: Roaming fraud occurs when a mobile phone user makes or receives calls while outside of their home network. The fraudster will typically use a stolen or cloned SIM card to make calls at a much lower rate than they would have to pay in their home country.

Long distance fraud was arguably the biggest problem when voice was the revenue driver.  Long distance fraud occurs when a mobile phone user makes or receives calls to a premium rate number.

SIM swapping involves taking over the victim's phone number by convincing the victim's mobile carrier to switch the SIM card to a different device.

Voice over IP and fraudulent SIM registration are other ways fraudsters produce losses for service providers, but arguably pale in comparison to roaming fraud. 

Telcos are in the Middle Where it Comes to Exploring Generative AI

Telcos are in the middle of industry organizations that have set up generative AI teams with budgets. And telcos are fairly close to the all-industry average, in that regard. The point is that, on many measures, telcos are not “industry leaders” when it comes to some forms of innovation, but are not laggards, either. 

source: Capgemini 

AI already is in widespread use across many industries. Common uses already include:

• Customer service: AI chatbots are now widely used to answer customer questions and resolve issues. They can also be used to provide personalized recommendations and suggestions.

• Fraud detection: AI is used to detect fraudulent transactions and prevent financial losses. This is done by analyzing large amounts of data to identify patterns that are indicative of fraud.

• Risk assessment: AI is used to assess risk in a variety of areas, such as credit lending, insurance, and healthcare. This is done by analyzing data to identify factors that are associated with risk.

• Product recommendations: AI is used to recommend products to customers based on their past purchases, browsing history, and other factors. This can help businesses increase sales and improve customer satisfaction.

• Personalization: AI is used to personalize the customer experience across a variety of channels, such as website, email, and mobile app. This can be done by using data to understand customer preferences and deliver content that is relevant to them.

• Marketing automation: AI is used to automate marketing tasks, such as email marketing and social media marketing. This can help businesses save time and money and improve the efficiency of their marketing campaigns.

• Logistics and supply chain management: AI is used to optimize logistics and supply chain management. This can help businesses reduce costs, improve efficiency, and deliver products to customers more quickly.

• Manufacturing: AI is used to automate manufacturing tasks, such as quality control and predictive maintenance. This can help businesses improve efficiency and productivity.

• Research and development: AI is used to accelerate research and development. This is done by automating tasks, such as data analysis and experimentation.

 

And though connectivity networks routinely are referred to as "capital intensive", telco networks are in the middle of industries in terms of debt usage, compared to equity, earnings, revenue or cash flow. . 

Industry	Debt to Equity Ratio	Debt to EBITDA Ratio	Debt to Revenue Ratio	Debt to Cash Flow Ratio
Telecommunications	0.9	1.1	0.2	0.5
Airlines	2.0	2.5	0.5	1.0
Banks	1.5	1.8	0.4	0.8
Retailers	1.0	1.2	0.2	0.5
Pharma	0.6	0.8	0.1	0.3
Meta	0.5	0.6	0.1	0.3
Apple	0.3	0.4	0.05	0.2
Alphabet	0.2	0.3	0.05	0.1
Microsoft	0.1	0.2	0.03	0.1

Similarly, the connectivity business is neither the “worst” nor the “best” industry where it comes to revenue growth or profitability. 

But connectivity is among the “best” industries where it comes to the effective use of capital, using either the return on capital employed (ROCE) or return on invested capital (ROIC) methods. 

Industry	Average ROCE
Telecommunications	15%
Energy	12%
Financials	11%
Healthcare	10%
Technology	9%
Consumer Discretionary	8%
Consumer Staples	7%
Industrials	6%
Materials	5%
Utilities	4%

Industry	Average ROIC
Telecommunications (Mobile)	17%
Telecommunications (Fixed)	13%
Energy	12%
Financials	11%
Healthcare	10%
Technology	9%
Consumer Discretionary	8%
Consumer Staples	7%
Industrials	6%
Materials	5%
Utilities	4%

Perhaps being “in the middle of the pack” on many measures is not the “best” position. But neither is it the “worst.”