Why "Winner Takes All?"

Digital markets often are characterized by a winner take all market share, led by oligopolies. One might argue that always is the case, which is why antitrust law exists. Power laws and Pareto outcomes seem quite prevalent in nature and economics.  

“Winner take all” outcomes arguably also happens for incomes of sports stars or other celebrities as well. The issue is why. Many cite the shift to intellectual capital as a reason for the rise of winner-take-all outcomes.

Other drivers are said to include economies of scale, scope and learning. Digital products and services have high fixed costs and low-to-zero marginal costs. That is scale. Scope is different. Such economies result when the production of one good makes production of another easier.

There also are economies of learning. You might call that the experience curve, or Henderson's Law. There is an inverse correlation between a product’s value added and the manufacturer’s experience in its production and marketing. 

In other words, as a company’s understanding and knowledge increases, the cost per unit tends to decrease. That is going to apply for artificial intelligence as it gains knowledge of particular tasks. 

Network effects can exist, where each additional node on a network increases the value of the whole network. One can consider fads and fashion an example of network effect. 

Platform business models also seem to encourage oligopolies. Platform businesses create value and earn revenue by matching complementary sets of people. Platforms make their money by extracting a transaction fee of some sort for connecting buyers and sellers, once they reach critical mass. 

Platform business models also are“multi-sided” and facilitate transactions between many types of users. That is why the notion of “ecosystems” is so prevalent. Also, a multi-sided market allows users to switch roles. A renter can become a housing supplier. A rider can become a ridesharing driver. A message sender is normally also a message recipient. Content producers are content consumers.

Some might credit sophisticated analytics, allowing mining of huge data sets, as also contributing to winner-take-all outcomes. As some colloquially note, “he wins who has the most data.”

Brand value also will tend to reinforce winner-take-all outcomes. Sometimes that happens because brands create ecosystems of goods, services and suppliers that “lock customers into a platform” by creating higher switching costs. 

Ironically, lower barriers to scale and competition might be among the reasons the “rule of three” (for example) seems so prevalent in digital markets. But some would note that most markets are oligopolies, digital or not. 

Others might mention the prevalence of leverage, luck or feedback loops as causes of the outcomes. Whatever the reason, relative performance, rather than absolute performance, is key. 

“Top performers might be only slightly more skilled than the people one level below them, yet they receive an exponential payoff,” says Farnam Street. “A small difference in relative performance—an athlete who can run 100 meters a few microseconds faster, a leader who can make better decisions, an opera singer who can go a little higher—can mean the difference between a lucrative career and relative obscurity.”

Were value deemed absolute, then a product or service with 90 percent of the performance of the “best” option would sell for 90 percent of the price of the “best” option. 

But what we often see in digital markets is that the “best” product gets 50 percent to 80 percent market share. The 10th-best product, with 90 percent of the best product’s utility, gets almost no share. 

That “winner take all” pattern arguably is true of digital goods in part because marginal costs are close to zero and because network effects often are generated. The former means the cost of supplying very-high demand is not onerous. The latter means utility grows exponentially with scale. 

Connectivity services tend not to exhibit that pattern as well, sometimes because high capital investment requirements discourage competitors. 

Back in 1981 economist Sherwin Rosen wrote about the phenomenon. Convexity is the principle that small differences in talent become magnified in larger earnings differences. 

source: Medium

 

source: Medium

Imperfect substitution is another concept that explains why “winner take all” markets develop. That can be driven by the value of an ecosystem of goods and services, brand preferences, switching costs or any other combination of values that create product differentiation. 

“Winner take all” might always have been an issue, but it is magnified in an era where digital economics operate.  

U.S. Gigabit Take Rates Surpass Typical Inflection Point

Openvault says 10.5 percent of U.S. customers now buy home broadband service operating at one Gbps or faster.  Openvault says gigabit take rates have doubled since the second quarter of 2021, when 4.8 percent of households bought gigabit service. 

That is important because 10-percent adoption often is the inflection point for adoption of consumer products.

source: Openvault 

Perhaps more important, the “typical” speed plan purchased by consumers has continued to shift upwards as well. 

Openvault says 80 percent of subscribers have chosen a speed tier of 100 Mbps or faster, with more than half of them (47.5 percent) subscribing to a 100-to-200 Mbps tier.

In the first quarter, using a slightly different reporting method, Openvault said 80 percent of U.S. households were buying internet access at 100 Mbps or faster. 

The percentage of subscribers provisioned for speeds of 100 Mbps or less fell by half, from 39.9 percent in 2Q20 to 20.1 percent through the same period, says Openvault. 

The first quarter 2021 report showed 9.8 percent of U.S. households were buying gigabit or faster connections.

source: Openvault

At Least for the Moment, B2B Sales Professionals See No Return to Old "In-Person" Sales

By the start of 2022, only 15 percent of business-to-business sales executives expect in-person sales meetings to be the norm, a McKinsey survey suggests. About 83 percent of the time, it appears B2B marketing and sales personnel believe digital and remote channels are as effective, if not more effective, than before the Covid pandemic, for reaching new customers. That is the same figure professionals claim for sales to existing customers. 

The obvious issue is whether this becomes the new normal for B2B sales and marketing. Currently, about 70 percent of sales interactions are remote or digital, only about 30 percent in person. 

There are lots of implications for spending priorities, skill sets and channels. 

source: McKinsey 

source: McKinsey

What if Connectivity Providers Have Done About as Well as AWS, Azure, Apple in Innovating?

“How to increase the value of connectivity services” has been a top concern of most telecom executives for a couple of decades. It can be argued that not so much as changed, for all the attention and effort. Nor should that come as a surprise.

Looking at all manner of revenue and market share growth winners, very few firms actually have managed to change their fundamental position in the value chain, or their revenue growth drivers. Google and Facebook still rely mostly on advertising. Microsoft still relies mostly on software sales. Apple still relies mostly on device sales. 

Several of those firms have made strides, however. Apple gets a bit more revenue every year from “services.” In recent quarters “service” revenues have been more than 20 percent of Apple’s total revenues. 

source: Statista 

Amazon Web Services now represents 11 percent of Amazon total revenue. Microsoft’s cloud computing revenue is growing, but it is hard to say by how much, as those contributions are grouped with Xbox and server revenues. Some estimate Azure revenue at about 17 percent of total Microsoft revenues.  

Slideware is not “product” or “service.” “Value” is not its representation on a chart but its role in a customer or user’s life and work. Generally speaking, observers say the highest value comes as one gets closer to the end user in the function stack. 

Basically, supplying bandwidth is lower in the value chain than content, advertising or apps. Every app requires connectivity, but differentiation and value tend to be clearest in the “business” layer that links a user or customer’s problem and the application that solves that problem. Google Maps; Amazon or Alibaba; iPhone; Airbnb or a productivity suite rather than internet access, for example. 

source: Cisco 

The “problem” for connectivity providers is slow revenue growth in the core connectivity services business. Under the best of circumstances, revenue grows in line with growth of gross domestic product. But connectivity and computing  services also exhibit a trend towards lower prices and profit margins over time. 

That implies a constant search for new products and revenue streams with higher value. And, to be sure, argues GSMA, that has happened. Connectivity providers have added significant non-core revenues. 

But that remains a challenging task. It is no mistake that some representations of the internet value chain place “connectivity” where it appears to be closer to the end user or customer. Our general notion is that this is where value is highest. 

source: GSMA 

As a practical matter, there are key disagreements about how much emphasis to place on chasing new roles and how much on operational improvements in the core business. Innovation can be quite hard, and very risky. 

Mobile operators have more options than fixed network providers. But every connectivity provider faces long-term risk of commoditization. Still, telcos have not done much worse than Microsoft, Apple and Amazon or Google at innovating, looking at the percentage of "new" revenue sources created over the past two decades.

"Higher Value" Often is the Key to Revenue Growth, in IoT and Elsewhere

Most firms that are leaders in any segment of the computing and communications ecosystems eventually find they must consider moving into new segments of their existing value chains to maintain growth. “Up the stack” or “across the value chain” are typical options. 

For suppliers positioned high in the value stack (application providers, for example) movement down the value chain also makes sense. The older notion of “vertical integration” applies, even when there is no attempt to completely vertically integrate operations. 

For connectivity providers, the movement is almost always going to involve moving up the stack, from physical layer IP connectivity towards application and platform supply roles. It almost never makes much sense to integrate “lower” in the value chain by becoming a manufacturer of radios, cable, fittings, connectors,  servers, chipsets or other infrastructure used to create and support IP connections.

Generally speaking, that also is the route for adding new value to internet of things business operations, argue researchers at Beecham Research, in a paper prepared for Sierra Wireless. That path corresponds to the way greater value can be added to any connectivity-oriented service, others might add.

To the extent that solutions to business problems ultimately drive the value of any information technology or communications investment by enterprises and businesses, the more complete and simple a solution, the higher the value to the customer. 

In other words, people want connections with other people, so they buy iPhones, not chipsets.  People want transportation, so they buy automobiles, not tires, batteries and engines. 

Though all IP communications require access to IP networks, and therefore internet service provider connections, nobody buys broadband except as a means to use higher-value applications for communications, e-commerce, entertainment, discovery or learning. 

Beecham researchers argue that  “as the application becomes more important to the business operations, it is increasingly important to manage all aspects of the connectivity (Connectivity Management) to ensure the application is not disrupted by the connection being broken, not working correctly or being hacked.”

Beecham Research, Sierra Wireless

In other words, a mission-critical function requires mission-critical reliability. And that might extend beyond the physical internet connection to the device and the applications and insight  built on the generated data. 

There is another way to phrase this observation: value is generated at many points in a complete value chain, for IoT or any other solution or product. Entities can generate more value and revenue by supplying greater portions of the value chain. 

Surprising Data on Online Video Viewing

Rarely, if ever, do statistics about video and television consumption surprise me. This does. Assuming it is not mislabeled, the data suggests that older consumers spend more time watching online video on mobile devices than do younger viewers.

source: Parks Associates 

I would have guessed just the opposite.

Global Telco Capex Will Grow at 3% CAGR Through 2025, Says Dell'Oro Group

Global spending on broadband access equipment and customer premises equipment  is expected to have a three-percent compound annual growth rate 2020-2025, an upward revision from the zero-percent CAGR capital investment  pattern Dell'Oro Group had expected in January 2021. 

Service providers seem to be  pulling forward some of their upgrade projects, including those involving the transition from copper to fiber, Dell’Oro says. 

An expectation of falling capex would not have been unreasonable. Some observers had predicted a dip in cable operator capex, for example. Other global capex forecasts called for flat spending. Also, fixed network capex has been relatively constrained, compared to mobile capex, for some years.   

source: S&P Global

How Feasible is "Orchestration" as a Business Model?

“Rather than continuing down the road of being a connectivity provider, CSPs need to transition to become an intelligent service orchestrator ,” says Bengt Nordström, Northstream managing director. “Taking a connectivity and wholesale approach in 5G, or becoming a reseller to the edge, will put CSPs in danger of seeing revenues dry up.”

“Becoming an orchestrator” is viewed by some as a move up the stack to becoming a “service enabler,” presumably allowing additional value creation and revenue generator possibilities. 

source: Ericsson 

Also, there is a less customer-facing understanding of “orchestration” that involves the internal operations of the access and transport network itself. In that sense, orchestration is about automation more than service creation, the creation of end-to-end service more than occupying a new role in the value chain.   

Orchestration makes sense, no matter which definition is used. But one form is easier than the other. Orchestrating the internal operations of the communications network is one thing, taking on a new role in the value chain, ecosystem or functions stack is a different matter. 

The issue is how feasible it is that the access provider becomes the app orchestration supplier, “positioning themselves as the key to enabling a wide range of services through their ability to connect a complex ecosystem of new digital offerings,” notes Nordström. 

Otherwise, 5G is likely to turn out as did 4G, with much of the new value reaped by over the top app, commerce and content suppliers, not access providers. 

source: Ericsson 

“For many CSPs (communication service providers), the aim of digital transformation programs is to empower them to become service enablers or service creators, which increases their commercial enterprise opportunity,” Ericsson says. 

But it might also be fair to point out that this requires moves akin to becoming system integrators, especially integrators with vertical market domain expertise. Some might note that others in the value chain already have staked out this position, requiring access providers to muscle out other existing competitors. 

Connectivity providers might assemble multi-cloud access capabilities, for example. But most access executives would be happier with a more-developed role as service creators. 

Also, in the internet era, it has proven easier to move down the stack than up the stack. That is to say, it has proven easier for entities with domain knowledge to add lower layer functions to create new offers, than to assemble new offers from below. 

In itself, the advice to “orchestrate” makes operational sense. All virtualized networks require orchestration. Whether the use of the term also extends to business role, and the odds of succeeding in such roles, is a different matter.

"Platform Business Models" Were Proposed 20 Years Ago

Enron attempted to become a bandwidth trading platform two decades ago, but imploded in 2001 for reasons directly related to accounting fraud. Highly controversial at the time, Enron’s notion of creating a platform for buying and selling bandwidth services appears not to be ahead of its time, but very much in line with current thinking about platform business models. 

It is hard to say what could have happened had Enron not gone bankrupt. The basic model of creating an exchange seemed to work in a number of other markets. And there was a logic to the hope of creating a more-liquid marketplace where complex products could actually be traded. 

source: Enron

But there was significant resistance from many other partners, for one obvious reason: the exchange was touted as leading to lower prices, and it was not deemed to be “in our own best interest” for other big carriers to put their assets into such an exchange. 

The offer was real-time contracting and delivery of capacity at network pooling points. It was one example of “on demand bandwidth provisioning,” long an industry hope. Enron promised an “efficient” market. But other potential partners might have heard “commodity pricing.”

About the last thing capacity executives wanted was anything that would turn their core product into a “commodity.”

Two decades later, businesses in many industries are urged to “become platforms,” earning their revenue by facilitating transactions. 

It’s the same idea, twenty years later.

Digital Customer Experience is Way Easier than Recreating a Business

For most people, all-digital customer fulfillment or “all-digital marketing” is an easier concept than “becoming a different kind of business” (customers, products, industries). In the mobile business, the concept is captured by “all-digital customer activation. 

source: Capgemini 

For that reason, much thinking about “digital transformation” focuses on customer-facing marketing and sales. 

source: Capgemini 

The notion of finding, ordering and taking delivery of a product all online is now familiar, and is the preferred activation model used by many upstart mobile service providers. 

Transforming a business model (revenue, customers, products, position in value chains) is far more difficult, and arguably quite a bit riskier. For connectivity providers there is another danger: transformation by rivals who then compete directly with connectivity providers. 

Going direct to customer can be helpful for connectivity providers, but arguably is more useful for other e-commerce, content or application providers. Indeed, one might argue that is precisely what the internet promises: direct access to end users and customers.

Cable Leads U.S. ISP Accounts, Telcos have Most to Gain

AT&T, Comcast and Charter Communications own networks covering the most U.S. residents (and presumably housing units). But it is difficult to directly correlate population covered with “homes passed” or even “locations served.” 

Most locations are served by at least two different terrestrial providers. Subscriber percentages hinge at least partly on marketing strengths (cable operators have at least 70 percent of the installed base). 

U.S. ISP Population Coverage and Subscribers
Firm	Population %	Subscriber %
AT&T	42	15
Comcast	36	29
Charter	34	27
Verizon	17	7
Lumen	9	2
Cox Comm.	7	5
Altice	5	4
sources: FCC, Leichtman Research

Still, the numbers suggest telcos have lower take rates than do cable operators, which corresponds to reality. Also, the telcos arguably have the greatest upside from network upgrades that would match or exceed cable performance metrics. With appropriate pricing, that should allow telcos to eventually close the take rate gap with cable companies, all other things being equal.