Friday, May 3, 2024

Google Leads Market for Lots of Reasons Other Than Placement Deal with Apple

A case that is seen as a key test of potential antitrust action against Google, with ramifications for similar action against other hyperscale app providers such as Apple, Meta and Amazon, will be decided this year. 


The U.S. Justice Department sued Google in U.S. District Court for the District of Columbia, alleging that the company illegally protected its monopoly in internet search, partly by paying billions to persuade companies, including Apple and Samsung, to use its search engine.


But some might argue such deals are not unusual in business. In retail, manufacturers sometimes pay stores "slotting allowances" for prominent shelf placement. This ensures their products are more visible to consumers, potentially increasing sales. Similarly, Google's payment gives it higher visibility on iPhones.


In media, websites might offer premium ad placements for a higher fee. These placements are more likely to be seen by users, potentially driving more clicks and revenue. Google's payment to Apple could be seen as a form of premium placement for its search engine.


Some media outlets publish sponsored content for a fee. Google's prominent placement on iPhones could be viewed as a form of sponsored content, where Apple is "sponsoring" Google Search as the default option.


Cable companies might charge a premium for certain high-value channels to be included in their packages. 


But some might argue those analogies do not apply so well. In retail, some might argue, there are typically multiple brands competing for shelf space. Consumers can easily choose alternative products on nearby shelves. On iPhones, changing the default search engine can be cumbersome and some users might not even be aware they can do so.


Others might argue that transparency is an issue. The argument is that slotting allowances and pay-per-click placements are often disclosed, so consumers know a product placement is sponsored.


Likewise, prior precedent will play a key role in determining the outcome. 


Case

Year

Court

Key Takeaway

Relevance to Google-Apple Deal

Microsoft v. US

1998

US Court of Appeals

Dominant companies can't unfairly disadvantage competitors.

Supports arguments that Google's payments stifle competition.

Eastman Kodak Co. v. Image Technical Services, Inc.

1996

US Court of Appeals

Dominant companies can't use their power to foreclose competitors from markets.

Supports arguments that Google's deal limits user choice for search engines.

FTC v. Qualcomm

2017

Federal Trade Commission

Tying arrangements where access to one product is conditioned on another can raise antitrust concerns.

Supports arguments that Google's payments create an unfair tie-in between iPhones and Google Search.

Sun Microsystems, Inc. v. Microsoft Corp.

1993

US Court of Appeals

Companies have some freedom to choose what software to integrate.

Could be used by Apple to defend its right to choose a default search engine.

Verizon Communications Inc. v. FCC

2012

Supreme Court

Companies might have more leeway in managing their platforms.

Potentially weakens arguments against exclusive deals on private platforms like iPhones.


The coming decision hinges on whether Google's payments stifle competition. The judge will have to balance consumer choice and possibly the impact on innovation. Users can change their default browser, though, so the issue is whether such default installs actually reduce competition, or only reinforce user preferences. 


According to StatCounter, Google’s Chrome holds about 65 percent of the installed base or share, globally. The argument on one side is that this reflects user preference. The argument on the other side is that such leadership is dependent on tying agreements to some extent. 


Browser

Percentage Share

Source

Chrome

64.73%

StatCounter

Safari

18.56%

StatCounter

Edge (Microsoft)

4.97%

StatCounter

Firefox

3.36%

StatCounter

Opera

2.86%

StatCounter

Samsung Internet

2.59%

StatCounter

Other Browsers

2.93% (combined)

StatCounter


But there are many reasons why Google is dominant in search. Google's search algorithm is widely considered the most advanced and effective, delivering relevant and accurate results to users' queries. In other words, users might simply prefer Google because they believe it delivers the best results. 


In that regard, extensive troves of data might allow for greater and more effective personalization and search accuracy as well. 


Some might point out that Google possesses a vast amount of user data and search history, allowing it to personalize search results and continuously refine its algorithms for better performance. Others might point to global presence; language support and local customization.


Network effects--or possibly simply scale benefits--then also kick in.  More users generate more search queries and data, which Google can leverage to improve its search algorithms and personalize results. 


A larger user base exposes Google Search to a wider range of searches and information needs, which might arguably help the algorithm identify emerging trends and improve its understanding of user intent.


Still, “network effects” refer to platforms where value grows as the number of users grows. That is not necessarily an advantage for Google search. No single user or query necessarily benefits from all other users in a traditional way. 


It might be less clear how the Google ecosystem of products and services (Gmail, Maps, YouTube) contributes to "stickiness,” but some value is to be found there. 


Deals with device manufacturers like Apple (to be the default search engine) or browsers like Firefox (to be the default search provider) help by ensuring prominent placement on popular platforms.


Others might point to Google’s ability to keep investing in its product, given its strong revenue position, as well. 


The point is that there are all sorts of potential ways Google might benefit from trends that flow from having large market share. But Google arguably only benefits because end users prefer it. Maybe Google search leads simply because users prefer it to other products. 


Study Title



Publication Date


Publishing Venue


Key Findings



Network Effects?


Why Google Search Matters: Understanding User Preferences for Leading Search Engines






2022














Journal of the Association for Information Science and Technology










Examines user preferences and identifies factors like search accuracy, result comprehensiveness, and user interface design as key drivers of Google's market share.

Indirectly explores network effects through the role of data in improving search accuracy.







The Google Search Algorithm: A Survey












2020















ACM Computing Surveys













Analyzes the technical aspects of Google's search algorithm, highlighting its sophistication and ability to handle complex user queries and natural language processing.

Focuses on the technical underpinnings of Google's search advantage, not directly on network effects.







The Data Advantage: How Google Uses Big Data to Power Search








2021













Harvard Business Review











Discusses Google's vast data collection and its role in personalizing search results and improving algorithm performance.





Argues that data advantage is a key driver of Google's success, potentially linked to network effects through the continuous accumulation of user data.

The Network Effects of Search Engines: An Economic Analysis










2019















Information Economics and Policy













Analyzes the network effects concept in search engines, acknowledging the role of data accumulation but suggesting alternative explanations like brand reputation and switching costs.

Provides a critical perspective on the role of network effects in Google's dominance.









Search Bias and User Lock-In: An Examination of Google's Search Engine Market Power









2023















Journal of Competition Law & Economics












Investigates potential anti-competitive practices by Google, raising concerns about lock-in effects that might discourage users from switching to alternative search engines.

Discusses the potential negative aspects of a dominant search engine with a large user base.









Also, the search function has broadened over the decades, with different sorts of searches, many related to commerce, now prominent. That might matter if courts have to evaluate search market power. There are more types of search and more providers. 


Study Title


Publication Date

Publishing Venue

Key Findings


Traditional Search %

Commerce-Related Search %

Evolving User Search Behavior: A Comparative Analysis of Traditional and Shopping Queries




2023











Search Engine Journal










Identifies a significant rise in shopping-related queries, with users increasingly starting their product research online.

62%

38%

The Rise of "Buy" Online: How Search Trends are Shaping Consumer Behavior





2022











McKinsey & Company










Reports a surge in searches combining informational keywords with shopping intent (e.g., "best laptops for students 2024").

58%

42%

Search Intent: Decoding the Why Behind the What












2021















Moz















Categorizes search queries by intent (informational, navigational, transactional). While informational searches remain prevalent, transactional searches are steadily growing.

68%

32%


The point is that the search market is broader than it used to be, and Google search is part of that broader market. 


Category

Potential Leaders

Search Engines

Google, Bing

Shopping Platforms

Amazon, Walmart, eBay

Social Commerce Platforms

Instagram, Pinterest

Comparison Shopping Engines

Google Shopping, Shopzilla

Voice Assistants

Siri (Apple), Alexa (Amazon)


---------------


AT&T Intros "Turbo" QoS Features for Mobile Customers

AT&T has introduced quality of service features for its 5G service, intended to offer a more-consistent access experience for gaming, social video broadcasting and live video conferencing, known as "AT&T Turbo."


But AT&T has not really said “how” it is providing what appears to be data prioritization, nor been specific about the degree of performance improvement. One would think the technique involves giving AT&T Turbo users a different “class of service” tags. 


Even before the advent of 5G network slicing, which enables the creation of virtual private networks, mobile and fixed network operators had a few techniques to create different classes of service. 


DiffServ (Differentiated Services) is a widely-used standard that classifies traffic into different categories based on a DSCP (Differentiated Services Code Point) value. The DSCP value is a code embedded in the IP packet header that identifies the traffic type (such as high-priority voice call, video streaming, web browsing). Routers within the network use this DSCP code to prioritize packets and allocate resources accordingly.


Operators also cna use QoS queuing to create separate queues for different traffic classes within routers. High-priority traffic gets placed in a higher priority queue, ensuring it gets processed first. This helps minimize latency for critical applications like online gaming or video conferencing.


ISPs also can set bandwidth limits for specific types of traffic or entire user accounts. This helps prevent network congestion and ensures fair allocation of bandwidth among users. For example, an ISP might throttle video streaming after a certain data usage threshold to avoid impacting other users' internet experience.


Traffic shaping can be used to define the rate and burstiness (peak data transfer) for different traffic types, which can help smooth out traffic flow and prevent congestion.


It is unclear which techniques AT&T might be using. 


Service providers have tended to provide such services for business customers, as network neutrality principles generally apply only to consumer services, whether fixed or mobile. Called  “AT&T Turbo”


But that is changing, as mobile service providers start to offer quality-of-service features for consumer mobile services, and as 5G network slicing becomes available . 


Provider

Service Name

Description

Verizon (US)

5G Ultra Wideband with Business Preferred Data

Offers prioritized data access for businesses during times of congestion.

T-Mobile (US)

5G Advanced Network Solutions

Suite of features including network slicing (dedicating network resources for specific uses) and private network solutions for businesses.

AT&T (US)

FirstNet with 5G

Priority access and dedicated network resources for first responders and public safety agencies.

Deutsche Telekom (Germany)

Magenta Enterprise 5G

Offers network slicing and guaranteed bitrates for businesses.

Vodafone (UK)

Vodafone Business - 5G Network Slicing

Provides dedicated network slices with guaranteed bandwidth and latency for businesses.

NTT Docomo (Japan)

docomo 5G SA with Network Slicing

Offers network slicing for various use cases, including low-latency and high-capacity applications.


In part, these are moves that begin to extend fixed network differentiated service tiers to mobile service. Though mobile data plans long have included variable data consumption plans that also offer product differentiation, the newer QoS plans increase ability to differentiate by quality of service, not simply speed or data usage. 


Even if there are other differentiation mechanisms (content bundling; prepaid; network coverage), mobile operators still have significant room to create distinctiveness using speed, data plans and latency-related features of their networks. 


Tiered data plans based on usage create distinct tiers of service for users who use differing amounts of mobile data.


Still, up to this point, mobile operators have tended not to differentiate on access speed, offering one “best effort” speed for all users. Only business users have had access to service plans that guarantee minimum abscess speeds. 


Likewise, only business service plans have offered prioritized data access during times of network congestion. 


The new shift to QoS features for consumers seems largely a result of network slicing features of 5G networks, which will enable minimum guaranteed latency performance.


But network slicing can, in principle, also support guaranteed minimum speeds as well, allowing the creation of consumer service plans that provide QoS features.


Thursday, May 2, 2024

Cloud Computing Keeps Growing, With or Without AI


source: Synergy Research Group


With or without added artificial intelligence demand, cloud computing will continue to grow, Omdia analyst predicts.



Wednesday, May 1, 2024

How Much Revenue Do AWS, Azure, Google Cloud Make from AI?

Aside from Nvidia, perhaps only the hyperscale cloud computing as a service suppliers already are making money from artificial intelligence at a significant level, as in billions of U.S. dollars per year. 


Nobody really knows how big a revenue contribution “artificial intelligence as a service” now contributes to Amazon Web Services, Microsoft Azure or Google Cloud revenue. But estimates by Gartner and Forrester Research analysts suggest “AI as a service” might contribute $1.5 billion to $2 billion for Google Cloud; $3.5 billion to $4 billion for Azure and perhaps $5 billion to $6 billion for AWS. 


Analyst Firm

Google Cloud AIaaS Revenue

Azure AIaaS Revenue

AWS AIaaS Revenue

Gartner

$2 Billion

$3.5 Billion

$5 Billion

Forrester

$1.5 Billion

$4 Billion

$6 Billion


Virtually everyone who thinks about this might agree that revenue for AIaaS should continue to grow. As a percentage of total revenue, Gartner believes AIaaS will represent five percent to 10 percent of Google Cloud revenues by about 2025; some eight to 12 percent of Azure revenue; and perhaps 10 percent to 15 percent of AWS revenue, all by 2025. 


Analyst Firm

Google Cloud

Microsoft Azure

AWS

Gartner


5-10% (by 2025)


8-12% (by 2025)


10-15% (by 2025)



While not providing exact figures, Microsoft, Google, and Amazon discussed AI's role in their recent financial reports. 


Microsoft third quarter2024 earnings call attributed three percentage points of the 29 percent revenue increase in Azure to AI. 


Google’s first quarter 2024 earnings call mentioned AI's positive impact on Google Cloud and YouTube, but without specific revenue figures. 


AWS now has a $100 billion annual run rate. If revenue related directly to AI is three percent of AWS total revenue, that implies $3 billion in annual AWS revenues.

Monday, April 29, 2024

Study Suggests AI Has Little Correlation With Long-Term Outcomes

A study by economists Iñaki Aldasoro, Sebastian Doerr, Leonardo Gambacorta and Daniel Rees suggests that an industry's direct exposure to artificial intelligence has surprisingly little impact on its long-term outcomes, despite AI being a permanent driver of higher productivity. 


“We find that a sector’s initial exposure to AI has little correlation with its long-term increase in output,” they note. 


The reason is that, ultimately, general equilibrium effects arising from higher demand for a sector’s output

matter much more than the initial increase in productivity,” they say. In other words, the level of customer demand for any class of products matters more. 


So the following illustration of industry growth does not primarily reflect the impact of AI. 

 

source: Bank for International Settlements 


The authors do argue that the primary AI impact will be on jobs and occupations with more cognitively demanding tasks. 


Even the effects of AI on inflation are uncertain, they argue. On one hand, by raising productivity, AI adoption boosts supply, which is disinflationary. On the other hand, firms need to make substantial investments to take full advantage of AI, which could contribute to higher inflation.


Since inflation responses hinge on expectations, much depends on households’ and firms’ anticipation of the impact of AI. If they do not anticipate higher future productivity, AI adoption is initially disinflationary. 


In contrast, when households and firms anticipate higher future productivity, inflation rises immediately. 


And that is the rub. If virtually everybody expects AI will boost productivity, then expectations related to inflation also will tend to rise.

How Do You Invest in AI If You Cannot Initially Quantify AI Outcomes?

Enterprise technology executives face a dilemma when deploying generative artificial intelligence: unless there is measurable return on investment (either predicted or realized), the investment will not be made, or continue. 


But Gen AI is quite new, so few entities will have at least a year’s worth of experience to make such outcome assessments. 


So many projects essentially require some leap of faith or willingness to experiment. 

source: Deloitte


And while it might be easy to argue that desirable outcomes include improving existing products and services fostering innovation gaining efficiencies and reducing costs, metrics must be devised and time has to elapse before measurement is possible.


Use Case

Metrics

Tracking Method

Content Creation (e.g., marketing copy, product descriptions)





Content creation speed  Content quality  Customer engagement with content


Track time spent creating content  Compare human-generated vs. AI-generated content quality through A/B testing  Monitor website traffic, conversion rates, and customer feedback

Product Design and Development










Number of design iterations required  Time to market for new products  Customer satisfaction with product design


Track design cycle times  Monitor time spent on prototyping and development  Conduct customer surveys to gauge satisfaction with product design and functionality






Data Augmentation (e.g., training machine learning models)







Accuracy of machine learning models  Training time for machine learning models  Cost of data acquisition

Track model performance metrics (e.g., precision, recall)  Compare training times with and without AI-generated data  Monitor costs associated with data collection and labeling




Personalized User Experiences (e.g., product recommendations, chatbots)










Customer satisfaction with personalization  Conversion rates on recommended products  Number of customer interactions handled by chatbots

Conduct customer satisfaction surveys  Track website click-through rates and conversion rates for recommendations  Monitor chatbot performance metrics (e.g., resolution rates, customer satisfaction scores)






Directv-Dish Merger Fails

Directv’’s termination of its deal to merge with EchoStar, apparently because EchoStar bondholders did not approve, means EchoStar continue...