For most of its history, Google's economic engine was advertising. In 2025, advertising accounted for 73% of Alphabet's revenue, Google Cloud, including its enterprise infrastructure and Workspace products, contributed around 15%, and subscriptions, platforms, and devices another 12%. These proportions reflect an arrangement that has defined much of the internet for three decades. Billions of people receive useful services for free, while advertisers pay to reach them. Google's extraordinary achievement was to build products that became indispensable to everyday life while monetizing the enormous volume of attention and commercial intent flowing through them.
But the economic environment in which this arrangement developed is changing, and so are the expectations of the people using these services.
A more demanding consumer
Consider the world consumers may be entering over the next few years. Energy and transportation costs remain vulnerable to geopolitical shocks, employment growth is unevenly distributed, and household incomes may struggle to keep pace with prices. Even without a major recession, families could find their disposable incomes under persistent pressure. People respond by becoming more selective about spending, comparing alternatives more carefully, and postponing purchases. Businesses face the corresponding challenge of attracting customers whose willingness to spend has become harder to predict.
As the cost of making a poor decision rises, so does the value of making a better one. Imagine planning a family holiday with a fixed budget, comparing flights, hotels, local transportation, cancellation policies, and seasonal demand across multiple websites. An AI system that understands your family's circumstances and can evaluate these alternatives intelligently would save considerable time and potentially money. The same applies to insurance, education, financial decisions, and everyday purchases. As people share more context about their circumstances, their expectations rise accordingly. They expect recommendations that reflect what the system already knows about them.
The rising cost of attention
We explored this changing relationship in The Attention Tax, particularly how AI can reduce the effort people spend navigating information and commercial choices. For years, searching, scrolling, comparing alternatives, evaluating advertisements, and moving between websites have been accepted as part of getting things done. Conversational AI changes that expectation. Once someone has explained their needs, budget, preferences, and constraints, repeatedly making them search and filter through alternatives begins to feel unnecessary.
This becomes particularly interesting when household budgets are under pressure. Consumers may become more reluctant to spend money while placing greater value on services that help them avoid waste. A €20 monthly subscription becomes attractive if it consistently prevents expensive mistakes, identifies better alternatives, or saves hours of work. As AI systems develop richer context about their users, the quality of assistance can improve across an expanding range of activities, making the subscription increasingly valuable through repeated use.
The emerging subscription habit
We have spent decades becoming accustomed to paying monthly for entertainment, music, cloud storage, and productivity software. Consumer AI introduces a capability that can be applied across many aspects of daily life. Someone might begin subscribing for writing, studying, or research, then gradually discover its usefulness for travel, financial decisions, shopping, and everyday administration. The subscription acquires value as the user discovers more situations in which the service can help.
This is why I suspect we are underestimating the potential rate of growth in paid consumer AI subscriptions. The market is still commonly discussed in terms of developers, professionals, students, and early adopters, although the larger opportunity lies with ordinary consumers who increasingly depend on these systems to make better decisions. As the services become more dependable and capable of completing tasks, paying for intelligence could become as familiar as paying for music or entertainment, with the additional advantage that its usefulness extends across multiple categories of spending and activity.
Google's second customer
Google enters this transition with an advantage that is difficult to reproduce. Search, Maps, Gmail, Calendar, YouTube, Android, and Chrome already occupy an enormous portion of people's digital lives. These services provide distribution and opportunities to introduce AI assistance within activities people already perform. With appropriate permissions, context from across these services could make Google's recommendations considerably more relevant. Someone planning to relocate abroad, for example, might begin by researching cities with affordable housing and access to nature, then gradually involve the system in comparing employment opportunities, finances, transportation, and practical arrangements.
Google already sells consumer subscriptions, including AI services. The opportunity is to develop a much larger direct-paying relationship around a capability whose usefulness can grow with accumulated context and repeated interaction. Historically, consumers generated the attention and commercial intent that attracted advertisers. Increasingly, they may also pay Google directly for better judgment, planning, and assistance. Across Google's enormous user base, even modest subscription conversion could create a substantial revenue stream, depending on pricing, retention, and the cost of delivering the service. The relationship becomes particularly valuable as consumers incorporate AI into more of their everyday decisions.
What happens to the first customer?
Advertisers are navigating many of the same changes as consumers. Tighter household budgets increase pressure on businesses to make marketing more effective, while conversational AI allows people to express commercial intent in far greater detail. A family describing its travel dates, budget, preferred destinations, and accommodation requirements provides more actionable information than someone entering a generic search for cheap flights. Google and Meta could use this richer understanding to improve commercial matching, potentially attracting advertising expenditure from less effective channels even when overall budgets remain constrained.
The interesting challenge is that these customer relationships place different demands on the same system. Advertisers want their products recommended to potential buyers, while paying consumers expect assistance that serves their interests. As users share increasingly personal context, trust becomes central to the value of the subscription. Google's ability to preserve the integrity of its recommendations while supporting commercial discovery will influence how successfully the two businesses reinforce one another.
The opportunity ahead
The market for paid consumer AI is still developing, with rapidly improving free services and intense competition. Nevertheless, people are becoming accustomed to explaining complex intentions to software and expecting useful outcomes in return. Macroeconomic pressures could accelerate this behavior as households seek better ways to manage their resources and make informed decisions. Entrepreneurs will respond by improving the quality, reliability, and accessibility of these services, further strengthening the case for subscriptions.
For Google, this creates the possibility of a more diversified business built around a deeper relationship with the consumer. Advertising can remain a powerful revenue engine, enterprise AI can expand through Cloud and Workspace, and direct consumer subscriptions can become substantially more important. Google spent more than two decades building one of the world's most successful businesses around helping people find what they were looking for. Its next opportunity may come from helping them decide what to do once they find it, and increasingly, getting it done.
And for that, the user may be willing to pay.