From Attention to Recommendation Economy: How AI Is Changing Our Relationship with Influence

For the past two decades, the internet has increasingly been shaped by what has come to be known as the like economy: a digital environment in which attention, engagement, likes, other attention buttons, shares and followers have become valuable currency forms. Social media platforms have built enormous businesses around capturing and retaining our attention, while brands have learned to compete not only for consumers' money, but for a place in their alarmingly crowded mental space.[1]

Artificial intelligence may be about to change this relationship fundamentally.

The next generation of AI is not simply another channel through which advertising and marketing messages are delivered. This is our profound reality already. Increasingly, AI systems are becoming a part of the decision-making process itself: Instead of searching Google, browsing dozens of websites, reading reviews and asking friends for recommendations, a consumer can simply ask AI application:

  • Which car should I buy?

  • Which hotel should I stay at?

  • Which jacket would suit me?

  • Which laptop is worth the money?

The AI does not merely provide information here; it can interpret preferences, compare alternatives, explain different trade-offs and, increasingly, help the user arrive at a decision. It uses the history of earlier discussions with the user, and observations made in that (usually lengthy) line of dialogue.

That creates a profound shift in the economics of influence. In the traditional like economy, brands competed for attention of users. In an AI-mediated economy, they may increasingly compete for recommendation. The valuable moment is no longer necessarily the moment when a consumer sees an advertisement or likes a social media post. It may be the moment when AI decides which three products, restaurants, hotels or other services are worth mentioning to that particular person. [2]

This is particularly interesting because AI recommendations can feel fundamentally different from advertising. When a friend says, "You should try this," we interpret the recommendation through the relationship we have with that person. When an AI knows our previous questions, preferences and constraints, its recommendation can begin to occupy a similar psychological space in our thinking. Thus, the interaction here becomes conversational rather than transactional, even compared to social media interaction.

And this is where AI applications such as ChatGPT start to look surprisingly similar to social media.

Certain differences still prevail. AI apps do not look like traditional social media. There may be no feed to scroll, no visible follower count and no obvious stream of posts from friends. Yet the underlying behaviour can begin to resemble the similar pattern. People return regularly, develop habits, discover products, discuss ideas, ask for opinions, share experiences and increasingly use the same environment to decide what to buy, where to go and what to believe which is worth their attention.

The crucial difference is that traditional social media asks, "What should I show this person?" AI then increasingly asks, "What does this person actually need?"

That distinction will have extensive consequences for brands, retailers and consumers alike, which we explore later.

From the Attention Economy to the Recommendation Economy

For much of the history of the commercial Internet, the fundamental question for companies and brands was: How do I get noticed?

Search engines created the first answer: companies invested in Search Engine Optimisation because an appearance near the top of a search result meant being visible at the precise moment when a potential customer was looking for something.

Social media introduced another answer: build an audience, generate engagement and become part of people's daily attention.

Both models were ultimately built around visibility.

AI may be creating a different model altogether.

The emerging question here is no longer 'How do I get noticed?' but 'How do I become recommended?'

This is a subtle, but profound change. A search engine can give a consumer hundreds of possible answers. A social media platform can expose them to thousands of products, opinions and pieces of content. An AI assistant can instead narrow the field. It can say: "Based on what you've told me, these are the three options I would consider."

The commercial value of that moment could be enormous.

In the attention economy, the scarce resource is human attention. In the recommendation economy, the scarce resource may increasingly be AI’s attention: the limited number of products, companies and ideas that an AI chooses to bring into a conversation. In our busy lives, we may even be grateful that some instance can actually narrow our options down and limit our attention span.

This not only gives us a new established way to search for product information, but also creates a new kind of digital visibility.

From SEO to AEO — and Beyond

For years, companies have optimised their websites so that search engines can find, understand and rank them. The logic in this practice called Search Engine Optimisation (SEO) has been relatively straightforward: you constantly improve your website’s content and structure to gain more organic search results for your site, and thus appear on the first results for the right search terms. This was of course done in order to have a better chance of winning the customer first.

Generative AI changes the interface between the customer and the information.

A consumer may no longer search for "best electric SUV under €70,000" and then visit ten websites. They may simply ask an AI to compare the alternatives. The AI becomes the interface through which the market is explored.

This has already given rise to concepts such as Answer Engine Optimisation (AEO) and Generative Engine Optimisation (GEO).[3] But perhaps these terms still describe this phenomenon too narrowly. The deeper change is not merely about optimising content so that an AI can find it.

It is about becoming eligible for recommendation.

A company can have excellent SEO and still be largely invisible in an AI conversation if its products, reputation, customer experience or publicly available information do not give the model enough reasons to mention it.[4]

The traditional search economy rewarded ranking.

The social media economy rewarded engagement.

The emerging AI economy may reward relevance, trust and contextual fit.

The End of the Present Marketing Funnel?

This also changes the traditional marketing funnel.

For decades, marketers have talked about awareness, consideration, conversion and loyalty as separate stages. Advertising creates awareness. Content builds consideration. Sales converts the customer. Customer experience creates loyalty.

An AI assistant can potentially compress several of these stages into a single conversation.

That is no longer a generic search query. It is a highly contextualised buying brief.

Imagine someone asking:

"I'm looking for a comfortable electric SUV. I drive around 18,000 kilometres a year, have two children, occasionally travel long distances and don't want something that feels enormous in the city. My budget is around €70,000. What should I look at?"

The AI can ask follow-up questions, establish priorities, compare products, explain compromises and eventually direct the consumer towards a purchase.

In that situation, the brand does not necessarily control the first impression. However, the AI does.

And that may be one of the most important changes in digital commerce since the rise of the search engines.

AI Apps as a Social Layer

There is another development that I find even more interesting.

As people use AI assistants repeatedly, they begin to develop something that resembles a relationship with the AI interface: They return to previous conversations. They explain their preferences. They ask for opinions. They share experiences. They ask the system to remember things. They use it to explore ideas before discussing them with other people or before finding information from other sources. In this role, the human and the AI interact directly to engage in a socially meaningful conversation and consideration of different options whether it is something to do with new interests, shopping or even more personal decisions. [5]

This starts to blur the traditional distinction between a search engine, a shopping assistant, a personal advisor and a social platform.

Social media historically connected people to other people's attention.

AI can connect people to contextualised intelligence.

And the more context the system has, the more useful the interaction can become.

That creates an unusual feedback loop. The user provides information about themselves in conversation. The AI uses that information to personalise future recommendations. Those recommendations influence what the user buys, reads, watches or visits. Those choices generate new experiences, which then become part of future conversations.

The result is not simply a personalised search engine.

It is something closer to a personalised discovery environment.

That could eventually make the distinction between "social media" and "AI assistant" increasingly difficult to maintain.

From Followers to Memory

There is also an intriguing parallel here.

The currency of traditional social media has been the follower.

Now, the currency of the AI relationship may be context.

A social media platform knows that you follow a particular brand. An AI assistant may know that you have considered buying its producto, say, three times, rejected it because of the price, prefer another material, dislike a particular design feature and are likely to reconsider the purchase maybe next year.

That is a fundamentally different type of commercial relationship.

It moves the centre of gravity from audience to individual.

And this is why the AI recommendation economy could become much more consequential than another evolution of digital advertising. Advertising traditionally tries to influence what people notice - recommendations can influence what people consider reasonable, relevant and worth buying.

The question for companies therefore becomes much more difficult than simply:

"How do we get people to like us?"

It becomes:

"How do we become the company AI thinks is relevant when the right customer asks the right question?"

That may be one of the defining marketing questions of the next decade.

The Emergence of AI Brand Equity

For decades, brand equity has essentially been a human concept. It describes the associations, perceptions, trust and expectations that people attach to a brand. A strong brand is not simply recognised; it carries meaning. Consumers know what to expect from it, what it stands for and, often, why they might choose it over an alternative.

But AI introduces a new participant into this equation.

Increasingly, brands are not only being evaluated by consumers. They are also being interpreted by machines.

Let’s consider a simple example. A consumer asks ChatGPT:

"I'm going to London with my family. We have three adults and two school-age children, and we're looking for a hotel in central London. We want something comfortable, family-friendly and within walking distance of major attractions, but we don't want to spend more than £500 per night."

The consumer has not asked about a particular hotel brand. They have not searched for a specific hotel. They have simply described a need.

Yet the AI may immediately narrow thousands of possible hotels down to a handful of recommendations. For the hotels that appear in that answer, something important has happened. They have entered the consumer's consideration set without necessarily having bought an advertisement or competed for a search ranking at that exact moment.

The recommendation here might be based on a combination of factors:

  • Location

  • Price

  • Room configuration

  • Reviews, reputation and available information on how the hotel appears to match the family's requirements mentioned

Now imagine two hotels with similar prices and similar locations. One has a very strong traditional brand among consumers, while the other has a much clearer digital footprint: its room types are well documented, its family facilities are consistently described across multiple sources, its location is clearly communicated, and customer reviews repeatedly associate it with families and convenience.

The second hotel may be easier for an AI system to understand and match to the user request.

That is where the concept of AI Brand Equity becomes interesting.

It isn’t simply about brand awareness. It is about the strength of the machine-readable associations surrounding a brand, product or service.[6]

When a consumer asks, "Which hotel is right for my family?", does the AI understand that your hotel is family-friendly? Does it know which room types are suitable for five people? Does it understand your location? Can it distinguish your proposition from competitors? And, crucially, does enough trustworthy information exist for the AI to confidently include you in its recommendation?

In the traditional brand economy, the objective was to create a strong association in the consumer's mind:

Family holiday → Hotel brand X

In the AI recommendation economy, the association might increasingly need to exist in another form:

Family holiday + central London + five people + £500 → Hotel brand X is relevant

That is a very different kind of mixture of needs for a brand asset.

The challenge for marketers is therefore becoming two-fold:

1.      Build brands that people remember and trust.

2.      Also build brands that machines can accurately understand and confidently recommend.

It suggests that the future of brand building may involve managing not only how a brand is perceived, but also how a brand is represented, understood and retrieved by AI systems.

The most powerful brands of the future may need to be both memorable to humans and legible to AI.

The New Influencer

As AI becomes an increasingly important part of our decision-making, it is worth understanding its place in the long history of influencers — from billboards and newspaper ads to radio and television commercials, social media, and beyond.

AI allows us to broaden our personal memory and discussion history beyond our own minds, bringing us an ability to use it as a sounding board for discussions, ideas, and the everyday dilemmas we face. This can lead to deeper-than-before insights and more considered suggestions. However, we should not forget that AI, too, can influence our choices in subtle ways. Certain brands and products may be given greater visibility than others, and we should remain mindful of the interests and agendas that may lie behind the recommendations we receive.

Tommi Kukkonen

Digital service professional, technology enthusiast & Doctoral candidate in Technology.

Next
Next

Meet the Founder: Suvianna Grecu and the Story Behind AI for Change