EXPLAINED: How AI platforms are building the next big advertising economy

As AI assistants become the new starting point for search, planning and decision making, the real question is no longer whether these platforms will monetise but how. Google, OpenAI, Perplexity and a growing number of enterprise deployments are already shaping a revenue model that blends ads, subscriptions, API usage and commerce. What changes is the fabric of the internet itself. What remains constant is that high intent always gravitates toward the most efficient interface.

Gemini and the Future of Search Monetisation

Google’s moves with AI Overviews inside Search have already signalled what a commercial future for conversational interfaces might look like. Even without confirmed ads inside the Gemini app, the infrastructure exists. AI Overviews currently run ads powered by the same Search and Shopping systems that fund most of Google’s business.

Akshay Mathur, Founder & CEO, Unpromptd, captures the strategic pivot clearly: “If Google introduces ads into Gemini, does it become a third performance channel or simply a UX layer on top of Search?” His assessment suggests that Gemini will likely emerge as a powerful high intent surface, even if it uses the same underlying auction and measurement engine as Google Ads.

For marketers, Gemini will feel meaningfully new because conversational requests have a different rhythm and depth than keyword searches. Yet operationally, it sits within the same monetisation logic: intent signals, automated bidding, inventory pricing and attribution built on Google’s existing stack. The interface changes. The business engine does not.

How Conversational Interfaces Pull Traffic Away from the Open Web

Much of the industry anxiety today revolves around the erosion of organic traffic. AI Overviews already answer a growing percentage of queries without requiring a user to click. Assistants accelerate that shift by resolving tasks inside the conversation. Adding ads only increases the likelihood that user journeys remain inside the platform.

Mathur notes that this movement is not triggered by ads but will be amplified by them. Once assistants deliver comparison charts, recommendations, next steps and commerce paths directly inside the interface, publishers and merchants see less traffic and less discoverability. Value concentrates in the assistant layer, which becomes the new gateway to high intent activity.

What Advertising Inside AI Assistants Will Actually Look Like

While no AI assistant has launched a full scale advertising product, early prototypes already exist in plain sight. Perplexity’s sponsored follow up questions, Google’s contextual cards and utility based prompts in AI Overviews provide a preview of the formats that can exist without breaking user trust.

The formats most likely to succeed are those that sit adjacent to answers rather than merging with them. Sponsored next steps, product suggestions, service cards and utility recommendations work because they map naturally to what users expect from a digital assistant. The formats that will fail are ones that blur the line between organic output and paid messages. Any erosion of trust invites regulatory scrutiny and risks user abandonment.

The Shift in Auction Dynamics When Moments Become Ad Slots

The biggest commercial shift inside assistants is not the appearance of ads but the change in what constitutes inventory. A single keyword becomes less important than a moment inside a conversation. Eligibility depends on the flow, tone, depth and direction of the session. Sponsored follow up questions create fewer but more contextually valuable ad slots, which attract premium bidding.

Instead of competing on a single input, brands will compete across the intent signals generated through multi turn interactions. Google’s existing automated bidding systems can already optimise across these signals. The assistant simply provides a richer canvas of user behaviour.

Targeting and Attribution in a Multi Turn World

A conversational thread provides more information than any single query. Preferences, constraints, comparisons, hesitations and readiness all surface naturally. Targeting becomes context driven, and bidding aligns to the highest probability point of influence. Attribution cannot rely on last click logic. It must trace contribution across the full conversation.

What emerges is a more precise understanding of intent but also a more complex landscape for marketers who must learn to operate inside fluid, dynamic journeys rather than predictable keyword funnels.

How AI Platforms Will Make Money: The Four Pillars

Sajal Gupta, CEO, Kiaos Marketing, outlines the revenue logic emerging across the industry. “When ads come into conversational AI, you are going to get very relevant advertising inventory. Today search gives you inventory based on keywords. In conversation it will be based on intent and context.”

But he is clear that ads are only one pillar. The free layers available today will not be permanent. “Going forward, they will restrict how much you can do for free. You will have to pick one AI engine to pay for. You cannot pay for all of them.”

According to him, AI platforms will depend on four major revenue streams: subscriptions, ads, affiliate commissions and API usage. Affiliate revenue becomes especially important in high intent moments. “You ask for the best phone to buy. The AI recommends Amazon with a link. If you purchase, the AI platform gets a commission.”

This model echoes the evolution of search but moves it into a multi turn environment where the assistant guides the journey, not just the query.

Why Advertisers Will Choose Some Platforms Over Others

As Gupta points out, distribution becomes the deciding factor. Gemini will seep through Android and Chrome. Copilot will be embedded inside Office 365. OpenAI leads in API usage and enterprise adoption. Perplexity is gaining credibility for deep research tasks. Each platform will attract advertisers based on the environments users already inhabit.

Creatively, agencies will differentiate not by choosing a specific model but by how they deploy it. If two agencies using the same model produce identical work, Gupta says the failure lies with the agency, not the AI. Human interpretation remains the creative differentiator.

The Road Ahead: Specialisation Over Generalisation

The long term trajectory is toward sharper differentiation. Users will subscribe to one primary AI engine based on what it does uniquely well. Platforms will invest deeply in specialisation because that becomes the reason users pay.

The advertising economy that once lived inside Search is now being rebuilt inside conversations. The platforms that balance utility with monetisation, and trust with commercial ambition, will shape the next decade of the internet.