About the Interviewee
Andrea Tortella is the founder of Thrad.ai, the advertising infrastructure purpose-built for the AI era. Thrad enables AI apps to monetize through in-chat advertising and gives brands a direct channel to reach millions of daily users inside LLM-powered applications — at the exact moment those users are most engaged and most intent-driven. Unlike SEO-adjacent AI plays or organic content strategies, Thrad operates squarely in paid media: real ad placements, inside real AI conversations, reaching real people at peak intent. Andrea’s thesis is that the shift from search to conversational AI represents a generational change in how attention is monetized — and that the infrastructure to capture that shift does not yet exist at scale. Thrad is being built to be that infrastructure.
Executive Summary
The internet’s most durable business model has always been advertising. Google built one of the most valuable companies in history by placing relevant ads alongside search results — by inserting commercial signals into the moment a user expressed intent. The question for the current decade is whether the same model can be rebuilt for a world where search is being replaced by conversation.
Andrea Tortella believes it can, and that the opportunity is bigger than most people realize. This paper draws on a conversation with Andrea to explore the fundamental shift in user behavior driving the AI advertising opportunity, how in-chat advertising differs from every ad format that came before it, why task-based targeting represents a step change in advertising precision, and what it takes for brands to compete effectively in this new environment. The conclusion is that AI-native advertising is not an iteration on digital advertising — it is a reimagination of it.
A Paradigm Shift, Not a Product Update
From Clicking to Conversing
The premise behind Thrad is not that AI chatbots are a new channel to add to an existing media mix. It is that the fundamental nature of how people interact with the internet is changing — and that this change has consequences for every business model built on top of it, advertising included.
For decades, the internet ran on clicks. Users searched for things, clicked on links, scanned pages, and clicked again. The entire architecture of digital advertising — keywords, display units, programmatic bidding, retargeting — was built around that click-based interaction model. Brands competed for position on search result pages and banner placements because that is where attention was.
That model is now under pressure from a different direction. Users are spending more time in extended, involved conversations with AI systems. They are not clicking and scanning — they are asking, elaborating, and following threads of inquiry that can span dozens of exchanges. The session depth is fundamentally different, and with it, the nature of the commercial opportunity.
“There’s less clicking, less searching in the classic way. Everybody’s just asking longer questions, more involved conversations. It follows logically that you can monetize and democratize that technology through advertising — that’s the revenue stream.”
The Google Parallel
Andrea draws a direct line from Thrad’s model to the story of Google. Before Google, online advertising was display-based — banners, pop-ups, interstitials that interrupted the user’s experience without regard for what the user was actually doing. Google’s insight was that intent-based advertising, placing commercial messages alongside expressed needs, was categorically more valuable than interruption-based advertising.
The same logic applies to AI. A user in a conversation with an AI system is, by definition, expressing intent — often in far more granular detail than a search query ever could. A search query might be three words. A conversation might be three hundred, with context, history, and nuance that reveals not just what someone wants but why they want it, what constraints they are working under, and how ready they are to act. The commercial signal embedded in that kind of engagement is orders of magnitude richer than anything search advertising ever produced.
What Makes In-Chat Advertising Different
No Context Switching
The defining characteristic of every major digital ad format before in-chat advertising is interruption. A display ad pulls the user’s eye away from the content they came to see. A social media ad breaks the scroll with something unrelated. A pre-roll video forces the user to wait before accessing the content they wanted. Even the most carefully targeted programmatic ad is, at its core, a context switch — it asks the user to redirect their attention.
In-chat advertising works differently. The user is already locked into a specific conversation. They are engaged, focused, and actively processing information. An ad that surfaces within that conversation — one that is relevant to the thread being discussed — does not ask the user to redirect attention. It arrives inside the attention that is already fully committed.
“You’re not context switching the entire time. You’re actually locked into a specific conversation. Any ads that get served in that context are just way more powerful. That’s the main differentiator.”
Dynamic Creative at the Individual Level
Traditional advertising operates on a one-to-many model. A brand creates a piece of creative — an image, a video, a headline — and pushes it across channels to as many relevant people as possible. The best campaigns are carefully targeted and well-crafted. But they are still fundamentally static: the same message delivered to a segment, not to an individual.
In-chat advertising inverts this completely. Because the AI understands the specific user, the specific conversation, and the specific moment within that conversation, the creative does not need to be pre-built and distributed. It can be generated in response to context. Every single impression is its own iteration — adapted to the individual, informed by everything the conversation has revealed about who they are and what they need right now.
This is not personalization in the conventional marketing sense, where a user’s name is inserted into an email subject line or a retargeting pixel follows them around the internet. It is personalization at the level of meaning — where the commercial message is shaped by the actual cognitive state of the person receiving it.
Knowing Where the User Is in Their Decision Journey
One of the most persistent challenges in digital advertising is timing. Brands want to reach consumers when they are ready to act, not just when they are casually browsing. The tools available to address this — behavioral targeting, lookalike audiences, purchase intent signals — are approximations at best. They infer readiness from proxies rather than observing it directly.
A conversational AI context offers something those tools cannot: direct, real-time evidence of where a user is in their decision process. Is the user exploring a topic for the first time, or have they been researching it across multiple sessions? Are they comparing options, or have they narrowed to a specific need? Are they ready to act, or are they still forming a view? These signals surface naturally in the language and structure of the conversation itself — and they are available to a well-designed ad system without any inference required.
Task-Based Targeting: The Next Frontier
From Demographic Buckets to Actual Intentions
Demographic targeting was the dominant paradigm of digital advertising for years. Reach men aged 25 to 34. Reach users who have visited a competitor’s website. Reach people who match a behavioral profile associated with purchase intent. These approaches work at scale precisely because they do not require understanding any individual user — they require only that a large enough population of roughly similar users be assembled and served the same message.
The problem is that demographics are a crude proxy for need. A 32-year-old man in a major city might be planning a holiday, managing a health condition, making a business decision, or doing none of those things at any given moment. His demographic profile tells a brand almost nothing useful about what he needs right now. AI conversations tell a brand exactly what he is thinking about, in his own words, in real time.
Thrad’s approach is built around what Andrea calls task-based targeting — and it represents a fundamental departure from every targeting methodology that preceded it. As AI systems become more agentic, users are increasingly delegating specific tasks to them: book this flight, draft this email, research these options, post this update, manage this project. Each task is a precise, declared intention. The gap between that intention and a commercial opportunity is, in many cases, negligible.
“Targeting specific tasks is extremely powerful. I’m telling my AI, go write this, go post this, go book my holiday. All these tasks have constraints — a human needs to achieve a goal. We show them brands that can actually help them achieve that goal.”
How Thrad’s API Integration Works
Thrad integrates directly with LLM infrastructure via API, inserting sponsored messages into the conversational flow at appropriate intervals. The system receives a call at every conversation turn, giving it visibility into what is being discussed and the ability to make real-time decisions about whether and what to serve. This is not display advertising bolted onto an AI interface — it is advertising infrastructure built from the ground up for the conversational context.
The targeting logic operates on the content and context of the conversation rather than on user profiles assembled from third-party data. What is the user talking about? What task are they trying to accomplish? What constraints are they working under? What has the conversation revealed about their intent? These signals drive the ad decision, and they are refreshed at every turn of the conversation.
Real-Time Signals and the Speed of Advertising
Beyond Demographics: Environmental and Behavioral Signals
The richness of in-conversation data extends well beyond the explicit content of what a user says. Context layers that would be difficult or impossible to integrate into traditional ad targeting become immediately actionable in a conversational AI environment. Andrea offers a straightforward example: weather as a signal.
A brand selling outerwear can instruct the system to show a jacket ad only when the weather at the user’s current location is below a certain temperature threshold. In a traditional display campaign, building that kind of real-time environmental trigger requires significant technical infrastructure and still only approximates relevance. In a conversational AI context with API-level access, it is a simple conditional: current temperature, yes or no, serve or suppress. The same logic applies to an expanding universe of signals — time of day, device context, conversation history, prior purchase signals, and dozens of others.
Advertising at the Speed of Thought
Andrea’s framing of this capability goes beyond targeting precision into something more strategic: the ability for brands to advertise at the speed at which the world is actually moving. The best marketing teams, in his view, are not the ones with the largest budgets or the most polished creative. They are the ones that can move fast — fast enough to respond to cultural moments, fast enough to match the speed of their own product development, fast enough to learn from one campaign and apply that learning to the next before the moment has passed.
“The best marketing teams are able to advertise at the speed of thought. At the speed of product development. At the speed of news. At the speed of culture. Real-time active decisioning allows for that.”
Real-time decisioning in an AI ad context is what makes this possible. When the decision about what ad to serve is made at the moment of the conversation turn — not hours earlier in a planning session or a bidding auction — brands can respond to signals that traditional ad infrastructure would never see in time to act on. A news event, a product launch, a cultural moment: the gap between the signal and the response can collapse to seconds.
Which Brands Will Win in AI Advertising
Speed and Agility as Competitive Advantages
The brands that will establish early leadership in AI advertising are not necessarily the largest ones or the ones with the most sophisticated existing ad operations. They are the ones with the organizational capacity to move quickly — to test a new format, learn from the results, and iterate before competitors have finished debating whether to try it at all.
Andrea is direct about this: the AI advertising landscape is being shaped right now, and the window for early movers is open. The brands that treat AI advertising as a channel to eventually explore, once it is more proven and more mainstream, will find themselves competing against incumbents who have already accumulated months or years of learning, optimization, and audience relationships. In advertising, as in most technology transitions, the learning curve itself is a competitive asset.
Closeness to the Customer
Speed alone is not enough. The brands that win will also be the ones that remain genuinely close to their customers — not in the sense of having rich first-party data files, but in the sense of actually understanding what their customers are trying to accomplish, what language they use to describe their problems, and what kinds of commercial interventions feel helpful rather than intrusive in a conversational context.
In-chat advertising surfaces this kind of customer understanding in real time and at scale. Brands that pay attention to what the data is revealing — which conversations are converting unexpectedly, which signals are predictive of purchase intent, which messages land and which fall flat — will compound that knowledge into an increasingly precise and effective advertising operation. Brands that treat AI advertising as a media buy rather than a learning opportunity will miss the deeper value entirely.
“The brands that are going to win are the ones that move. Agile in adopting new trends, close to their customers, learning from their customers.”
The Unexpected Conversion: What the Data Is Already Showing
One of the more striking findings from Thrad’s early operational experience is that conversions are occurring in conversations where brands would not have predicted them. The conventional wisdom in ad targeting is that you serve ads when explicit intent signals are present — when someone searches for a product category, visits a relevant website, or exhibits purchase-adjacent behavior.
What AI conversation data is revealing is that intent is often embedded in context that has nothing overtly to do with purchasing. A user discussing a packed calendar and too many meetings is not explicitly shopping for a productivity tool. But the signal is there — and an AI system sophisticated enough to read it can surface a relevant commercial message at a moment when the user is more receptive than any demographic or behavioral proxy would have predicted.
This is the genuine frontier of what Thrad is building toward: an advertising system that understands not just what users are saying, but what they need — and that can connect that need to a brand solution before the user has even consciously recognized the connection themselves. The distance between that capability and where advertising has historically operated is considerable. The infrastructure to close that distance is what makes Thrad’s position meaningful.
About Thrad.ai
Thrad is the advertising infrastructure for the AI era. The platform enables AI apps to monetize through in-chat paid media placements and gives brands a direct channel to reach millions of users daily inside LLM-powered applications — at the moment of peak intent, inside the conversation itself.
Thrad operates exclusively in paid media, not SEO or AI-organic strategies. Its API integrates directly with LLM infrastructure, enabling real-time ad decisioning based on conversational context, task signals, and behavioral data surfaced within the chat. Every impression is specific to the individual user and the specific moment in their conversation — a level of relevance that legacy ad formats cannot replicate.
As AI systems become increasingly agentic and conversational interfaces replace traditional search for a growing share of daily user activity, Thrad is building the infrastructure layer that connects that attention to the brands and businesses that can serve it. Visit thrad.ai to learn more.
About StartupSword.com
StartupSword.com is an editorial platform publishing candid, experience-first conversations with the founders, operators, and builders shaping the next generation of business. This white paper is part of the Entrepreneurship & Innovation Series, which profiles practitioners with a track record of doing the work — not just talking about it.