Frequently Asked Questions (FAQ)
Thrad FAQ
Everything you need to know about running paid ads inside AI conversations, and monetising AI products with Thrad.
About Thrad
What is Thrad?
Thrad is the advertising infrastructure for AI. We connect advertisers who want to reach users inside LLM apps with AI publishers who want to monetise their product without charging users. Our engine analyses prompts in real time, scores intent, and deploys the most relevant ad natively inside the AI response.
What does "paid ads in AI" actually mean?
Instead of buying ads on search engines, social feeds, or display networks, advertisers reach users at the moment they are asking an AI assistant a question. When a user asks a chatbot for running shoe recommendations, a running brand can appear natively inside the answer. The user gets a relevant suggestion: the brand gets a high-intent touchpoint.
Who is Thrad for?
Two audiences:
Brands and agencies that want to reach high-intent users inside AI conversations, whether performance marketers, programmatic buyers, or direct advertisers.
AI publishers: anyone building a chatbot, AI assistant, AI search product, or LLM-powered app who wants to monetise without a paywall.
Where is Thrad based?
Thrad is headquartered in San Francisco, with a global team and clients across North America, Europe, APAC, and the Middle East.
Who founded Thrad?
Thrad was founded by Andrea Tortella who was doing advertising at Perplexity and Marco Visentin was doing his PhD in machine learning at Imperial.
How is Thrad different from Google Ads or Meta Ads?
Google and Meta were built for a world where users browse search results and social feeds. Thrad is built for a world where users talk to AI. The targeting signal is not a keyword or a profile: it is the intent expressed in a natural-language prompt. This means higher relevance, higher conversion, and lower waste.
What is Thrad focused on?
Full-stack infrastructure: Thrad runs both the supply and demand side through one engine, plus an SSP with programmatic open bidding.
Native ad formats: Thrad deploys endemic ads that match the publisher's UI, not banner-style insertions. Two times the engagement of standard formats.
Speed to launch: campaigns go from idea to live in seconds, with AI-optimised creatives generated on the fly.
Privacy by design: AI conversations are deeply personal, and trust is the foundation of this business. Thrad targets on real-time prompt context, not cross-site tracking, cookies, or third-party profiles, and we never retain conversation content beyond what is needed to serve a relevant ad.
Thrad also serves millions of ads daily to Fortune 500 clients across retail, finance, media, and enterprise software.
For Advertisers
How do I launch a campaign?
Sign up at platform.thrads.ai, upload your creative or let our AI generate one, set your budget and targeting, and go live.
What does the Thrad platform look like?
A self-serve dashboard with campaign builder, live performance metrics (CPC, CTR, ROAS, conversions), audience intelligence from prompt data, and creative management. Managed service and design partner tiers are available for advertisers who want hands-on support.
Is there a minimum ad spend?
No, Thrad has no minimum spend.
What is the pricing model: CPC or CPM?
Both are available. Choose the model that best fits your campaign goals. CPM is generally best for reach, CPC tends to be best for performance.
What inventory do you have?
Thrad serves ads across a growing network of AI publishers, from AI search engines and chatbots to vertical AI assistants, reaching millions of daily conversations. Inventory access is available on request.
How is targeting done?
Targeting is contextual and intent-based. Instead of building audience segments from cookies or third-party data, Thrad uses the user's context in real time and matches it against advertiser campaigns based on topic, intent, and commercial relevance.
What ad formats does Thrad support?
Native in-chat ads that blend into the AI response, carousel formats, image formats, and CTA-driven suggestions. All formats are endemic to the publisher's UI, meaning they match the look and feel of the chat interface.
How fast can a campaign go live?
Campaigns can be live in seconds. This is one of the fastest launch times in the industry and is one of Thrad's core advantages versus traditional programmatic workflows.
What results can I expect?
Typical performance across our network:
CTR: 1 percent to 5 percent, depending on category and format
CPC: From 0,5$
CPM: From 10$
Exact benchmarks vary by vertical and creative.
How do you measure attribution?
Thrad provides pixels and UTMs. Click tracking, impression tracking, and post-click conversion tracking via pixel or postback. For multi-channel advertisers, we offer multi-channel attribution through our Addendum framework. Live dashboards show CPC, CTR, ROAS, and conversion data in real time.
How do you ensure brand safety?
Every prompt is analysed before an ad is served. Ads are blocked from appearing alongside sensitive or unsafe content: politics, violence, adult, medical emergencies, and other excluded categories. Advertisers can set category-level exclusions, and publishers can set their own brand safety rules. We maintain a full safety policy on our website.
Can I get case studies?
Yes all our case studies are public and available on our website. We publish case studies across multiple verticals, including travel, finance, media, e-commerce, and enterprise software. Visit thrad.ai/case-studies.
For Publishers
Why should I monetise with Thrad instead of charging users?
Subscription models alone do not scale. 95% of users do not pay for AI and advertising subsidises tokens. Advertising unlocks a revenue stream from your entire user base without gating access. Thrad lets you earn from conversational traffic the way web publishers earn from page views.
How does integration work?
Thrad provides a lightweight API or SDK. Integration is typically a few lines of code and can be completed in a day. We support custom UI, ad frequency controls, category filters, and data privacy configurations.
How long does integration take?
The itegration can be done in minutes. To be in production, most publishers are live within 24 to 72 hours. Larger enterprise integrations with custom requirements can take longer, and our team provides hands-on support throughout.
What integration models do you offer?
We offer multiple ways to integrate, depending on the publisher's preference and how deeply they want to embed Thrad:
Self-serve API or SDK: lightweight, fast to integrate, standard revenue share, full control over ad frequency and categories.
Managed integration: hands-on support from our team, custom UI work, optimised configuration, and dedicated account management.
Design partner: deeper commercial and product collaboration, guaranteed minimums, co-development of new ad formats, and joint go-to-market.
The right model depends on your traffic volume, technical capacity, and how much of the ad experience you want to own versus delegate to Thrad.
Do I control which ads appear?
Yes. Publishers set category blocks, competitor exclusions, ad frequency caps, and presentation rules. You can preview all ads that run on your platform through the Thrad dashboard.
Will ads slow down my app?
No. Thrad is engineered for near-zero latency so ad delivery does not degrade the user experience. The engine runs in parallel with the LLM response.
When do publishers get paid?
Net 30 from the end of each month, with transparent reporting on impressions, clicks, and earnings inside the dashboard. Payouts via bank transfer or supported payment methods.
What is the Monetisation Fund?
Thrad operates a $500K Monetisation Fund for qualifying AI publishers. The fund provides guaranteed revenue, technical integration support, and co-marketing for publishers who join as early partners. Details at thrad.ai/monetisationfund.
Who owns the user data?
You do. Thrad does not store conversation content beyond what is required to serve the relevant ad. Publishers retain full ownership of their user relationships and conversation history. See our Privacy Policy for details.
Technology
How does Thrad analyse prompts?
Our engine processes context in real time, extracts intent signals, ranks available advertiser campaigns against those signals, and selects the most relevant ad via real-time bidding. The full pipeline runs in milliseconds.
What LLMs does Thrad support?
Thrad is LLM-agnostic. We integrate with products built on OpenAI, Anthropic, Google, Mistral, open-source models, and custom proprietary stacks. The integration sits above the model layer.
API or SDK?
Both. Use the API for web-based or server-side integrations, and the SDK for mobile, desktop, and embedded environments. Full documentation is at docs.thrads.ai.
Is Thrad open source?
Parts of our stack are open-sourced on Hugging Face at huggingface.co/Thrad, including research models and reference implementations. The core ad engine is proprietary.
What is the Thrad SSP?
Our Supply-Side Platform enables programmatic open bidding across the Thrad publisher network. Agencies and DSPs can integrate directly via standard protocols to buy AI conversation inventory alongside their existing media buys.
Privacy and Safety
How is user privacy handled?
Thrad processes prompts for ad matching in real time and does not retain conversation content beyond what is needed for the transaction. We comply with GDPR, CCPA, and other applicable data protection regulations. Publishers control what data is shared with the engine.
Do ads affect AI responses?
No, the ads are seperate from the AI responses and don't interfere with the LLM output. The ads are always clearly labelled and distinct from the AI answer.
Is Thrad GDPR compliant?
Yes. Thrad operates under GDPR-compliant data handling for all European traffic, with full DPA support for publishers and advertisers.
How do you prevent ads from appearing in sensitive conversations?
Every prompt passes through a safety classifier before ad selection. Conversations flagged as sensitive (politics, health emergencies, self-harm, violence, adult content, and other excluded categories) do not trigger ad delivery. Publishers can expand these exclusions.
Does Thrad use personal data to target ads?
No. Thrad uses the current prompt and session context for targeting, not cross-site tracking, cookies, or third-party profiles. This is a key privacy advantage over legacy ad networks.
Getting Started
How do I book a call?
Visit thrad.ai and click Book a Call, or email.
How do I try the platform without a call?
Go to platform.thrads.ai and sign up. You can launch a first campaign or explore the publisher integration with no commitment.
Where can I see recent news about Thrad?
Visit thrad.ai/newsroom for recent announcements, including our AWS agentic architecture case study, BCG partnership report, and CEO interviews at WSJ Tech Live Doha and ETN.
