A Field Guide to Chatbot Ads

Chatbot ads are disclosed sponsored placements that appear inside or alongside a chatbot's response. They are a real category now, with live inventory, active buyers, and standardized formats. This guide is the short version: what they are, who buys them, how to run a campaign, and what it takes to integrate them on the developer side.

What chatbot ads are.

A chatbot ad is sponsored content placed inside a conversational AI product. The placement is labeled as sponsored. The creative is usually a card, a text line, or a follow-up prompt. The payment model is CPM, CPC, or CPA depending on the network and the format.

Chatbot ads are not the same as in-app ads. In-app ads sit in fixed slots — banners, interstitials, rewarded video — inside an app shell. Chatbot ads live inside a conversation, which means the placement changes with the content of each user turn. A sponsored card shown to a user asking about hiking boots is different creative from the one shown to a user asking about flight deals.

For context on where the category stands today, see the state of chatbot advertising in 2026.

The formats.

Five formats ship in production today.

Text inside response. A short labeled line of sponsored text that appears within the chatbot's reply. The label prefixes the content.

Disclosed card after response. A card with a sponsored badge that renders below the chatbot's organic reply. This is the most common format because it keeps the model's output clean.

Sponsored citation. A labeled link in the reference list of a response. Less common, higher disclosure bar.

Sponsored follow-up prompt. A labeled suggested next question. The user taps it, and the brand pays for the click.

Sponsored tool-call suggestion. An agentic placement where the chatbot suggests a paid provider for an action like booking a flight or generating a quote. Still mostly in beta.

Each format has its own pricing, creative spec, and disclosure requirement. Most chatbot developers start with the disclosed card because it is the least invasive.

Who buys chatbot ads.

The buy side is concentrated in five verticals: DTC, travel, SaaS, fintech, and consumer packaged goods.

DTC brands run chatbot ads because the placement fits high-intent product questions. Travel brands run them because users ask chatbots for trip planning help. SaaS brands run them — especially on developer-focused chatbots — because a user asking a technical question is a warm lead. Fintech runs them within regulated disclosure constraints. CPG brands run them for awareness, funded out of brand budgets rather than performance.

Most chatbot ad networks sell on a self-serve basis with a minimum spend, usually a few thousand dollars to start. Buyers can see inventory, bid, and measure results without a managed sales rep at mid-sized budgets. Larger brands get account support. Advertisers can get started on the reach side.

How to run a chatbot ad campaign.

The workflow looks like paid search with conversational targeting.

1. Pick a network. Availability and fit vary by category. Networks with consumer-facing supply are different from networks with developer-tool supply.

1. Set a targeting signal. Most networks support keyword-level targeting (show this card when the user mentions these terms), intent-level targeting (show this card when the user is in a shopping context), or audience-level targeting (show to users who have opted into brand interest categories).

1. Upload creative. A card, a line of copy, a follow-up prompt, or a tool-call suggestion. Character limits are tight — most networks cap cards at 90 characters for the headline and 120 for the description.

1. Set a bid. CPM for brand, CPC for performance, CPA for direct response. Not every format supports every bid model. Tool-call suggestions are usually CPA-only because the intent is so high.

1. Measure. Networks report impressions, clicks, and post-click conversion via pixel or postback. Attribution is weaker than paid search because the conversational flow breaks click paths. Plan for assisted-conversion and lift studies, not last-click purity.

Budgets tend to start small and scale if the CPA holds. Most brands test for two to four weeks before committing ongoing spend.

How developers integrate chatbot ads.

For chatbot developers, integration is a small amount of work and does not require rewriting the app.

The common pattern is an SDK or an API call that runs after the model generates a response. The SDK takes the conversation context and the model's reply as input and returns a sponsored card (if one is eligible) as output. The developer's job is to render the card with the required disclosure and report impressions and clicks back to the network.

Three things matter on the integration side:

1. Latency. The ad call cannot slow down the chatbot's perceived response time. Networks that add more than 200 ms are hard to use. The best ones fire in parallel with the model stream and attach the card after the response renders.

1. Relevance. If the card is unrelated to the user's question, the click rate is low and the user experience suffers. Networks with weak relevance models produce weak revenue.

1. Disclosure. The card must render with a sponsored label. Networks supply the label, but the developer is responsible for placing it visibly. Regulatory exposure sits with the publisher too, not just the network.

Most developers finish integration in a day or two. See monetize an AI app for the implementation path.

What's next.

The category is still early. Three developments to watch over the next year.

Standardization. A creative and disclosure spec is likely to emerge from an industry working group. Once creative is portable, networks compete on yield rather than format, which is good for both brands and developers.

Agentic formats. Sponsored tool-call suggestions will mature. Travel is the likely first vertical to scale this because the booking flow is already structured around tool calls. Fintech and commerce follow.

Measurement. Conversational attribution standards will improve as more spend flows through the category. Expect lift-study products, conversational-intent metrics, and multi-turn attribution to ship from the bigger networks.

Chatbot ads are a real channel now. The brands and developers who build the muscle early will have a head start when the category shows up on the standard media plan.