Conversational Ad

What is a conversational ad?

A conversational ad is a paid placement formatted as a dialogue. Instead of a banner the user scrolls past or a video pre-roll they skip, the ad takes the form of an answer to a question, a suggested next step in a conversation, or a branded response integrated into a chat flow. The user experience of reading the ad looks and feels close to the user experience of reading an organic message from the platform.

The format has two distinct generations, which are worth naming clearly because the industry often conflates them.

Legacy conversational ads (2017-2019) lived inside messaging apps. Facebook launched Messenger ads that opened into chatbot-driven conversations. WhatsApp rolled out business messaging with sponsored replies. Brands built pre-scripted branching flows: user taps the ad, a chatbot greets them, the bot offers a product recommendation or lead-capture form, the user responds, another branch unfolds. The technology behind the dialogue was usually a rule-based chatbot, sometimes a narrow NLU model. Measurement was click-to-conversation and conversation-to-conversion.

New-generation conversational ads (2024-2026) live inside LLM-powered answer engines. When a user asks ChatGPT or Gemini or Perplexity a question, the answer can include a sponsored recommendation that was selected based on the conversation's context, not a keyword match. The ad is generated or inserted within the AI's natural response, often labeled as sponsored but visually close to the rest of the answer. The buyer is usually not scripting a branching flow; they are contributing creative and targeting criteria to a model-driven matching system.

Both share the format (dialogue-based, conversational tone, integrated placement). They differ in the substrate (messaging apps vs AI answer engines), the matching mechanism (keywords and user attributes vs conversational context), and the scale of reach (bounded by a single app vs potentially all AI-search queries).

How conversational ads work

The new generation has four moving parts that marketers should understand at working depth.

Inventory

The AI platform decides what share of responses can contain ads. Platforms make conservative choices early (ChatGPT launched with a small share of responses containing ads) because over-monetizing too fast damages trust in the answer. Expect the share to grow gradually.

Matching

When a user's query arrives, the platform decides whether an ad is eligible based on the conversation's topic and intent. This is closer to contextual advertising than to keyword advertising; the matching signal is the meaning of the conversation, extracted by the LLM, rather than a set of exact-match keywords the advertiser bid on.

Creative unit

Creative is typically a short recommendation, a link with an adjacent paragraph of rationale, or a card inside the response. Length and format are constrained by the platform. Because the creative appears next to organic answer content, there is a strong pull toward informative copy over hard-sell copy.

Measurement

Impressions, clicks, and conversions are all still measured. New measurement questions unique to the format include whether the user engaged with follow-up questions about the ad, how often the ad was referenced later in the same conversation, and how the ad's presence affected completion rates on the main task. Platforms will evolve these measurement layers over the next several product cycles. The first conversational-ad-platform-native pixel came from OpenAI: its ChatGPT Ads conversion pixel and a server-side Conversions API launched broadly on May 5, 2026 as a self-serve beta, tracking 10 events with a 30-day attribution window, backed by the April 30 privacy policy that formalized the data-flow disclosures (including a renamed “marketing partners” vendor category for cross-context behavioral advertising).

Legacy vs new-generation conversational ads

Major conversational ad platforms

• ChatGPT Ads - OpenAI's ad platform, launched February 9, 2026, with Criteo as first ad tech partner. The flagship new-gen conversational ad product. Self-serve Ads Manager went live April 10, 2026 and opened to all US advertisers on May 5, 2026 with no minimum spend (was $50K from April 13 to May 5; was $200K at launch); CPM ranges $25-$60 depending on access path; CPC bidding live with $3-$5 bid floors by category.
• Google Gemini sponsored responses - Google is integrating sponsored content into Gemini answers and AI Overviews, leveraging the Google Ads platform's existing advertiser base.
• Microsoft Copilot ads - Microsoft's Copilot integrates sponsored results drawn from the Microsoft Advertising platform, making it the easiest on-ramp for existing Bing Ads buyers.
• Perplexity sponsored answers - Perplexity has experimented with sponsored questions and brand partnerships; the ad surface is still evolving in 2026.
• Legacy Messenger, WhatsApp, and SMS ad products - Meta's messaging platforms continue to offer conversational ad formats for marketers who already run campaigns on Facebook or Instagram.

Why conversational ads matter

The format matters because AI answer engines are becoming a dominant interface for category research. If a buyer is using ChatGPT to decide between vendors, the placement opportunity inside that answer is at least as valuable as a paid slot at the top of a Google search results page was 10 years ago. That is the strategic case. Three tactical implications follow.

Intent quality is high

Users who trigger a conversational ad response are typically in active consideration: they have articulated a question, they are reading an answer, they are likely to act on the answer. Conversion rates in early ChatGPT Ads data are reported to exceed traditional search rates by multiples, though the sample sizes are still small.

Creative demands writing

Conversational ads cannot be a headline and a URL. They have to read like useful sentences. Marketing teams without strong in-house writing will struggle to produce creative that performs. The shift of advertising budget into conversational formats is also a shift toward marketing teams that can produce short, informative copy on demand.

Organic baseline affects paid performance

A brand with no organic AI visibility paying for conversational placement tends to feel like a cold-call. Users see the sponsored recommendation; they check the surrounding organic answer; the brand is absent from the organic answer; the ad loses credibility. Paid and organic work together in this format even more than in traditional search.

Common misconceptions

Conversational ads = ChatGPT Ads

Conversational ads is the category; ChatGPT Ads is one product. Gemini, Copilot, Perplexity, and legacy Messenger and WhatsApp formats are all conversational ads. A strategy that treats the category as only ChatGPT misses the rest of the surface.

Users dislike conversational ads because they feel intrusive

Early user research on new-gen conversational ads is actually more favorable than banner or pre-roll ads, provided the ads are informative and clearly labeled. Users dislike ads that interrupt; conversational formats have the advantage of being adjacent to the answer the user wanted in the first place. The execution risk is ads that hallucinate or misrepresent; the format risk is lower than prior ad types.

Keyword targeting still works

It works at a narrow level but is not the main matching signal. The LLM is extracting intent from the conversation, which is often broader and more nuanced than a keyword. Advertisers who move to conversational formats with a keyword-first mindset will miss inventory. Context-first thinking is the adaptation.

Frequently asked questions