# Conversational Ads: Messaging vs AI Answer Placements

Conversational ads now means three things: click-to-message, advertiser-built message units, and labelled sponsored placements beside AI answers. Here is how each works.

# Conversational Ads: Messaging vs AI Answer Placements

Conversational ads covers three different products: click-to-message campaigns that open a chat you run, advertiser-built message flows, and labelled sponsored placements beside an AI assistant's answer. Tell them apart by who writes the conversation, then match the format to your team's workload and where your buyers are.

"Conversational ads" now names three different products: ads that open a chat with your business, ads that are themselves a branching message, and labelled sponsored placements beside an AI assistant's answer. They share a name and almost nothing else, because each one asks you to brief, build, buy and measure differently.

That matters if you searched the term to buy something or to sell inventory. A paid-media lead who books "conversational ads" expecting a WhatsApp chat flow will be surprised by a sponsored card in an AI app, and the reverse is just as true. This guide separates the three, shows what each workflow demands from you, and then explains where Jolts fits, which is only in the third.

## What "conversational ads" means now: three things sharing one name

Conversational ads covers click-to-message campaigns, advertiser-built message units, and sponsored placements beside AI answers. The useful way to tell them apart is to ask who writes the conversation.

In click-to-message, you write nothing inside the ad. The ad is a button, and the conversation happens afterwards in a chat your team or bot runs. In a conversation ad unit, you author the message itself as a branching flow, and the platform delivers it. In an AI answer placement, the assistant writes the answer, and the advertiser supplies a separate, labelled sponsored unit that appears next to it.

You will also see "conversational advertising" contrasted with "conversational marketing." The first is a paid format: you buy an impression or a click that leads into a dialogue. The second is the broader practice of talking with customers in real time across site chat, messaging and support, paid or not. Conversational ads are one way to feed conversational marketing. They are not a synonym for it.

The rest of this guide takes the three formats in turn, then compares them.

### What did conversational ads marketing mean in 2024?

In 2024 the term mostly meant two older formats: click-to-message campaigns that open a chat in WhatsApp or Messenger, and advertiser-scripted message flows such as LinkedIn Conversation Ads. The third meaning, sponsored placements beside AI answers, was only emerging. When you read 2024 material, check which of the three it describes, because the buying model, creative work and measurement differ for each. Jolts focuses on the AI answer placement type.

## Click-to-message campaigns: the ad opens a chat you then run

A click-to-message campaign is a social ad whose call to action opens a chat with your business, typically in Messenger, Instagram Direct or WhatsApp. The ad's job ends the moment the chat opens.

In a typical workflow, the ad invites a conversation in a social feed, the customer taps the button and lands in a chat, and the brand uses what it learns to personalize recommendations and follow-ups.

**What you brief:** a goal such as inquiries, leads or support conversations, plus an audience.

**What you build:** the chat itself. That means greeting copy, qualifying questions, handoff rules to a human, and response-time targets. The ad creative is the easy part. An ad that opens an unanswered chat wastes the click.

**What you buy:** the ad impression or click on the social platform, on that platform's terms.

**What you measure:** conversations started, and what happens inside them: qualified leads, bookings, sales. You own the downstream data because the conversation lives in your messaging account.

The common mistake is treating this as an ad project. It's an operations project with an ad on the front. Staff the chat before you scale the spend.

## Conversational ad units: advertiser-built message flows (e.g. LinkedIn Conversation Ads)

A conversational ad unit is a sponsored message the advertiser designs as a branching flow. LinkedIn Conversation Ads are the clearest example: a sponsored message lands in a prospect's LinkedIn inbox and offers buttons, and each button can open a form, send the person to a page, or show a follow-up message with more choices.

The ad is the conversation. Nothing happens afterwards in a separate chat unless you design it that way.

**What you brief:** a B2B-style objective such as awareness, website visits or lead generation, and a sender identity that makes an inbox message feel legitimate.

**What you build:** the tree. Write the opening message, two or three button choices, the follow-up for each branch, and the end state (a form, a page, a thank-you). Keep branches shallow. Every extra level is a place for the reader to drop out.

**What you buy:** a send. LinkedIn charges by send, and the message carries a Sponsored label.

**What you measure:** opens, button clicks per branch, form completions and site visits. Because you wrote the tree, you can see which branch worked. That's the real advantage of this format.

The flow design rules are the same whether you're building an inbox unit or a chatbot: lead with one question, give choices instead of open text, and always offer an exit to a human or a page.

## Sponsored placements beside AI answers: the ad sits next to the answer, labelled and separate

A sponsored placement beside an AI answer is a paid, labelled unit that appears next to the assistant's response while the answer itself stays organic. It is the only one of the three where the advertiser neither writes the conversation nor opens a new one.

The person is already mid-conversation, asking an AI app to compare products or decide what to buy. A sponsored card, message, carousel or optional poll appears alongside the answer. The user did not click an ad to get there, and the advertiser is not running a chat afterwards. Platforms are moving in this direction broadly. Microsoft's advertising team describes [conversational AI changing how consumers find information and interact with brands](https://about.ads.microsoft.com/en/blog/post/may-2025/captivating-and-relevant-how-conversational-ai-is-changing-advertising), and Google has a separate [chat-based feature inside Google Ads](https://support.google.com/google-ads/answer/14145186?hl=en) that helps advertisers build Search campaigns. That Google feature is a tool for advertisers, not an ad format shown to users, so don't confuse it with the three formats here.

**What you brief:** the products and audience you want to be considered for, the contexts where your offer fits, and the contexts where it does not. The customer's question is the creative brief, as we argued in [The customer question is the creative brief](https://jolts.xyz/blog/the-question-is-the-brief).

**What you build:** the sponsored unit and its landing destination, not a conversation tree.

**What you buy:** CPM or CPC, depending on the network. The terms vary by platform and publisher.

**What you measure:** impressions, clicks and what happens after the click. Be careful with causal claims. [A click can be real without being incremental](https://jolts.xyz/blog/clicks-attribution-and-lift).

Trust defines this format. The placement must read as sponsored and stay visually and logically separate from the answer. [A sponsored offer should have its own place](https://jolts.xyz/blog/sponsorship-is-not-an-answer) covers why.

### Do AI chat platforms like ChatGPT run sponsored answers?

The model discussed here keeps the sponsored unit separate from the organic answer and labels it clearly, rather than rewriting the answer itself. Some platforms manage and sell their own conversational inventory directly. Jolts does not. Jolts works with other AI-app publishers, matching sponsored placements to relevant context and showing no ad when nothing suitable matches.

### Where do Amazon Alexa sponsored products fit in?

Amazon's Alexa-related sponsored products and brand placements are one example of a retailer or platform selling ads inside its own assistant. They run on that platform's inventory and are bought through its own ad tools. Jolts is a different route: it connects advertisers with independent AI-app publishers. Jolts cannot sell Amazon inventory, and this page does not compare reach or performance between the two.

## Side by side: who writes the conversation, what you buy, what you can measure

The three formats differ most on who authors the dialogue, so compare them on that and on what you actually purchase.

| | Click-to-message | Conversation ad unit | Placement beside AI answer |
|---|---|---|---|
| Where it appears | Social feed (Facebook, Instagram) | Recipient's message inbox (e.g. LinkedIn) | Next to an AI assistant's answer |
| Who writes the conversation | You, in the chat after the click | You, as a branching flow inside the ad | The assistant writes the answer; you supply a separate sponsored unit |
| What you buy | Ad impression or click | A send (LinkedIn charges by send) | CPM or CPC, per network terms |
| Main workload | Staffing and scripting the chat | Designing the branch tree | Matching rules, creative unit, landing page |
| Core measurement | Conversations started, then outcomes in chat | Branch clicks, form fills, visits | Impressions, clicks, post-click behaviour |
| Label | Standard ad label on the social ad | Sponsored label in the inbox | Must be clearly labelled and separate from the answer |

Two practical readings follow. If you need leads from a conversation you control, the first two fit. If you want to appear where someone is researching a purchase inside an AI app, only the third does. And if you are an AI publisher, only the third is something you can host.

One more difference: in the first two, your own team sees every message exchanged. In the third, you generally see the sponsored unit's performance, not the user's AI conversation. That's a privacy feature, not a gap.

### How do Alexa shopping prompts compare with other conversational ad types?

Sponsored prompts inside a shopping assistant are a platform-run placement: the platform owns the conversation, the inventory and the reporting, and you buy through its own ad tools. That differs from click-to-message campaigns, where you run the chat, and from advertiser-built message flows. Compare each on who writes the conversation, what unit you buy, and which results you can actually measure before committing budget.

## How Jolts approaches AI answer placements: matching, labelling, CPM or CPC, a scoped pilot

Jolts connects advertisers with AI publishers through relevant, clearly labelled sponsored placements in AI conversations. It operates only in the third format above. It does not run click-to-message campaigns or inbox conversation units.

**Matching.** Jolts matches on relevant context the app shares, user intent, campaign criteria and publisher rules. Full conversations are not required by default. Structured topics or redacted summaries can describe the context. For what that looks like in practice, see [What can an ad matcher learn from four fields?](https://jolts.xyz/blog/context-without-the-transcript).

**Labelling and separation.** Sponsored placements stay separate from the organic answer. The site illustrates messages, cards, carousels and optional polls.

**No match, no ad.** If nothing suitable matches, nothing is shown.

**Buying model.** Campaigns are priced CPM or CPC. Exact rates, counting rules and spending limits are agreed before starting. Managed campaigns and custom integrations have custom pricing, and CPA is not available yet. See [pricing](https://jolts.xyz/pricing).

**A scoped pilot.** Jolts is at an early stage and works with early partners. Getting started begins with a call to confirm audience fit, placement availability, scope and terms. Publishers agree categories, frequency, excluded topics, shared data, revenue share, event counting, fees and payout timing before going live. [Before the SDK: the publisher conversation worth having](https://jolts.xyz/blog/publisher-pilot-before-sdk) walks through that conversation.

If you are comparing networks for this format, [Contextual ads in AI conversations: platforms to compare](https://jolts.xyz/blog/contextual-ads-in-ai-conversations-platforms-to-compare) lays out the criteria.

### What formats and best practices apply to AI answer placements?

The formats shown on the Jolts site include messages, cards, carousels and optional polls, though these demonstrations do not establish availability in every interface. Practical rules: label sponsored content clearly, keep it separate from the organic answer, match to relevant context and intent, set frequency and category limits up front, and show no ad when nothing fits. Start with a scoped pilot and agree how results are counted.

## Which type to start with: a short decision checklist for advertisers and AI publishers

Start with the format whose workload your team can actually carry, then match it to where your buyer is.

**Choose click-to-message if:**
- Your sales or support team can answer chats quickly, during real hours.
- Your buyers already use Messenger, Instagram or WhatsApp.
- You want the conversation and its data in your own account.

**Choose a conversation ad unit if:**
- Your audience is professional and reachable by inbox message.
- You can design a short branching flow and want branch-level reporting.
- A form fill or page visit is the goal.

**Choose an AI answer placement if (advertiser):**
- Your buyers research and compare products inside AI apps before they search or shop elsewhere.
- Your offer can stand as a labelled sponsored unit without a chat behind it.
- You are comfortable with a first campaign that is a learning budget. See [Your first AI-ad campaign needs a learning budget](https://jolts.xyz/blog/first-campaign-learning-budget).

**Choose an AI answer placement if (AI publisher):**
- Subscriptions or usage credits do not cover your costs, and a free tier is a drain. [What should pay for your AI app's free tier?](https://jolts.xyz/blog/funding-the-free-tier) frames the options.
- You can define eligible categories, frequency and excluded topics, and you want a no-ad fallback when nothing fits.

You don't have to pick one. A brand can run click-to-message for lead capture and a placement for research-stage reach. Just keep the briefs separate, because they are different products.

### What are the benefits and use cases of conversational ads?

The benefit depends on the type. Click-to-message campaigns suit lead generation, since they start a chat you can qualify. Advertiser-built message flows suit engagement with a known audience. Sponsored placements beside AI answers reach people while they research, compare and choose products. For AI publishers, the use case is ad revenue alongside subscriptions or a free tier. Pick the use case first, then the type.

## Pick the format by who writes the conversation, then book the call

Conversational ads is not one channel. The cleanest test for any offer is to ask who writes the conversation: you in a chat, you in a message tree, or the AI with a labelled sponsor beside it. If the third one is yours, whether you want to appear where people research and compare, or you run an AI app that needs revenue beyond subscriptions, book a call with Jolts at contact@jolts.xyz. The call confirms audience fit, placement availability, scope and terms before anything goes live.