We run your conversational AI ad campaigns

Advertising that runs inside AI assistant conversations instead of search results or social feeds. When a buyer asks ChatGPT, Gemini, or another assistant a question, the answer can include a native sponsored card. You buy that placement against the intent in the prompt, not a keyword or an audience profile. It captures demand the moment it emerges, while the buyer is still describing what they want.

Three things end-to-end. Strategy: we define your conversational funnels, intent clusters, and KPIs using mapou's category benchmark data. Execution and optimization: we configure the ads manager, engineer your knowledge set, build proof-led sponsored card creative, manage bids and pacing, and watch citation alignment between the organic answer and the paid card. Reporting and insights: a live app.mapou.ai workspace with AI Discovery and AI Media Planner, plus a monthly strategy review. We run as an extension of your media team, not another tool to staff.

ChatGPT is where paid placement is live today. OpenAI opened its self-serve Ads Manager in May 2026 with CPC bidding, a tracking pixel, and a Conversions API. We build every campaign so it ports to Gemini, Claude, and the rest of the AI open web as their ad inventory opens, because the underlying work, intent mapping and entity authority, is the same across engines. The dashboard already measures your visibility across all four engines today.

You do. The ChatGPT Ads Manager account is yours and you pay OpenAI directly, with no markup; mapou works in it with manager access. You set a monthly cap. Inside it, we move spend between intent clusters each week toward what performs, and every change is logged in your workspace. Raising the cap always needs your sign-off.

Every brand starts with a six-week pilot at a fixed $7,500: two weeks of setup (category read on the mapou panel, intent mapping, account setup, tracking, creative) and four weeks live. Our first five clients pay $5,000 in exchange for a published case study. If the campaign is not live by day 14, the fee comes back. After the pilot, ongoing management is $4,000 a month, with half the pilot fee credited to the first month. Ad spend is paid directly to OpenAI by you, not marked up by us. Media is bought on CPC, with CPA bidding to follow as the platform ships it.

Before kickoff we agree one number to judge it by, such as cost per click or cost per sign-up, plus a target, and we track how often the AI names you on the same questions without the ad. At week six you pick one of three: keep going on the monthly retainer, run four more weeks on new questions if the signal is promising but thin, or stop. If you stop, you keep the category read, the ad account and the creative.

Native sponsored cards inside the AI answer. They read as part of the conversation, so the copy has to earn trust the same way the assistant's answer does: proof-led claims backed by third-party reviews, ratings, certifications, and citation provenance. Display-style or pattern-interrupt creative underperforms here. OpenAI starts testing visual ads in ChatGPT's image-generation experience in October 2026 with a first group of US advertisers; they sit beside the generated image, clearly labelled. The gallery on this page shows illustrative examples across travel, finance, wellness, and beauty.

AI Discovery: which prompts your brand wins and loses across ChatGPT, Gemini, Claude, and Grok, and which competitors are appearing in your place. AI Media Planner: prompt clusters scored by commercial intent, mapped to the campaigns we run. You see exactly what the campaign is bidding on and why, with the same benchmark data we use to make decisions.

Free-tier and ChatGPT Go ($8/month) users. ChatGPT Plus ($20), Business, and Enterprise tiers remain ad-free. The rollout started in February 2026 in the US, expanded to UK, Brazil, and Japan in May 2026 with more markets to follow.

Wired ran 500 prompts on the ChatGPT mobile app and reported that roughly one in five questions in a new conversation thread triggered a sponsored card (Reece Rogers, Wired, 2026). Travel and shopping prompts trigger most often. Separately, Profound found that across 260M prompt-level results only about 6% trigger the shopping surface reliably, and eligibility is a stable property of the prompt. Which is exactly why prompt-level intent mapping is the whole game.

When the AI cites your brand as the organic recommendation and your sponsored card appears alongside saying the same thing, trust compounds and conversion rate rises. We build the citation foundation first using app.mapou.ai's AI Discovery data, then time paid activation to match the authority already established.

AI Share of Voice (the percentage of category prompts where the brand is cited or appears in a sponsored card across ChatGPT, Gemini, Claude, and Grok), citation alignment between organic answer and paid card, the platform pixel and Conversions API for purchase and sign-up tracking, and prompt-to-purchase attribution from AI-sourced clicks. OpenAI is adding support for attribution partners such as Kochava, so ChatGPT results can sit inside the web and app measurement you already use rather than in a separate silo. Impressions are counted on the IAB viewability standard (at least 50% of the card in view for at least one second), so reported reach means cards a buyer could actually read. CPC and CTR lead the report, with CPM alongside.

OpenAI has brand-suitability partnerships with DoubleVerify and Integral Ad Science, so you can verify where your ads ran with the same vendors you use on other channels. It is also adding exclusion targeting, which lets you keep campaigns out of contexts that don't fit the brand. We set your exclusions at launch, review the suitability reports weekly, and tighten them as the campaign learns.

Fourteen days from kickoff to first campaign live. Days 1-3 audit and intent mapping; days 4-7 account setup, knowledge-set engineering, and creative; days 8-12 pilot launch and citation-alignment QA; days 13-14 attribution and 90-day roadmap. Then four weeks live inside the six-week pilot, and weekly optimization under the retainer after that.

Technically yes, in practice no. A sponsored card that contradicts what the AI says organically destroys trust faster than no card at all. We always start with a baseline citation audit so paid placement and organic answer point at the same brand story. That audit is included in setup.