Plain-English, citation-backed guides written by a media buyer with twelve years of experience. Start with the cornerstone, then drill into the specific surface you're working on this week.
The cornerstone. What context hints are, how OpenAI's relevance-weighted second-price auction scores them, how to write them, 50 worked examples, bulk-upload schema, pricing benchmarks, and common mistakes.
Read the guideHow context hints work in OpenAI ads and exactly how semantic intent matching differs from exact-match keywords — with a side-by-side table and a paid-search translation map.
Read the guideWhere ads appear, who sees them, the ad unit anatomy, eligibility, and the three trust principles OpenAI uses to keep the experience safe. Plain English, no marketing fluff.
Read the guide2026 pricing in plain numbers — CPC and CPM bids, the minimum spend that disappeared, budget scenarios by business size, max-bid strategy, and an honest read on whether it pays back.
Read the guideIs it right for your business? The 2026 self-serve launch and no-minimum reality, a verdict by business type, a lean first-campaign playbook, and when to choose organic instead.
Read the guideThe channel decision — search intent vs conversational intent, a full side-by-side, cost and measurement, a choose-X-if framework, budget-shift triggers, and when to run both.
Read the guideThe first-party-data layer — upload customer lists for inclusion, exclusion, and 0.1x-10x bid multipliers. File rules, the 25,000 minimum, SHA-256 hashing, and five operator playbooks.
Read the guideThe merchant catalog layer — the full-snapshot model, SFTP delivery, Parquet and shard rules, explicit removals, the three ingestion failures, and the phased rollout from sample to daily automation.
Read the guideThe field manual. Frameworks for the audience, intent, and topic layers of a hint, templates by funnel stage, and 30+ worked examples you can copy.
Read the guideAccount creation, Persona verification timing, billing, account info, and team-member permissions. Written for a real first-time advertiser, with the failure modes called out.
Read the guideThe four ways AI search and chat products make money from advertisers — conversational ad units, sponsored placements, product feeds, and agentic checkout — and what you can actually buy today.
Read the guideBuyers now shortlist inside ChatGPT before they visit your site. The comparison-query grammars, the content that gets lifted into shortlists, and the paid backstop at the moment of choice.
Read the guideHow the relevance-weighted auction reads conversational intent, where it beats keywords and audiences, where it fails, and the ad-group structure that turns intent into buyable funnel stages.
Read the guideEvery element of the sponsored unit, explained: the 16-character headline, the 32-character description, the favicon, the image creative — and why OpenAI says not to lead with your logo.
Read the guideThe 30-character headline, the 60-character body, the 256×256 square image, and the single biggest mistake advertisers make when they import a Google Ads voice. With a before-and-after rewrite.
Read the guideNative attribution is limited and CTR is the wrong headline metric. What OpenAI's Conversions API gives you, a matched geo-holdout method, brand-search lift, and the five metrics that survive a finance review.
Read the guideSix months of public spend data — the 15-day $60k case, the pilot's rocky start, the March 2026 volume scale-up, category benchmarks, and a four-question framework for judging any case study.
Read the guideAn honest map of every AI advertising surface in 2026 — ChatGPT, Google AI Mode, Microsoft Copilot, the contested Gemini, and the third-party networks selling into AI apps. With a where-to-start recommendation.
Read the guidePerplexity killed its ads in February 2026 citing trust. Gemini's in-chat ads were briefed to advertisers, then denied publicly. The timeline, the real reasons, and how a serious 2026 media plan should treat both surfaces.
Read the guideThe organic side. Four signals ChatGPT actually uses to pick citations, a four-week playbook for a first cycle, and how to measure whether the work is showing up — without spending a cent on ads.
Read the guideFive metrics, three measurement methods (manual prompting, referral analytics, and paid tools like Profound, AthenaHQ, and Bluefish AI), and a weekly dashboard shape that ties visibility to pipeline.
Read the guidePlain-English definitions for context hint, ad group, Reach objective, Clicks objective, second-price auction, Persona verification, bulk-upload schema, and 25 more terms.
Open the glossaryForty answers to the most common questions about ChatGPT ads, context hints, eligibility, pricing, the auction, measurement, and policy.
Open the FAQ