Campaigns on OpenAI's advertising platform — context hints, conversion tracking and product feeds. Below, an honest explanation of how ChatGPT ads work, what they now cost, and whether your business should be testing them yet.
A straight answer on whether it fits your business. No obligation.
Check your inbox for a confirmation from info@adzlance.com.
The basics
Sponsored placements that appear below a ChatGPT response, matched to what the conversation is about. Clearly labelled, and shown only to free and Go tier users — never to Plus, Pro or Business subscribers, and never to anyone ChatGPT believes is under 18.
For the full picture, read our complete guide to ChatGPT Ads.
A sponsored card sitting underneath ChatGPT's reply, typically showing a product or service with name, price and a clear sponsored label. It doesn't interrupt the conversation — which also means it has to be genuinely relevant or it is ignored.
Paying subscribers on Plus, Pro and Business plans see no ads at all. That shapes the audience: consumer and small-business users rather than the enterprise seats.
A second-price auction that weighs relevance, not just bid. Buying your way to irrelevant placements does not work — the system is built to suppress ads that don't fit the moment.
Targeting
This is the part that catches out experienced media buyers. There are no keywords to bid on and no audience profiles to build.
The practical consequence is uncomfortable for anyone used to buying reach: if your product is only a loose fit for the questions people ask, no amount of budget fixes that. The brands that win are the ones that are genuinely the right answer to a specific question.
What it costs
Most articles about ChatGPT ads still quote the launch pricing, which is badly out of date. Here is the actual trajectory.
Reported: ~$60 CPM, ~$200,000 minimum
Reported: minimum drops to ~$50,000
Result: genuinely reachable for mid-sized advertisers
Means: spend can be justified against outcomes, not exposure
Means: e-commerce can scale without hand-building ads
Straight assessment
Most agencies selling ChatGPT ads services today have never run a campaign on the platform. Here is the honest position.
You sell something people research conversationally — software, considered purchases, services with comparison shopping — and you already have Google and Meta running profitably. This is an addition, not a replacement.
Your existing channels aren't yet measured properly. Adding an immature platform to a stack you can't measure produces noise, not insight. Fix tracking first.
You need predictable volume this quarter. Reporting is still thinner than Google or Meta, inventory is limited, and the platform changes month to month. This is early-stage buying, not a growth channel yet.
The service
The work is different from Google or Meta management, because the levers are different.
The core skill on this platform. Writing scenario descriptions precise enough to reach the right conversations and broad enough to get volume — then refining them against what actually delivers.
Installing and validating OpenAI's conversion pixel, so campaigns can be bought on CPC or cost-per-action rather than impressions. Without this you are buying blind.
A sponsored card below an answer has to feel useful. Outcome-oriented copy, the customer's language rather than internal product taxonomy, and concrete specifics — price, timing, delivery — that reduce decision friction.
For retail: structuring the catalogue feed that powers automated ad generation, with the same attention to titles, pricing and availability that Shopping feeds demand.
Understanding what ChatGPT actually contributed rather than what it claimed, measured against your Google and Meta performance rather than in isolation.
Formats, pricing models and available markets are all still moving. Part of the job is noticing when something changes and adjusting before it costs you.
Questions
