You Can Now Buy Ads Inside ChatGPT, With No Minimum Spend

In plain terms: OpenAI now lets any US business buy ads that show up inside ChatGPT, with no minimum budget. As of this month, those ads can be set to chase actual sales, not just clicks. This is a brand new place to advertise where most of your competitors are not yet, and below is how to think about testing it without wasting money.

There is a new place to run ads, and almost nobody you compete with is there yet.

Any business in the US can now buy ads that appear inside ChatGPT, the assistant hundreds of millions of people open every day. There is no minimum spend. You do not need an agency. You can start with a test budget the size of a nice dinner.

That alone would be worth knowing. The part that landed this month makes it more interesting: the ads can now be set to optimize for conversions, which means OpenAI tries to show them to people likely to actually buy or fill out a form, not just rack up clicks (Search Engine Land). That is the difference between paying for attention and paying for outcomes.

What actually changed

A quick timeline, because the headline hides the useful detail:

• The self-serve ad manager opened to all US businesses, and the minimum spend dropped to zero. It started at a quarter million dollars, fell to fifty thousand, and is now nothing.
• This month, conversion-optimized campaigns went live for advertisers who install OpenAI's tracking pixel or connect its Conversions API (OpenAI docs).
• The pilot is expanding country by country. It is already running in Canada, Australia, and New Zealand, with the UK as the first European market (PPC Land).

The honest catches

This is the part most coverage skips.

1. Paid ChatGPT subscribers never see the ads. Ads only show to people on the free and lower tiers. So the most affluent slice of ChatGPT users, the ones paying every month, are invisible to your campaign. Depending on who you sell to, that either matters a lot or not at all.
2. Conversion tracking is a real setup step. To optimize for sales you have to install OpenAI's pixel or wire up its Conversions API on your site. That is maybe an hour of work, or a quick job for whoever runs your website. It is not a five-minute switch, and skipping it means you are flying blind.
3. Nobody knows yet whether it converts for small or local businesses. Prices are low partly because this is unproven. That is the opportunity and the risk in the same sentence.

What to do with this

The move is a small, measured test, not a big swing.

Because the channel is new and uncrowded, a modest budget can teach you a lot cheaply. Set a number you would not miss, point it at one clear offer, install the tracking first so you can actually see whether it worked, and give it a couple of weeks. If it converts, you found a cheap channel before everyone else did. If it does not, you spent a little to learn that before your competitors spent a lot.

The thing not to do is treat it as a sure thing because it is new and shiny, or pour real money in before the tracking is in place to tell you what happened.

New channels like this are exactly where a smaller business can get an early edge, but only when the measurement is set up first. Wiring that up so a test actually tells you something is the kind of quiet plumbing that makes the difference between a real experiment and a guess.