How To Tell If Competitors Are Inserting Ads In ChatGPT Answers

This post is sponsored by Trendos. The opinions expressed in this article are those of the sponsors.

Are my competitors using ChatGPT ads?

Is there an ad library for ChatGPT sponsored results?

How do I track which advertisers respond to AI?

Your target customers are asking ChatGPT about your category right now.

Sponsored placements appear below the response, and if competitors have purchased them, you grab the click at that point when buyers are ready to decide.

Unless you run all the right commands yourself, opponents underestimate your AI visibility in the most important moments, and you can’t see any of it.

Including This Flow of Things

This is a manual process to find out who is bidding against your category, and where you can see who is buying ads from your ChatGPT customer responses without doing it yourself.

OpenAI launched ChatGPT ads for US users Free and Go on February 9, 2026.

In the spring, 600+ advertisers had the opposite placement and information for the top targets:

• Software comparison.
• Planning a weekend trip.
• “What is the best crm tool?”

These questions were always on Google; now they show inside ChatGPT as ads.

ChatGPT ads appear within the response as a sponsored card under the response.

After ChatGPT responds to the prompt, the sponsored card provides a response below, visually separated and clearly labeled “Sponsored.” The card includes the advertiser’s name, a thumbnail image, a short caption, a strong body description (~19 words on average), and a link to the site page.

OpenAI does not currently publish an ad library equivalent to Meta or Google, and no central searchable database of all active ChatGPT ads exists. To see who is running the ads, you must run the commands on the relevant US sessions and capture the impressions.

For monitoring purposes, four data points describe what a competitor is doing in a particular ad:

• The title of the ad: competing title copy is active
• Description of the ad: the body sentence under the title
• Last URL: where they send traffic to
• Sharing impressions: how often a competitor’s ad appears in a given notification across multiple runs

You need all four to learn the competitive picture.

The title and description tell you how they stand.

The final URL tells you whether it’s sending to the standard homepage, category page, or comparison.

Impression share, the percentage of total ads in a given ad that goes to a particular advertiser, turns “I’ve seen you” into “I own this information.”

Competitive intelligence is more important than raw impression statistics because it is common to all commands with different ad fill rates.

Step 1: Map the Questions Your Customers Are Already Asking

Create a data list that represents how your customers actually talk to ChatGPT. You don’t prepare impressions in broad terms. You highlight the work of competitors in the discussions that lead to your category.

Start with queries that you already know have converted to high-target paid and organic search.

Then translate them to how someone would express the same need in ChatGPT. People don’t search ChatGPT the way they search Google. They write complete sentences with context, constraints and purpose.

An effective command list for a paid search manager in any marketing category should reach 30 to 50 entries and cover:

• Direct comparison (“best [category tool] for [use case]”,”[Brand A] vs [Brand B]”).
• Recommendation recommendations (“I need a [tool] for [job to be done]what should I look for?”).
• Changing the information (“different from [Brand]”).
• Use case qualification information (“which [tool] even better [small team / enterprise / specific industry]”).
• Price information (“affordable [tool] for [audience]”).
• Long edge cases (“[tool] which includes [niche stack]”).

Pull in your SQL data with brand and category, top keywords, and any customer-facing input you have (support tickets, sales calls, site search logs, review mentions), so the list represents the buyer’s actual language, not what you think they’re saying.

If your competitors bid on information you didn’t write, you’ll never see it; Your ad library starts and ends with your content list.

Pro tip: Use Ad Radar to pull your inventory and keep it active.

Step 2: Start Each Prompt in a ChatGPT Session

Once you have a list of information, use it, and pay attention to the session setup, where the data becomes useful or noisy.

Run each prompt and take a screenshot of the answer, including any supported cards that appear below the answer.

Do not use each prompt once.

ChatGPT’s ad auction does not show the same ad to all users at the same time; different times show different advertisers depending on the bid, related signals, and rotation.

A single run captures the result of a single auction, not a competitive set.

To get a useful reading on any given data, schedule at least 20 to 30 runs over multiple days.

Change the session: clear cookies between batches, and the speed runs in the morning, afternoon, and evening. Run all 30 in 10 minutes from the same session and you take one piece of the auction.

Step 3: Capture the Four Data Points That Define a Competitor’s Ad

For every sponsored placement you display, record the same four fields, in the same place, every time. Otherwise you cannot compare all runs.

Four data points to capture with each view:

1. The title of the ad: a copy of the exact title on the sponsored card. Copy letter to letter. The headlines change.
2. Description of the ad: the body sentence under the title. About 19 words on average right now, but the range varies. Download the full text.
3. Last URL: the destination URL the card links to. Strip the UTMs to identify the canonical landing page, but keep the full URL in the second column to analyze tracking patterns later.
4. Sharing impressions: calculated, not directly observed.

For each information, count how many times each advertiser appeared in the run. If you ran the prompt 25 times and Contest A showed in 12 of them, that’s a 48% share of that prompt window.

Mark each row with the command that triggered the ad, the date and time it played, and the session details (Free or Go, Location). Set up your spreadsheet so you can cycle through share impressions by notification, competitor, and week.

Ad copy replicates quickly. The same marketer might launch three or four different titles against the same content in one week as their team explores creativity. Final URLs also change; A competitor may cycle between the home page, comparison page, and category landing page to check for conversions. Capture only the title and miss repetition patterns and URL strategies, which tell you a lot about what your competitor is doing.

Step 4: Repeat Often Enough to See Voice Sharing Over Time

A single image reading from a competitor’s ad activity will mislead you. You will catch whoever won the auction on the day you used the information and miss the rotation that happens every day. Decide on a budget from a one-day snapshot and decide on a sound.

To see voice sharing, which means who actually owns this section in ChatGPT, you need a recurring cadence. The minimum that gives you a signal:

• Every day it starts with your top 5 to 10 highest value notifications (the questions closest to the purchase intent)
• It runs weekly on a total list of 30–50 alerts
• Monthly trend draws to see how competitors are gaining or losing share over rolling 30-day windows

Pro tip: Use Ad Radar to run this cadence automatically and get a continuous reading of your competitor’s ad activity in ChatGPT, without a spreadsheet on top.

Stop Flying Blind in Paid AI Search

Paid search managers have auction data, ad libraries, and a host of third-party Google monitoring tools. For ChatGPT ads, they don’t have that yet. ChatGPT ads are a new auction against the same buyer intent, and currently many groups cannot see who is bidding on them. If competitors are already in your ChatGPT customer response, you will find out through your monitoring or pipeline gap that you see too late to fix it.

Ad Radar conducts continuous data monitoring and reveals every advertiser, every piece of information, every creative iteration. See the ongoing visibility of ChatGPT’s competitor ad activity in your category.

Photo Credits

Featured image: Photo by Shutterstock. Used with permission.