Key takeaways
- ChatGPT's Ads Manager only reports your own campaigns. It does not show which competitors are buying placements, what their ads say, or where clicks go. There is no public ad library like Meta's or Google's.
- Every AI ad tracker works the same way: re-run prompts on a schedule and record the sponsored placements that appear. All numbers are observed samples, not a census.
- Ad-bearing responses are not rare. Promptwatch Data found ads in an average of 20.1% of ChatGPT web-search responses between May 20 and Aug 17, 2026.
- The features that matter are full creative capture (headline, copy, landing page, triggering prompt), organic and paid in one view, gap reporting, and clear pricing.
- Tracking your own campaigns is a separate job from tracking competitors. You need both.
Why AI search ads need their own tracking
If you run Google Ads, you can open the Ads Transparency Center and see what a competitor is running. Meta has its Ad Library. ChatGPT has nothing comparable.
OpenAI started testing ads in ChatGPT in the US in February 2026, according to the timeline in Rankshift's roundup of ChatGPT ads trackers. It later opened a self-serve Ads Manager and rolled sponsored placements out to Free and Go users in dozens of countries, including most of Europe. The Ads Manager reports impressions, clicks, spend and conversions for your own campaigns. That is all.
So the questions most advertisers actually care about go unanswered by the platform:
- Which competitors show up on my prompts?
- What do their ads say?
- Which landing pages do they send people to?
- Which prompts trigger an ad at all, and which are still open?
Third-party trackers fill that gap. None of them has access to OpenAI's auction. They are all observers.
How big is the opportunity, really?
I would not trust any single number here, and the data sources disagree on timing. That is worth being upfront about.
Promptwatch Data's ChatGPT Ads Over Time report covers May 20 to Aug 17, 2026. Over those 90 days, ads appeared in an average of 20.1% of ChatGPT responses that ran a web search. The report saw no ads until May 27, when they showed up in 1.69% of responses. The rate rose in bursts: a brief spike on June 25 and 26, then a jump past 30% on July 1 that held. The daily peak was 43.9%, and the last 30 days of the period averaged 29.6%.
A few details from that report are useful when you set up your own monitoring:
- Only responses where ChatGPT searched the web, returned at least one citation, and completed ad extraction are counted.
- Of ad-bearing responses in the final week, 73.3% were organic prompts, 18.0% brand-specific, and 8.6% competitor comparison. That mix mostly reflects how many prompts of each type were being tracked, so adding more organic prompts will raise their share without any change at OpenAI.
- A single run can land on an outlier day. Look across several days before drawing conclusions.
Now the conflict. Rankshift dates the US test to February 2026, while Promptwatch saw no ads until May 27. These probably are not measuring the same thing, since Promptwatch only counts web-search responses with citations. Treat the start date as unsettled and judge each tool by how it samples.
SE Ranking's analysis of 50,006 commercial prompts from August 2026 (cited via Rankshift) found about 26% showed a paid placement. It also found that the advertiser was cited organically in only 3.63% of those placements. If that holds up, paying for the slot and earning the citation are very different things.
What to look for in an AI ads tracker
Rankshift's own selection criteria are reasonable, even though the company ranks itself first. I would boil them down to this:
- Detection on your own prompts. Generic market indexes are fine for benchmarking, but you need your buyer queries.
- Full creative capture. The advertiser name alone is not enough. You want headline, body copy, landing page URL, and the prompt that triggered it.
- Organic and paid together. Seeing who is cited and who is paying on the same prompt tells you where the real gaps are.
- Gap and share reporting. Which prompts have competitors advertising while you do not?
- Public pricing, because this category changes monthly.
- Multiple runs over time, so one noisy day does not mislead you.
The tools compared
Prices below come from vendor-authored roundups and change quickly. Check each vendor's site before you buy.
| Tool | Starting price (as listed) | What it shows | Best for |
|---|---|---|---|
| Promptwatch | See promptwatch.com/pricing; Ads Radar on Business, agency plans and Enterprise | Advertisers per prompt, ad copy, claims, landing pages, alongside citations and visibility | Teams that want ads, organic citations and content fixes in one place |
| Rankshift | From EUR 69/mo billed annually, 30-day free trial | Gap report of prompts where competitors advertise and you do not | Budget-conscious teams focused on ad gaps |
| SE Ranking | From $129/mo ($103.20 billed annually) | Ad detection inside Competitive Research, no add-on | Teams already on SE Ranking |
| Adthena | $399/mo after a 21-day trial | 300,000+ prompts monitored daily, greenfield opportunity detection | Enterprise paid search teams |
| Gauge | From $599/mo | Your OpenAI Ads account metrics next to competitor ads | Advertisers who want own and competitor data together |
| Writesonic AI Ads Index | Free | Daily index of how often ChatGPT shows ads, 90 days of history | Market benchmarking only |
One caveat on that table: most of those entries come from Rankshift's own comparison, which is vendor-authored. I could not independently verify features for every product, and I did not verify anything about ads in Google AI Mode, AI Overviews, or Perplexity.
Promptwatch
Promptwatch is an AI visibility and GEO platform, and its ChatGPT ads tracking (Ads Radar) shows which advertisers bid on your prompts, their ad copy and claims, and the landing pages they use. You can view it by ad or by prompt. The useful part is context: the same platform tracks organic citations, so you can see on one prompt whether a competitor is paying, being cited, or both. Ads Radar is available on Business ($579/mo), every agency plan (from Kick-off at $199/mo), and Enterprise. Confirm current pricing at promptwatch.com/pricing.

If the ad data points at a content gap, its Action Board and Content Agent can help you act on it. That matters given the SE Ranking finding that few advertisers are also cited organically. You can start with the 7-day free trial on monthly billing, or run the free AI visibility report first.
Rankshift
Rankshift is the source of most of the pricing data in this space, and it sells its own tracker, so keep that in mind. Its gap report is the feature to look at: prompts where competitors advertise and you do not. The entry price is low.
SE Ranking
SE Ranking builds ad detection into Competitive Research, which means no separate add-on if you already use it. That is convenient if your team lives in an all-in-one SEO platform. Its ChatGPT ads data also underpins the 50,006-prompt analysis cited above.

Gauge
Gauge connects to your OpenAI Ads account, so your own impressions and spend sit next to competitor placements. It is the priciest on the list, but the only one here pitched at closing the loop between your campaigns and the competitive view.
Adthena and the free index
Adthena is aimed at enterprise paid search teams and monitors 300,000+ prompts a day. Writesonic's AI Ads Index is free and useful as a market thermometer, but it tells you how often ads show up in general, not who is on your prompts.

Classic ad tools are not the answer
Much of the "best ad tracking" content you will find in 2026 is about something else. Roundups like Adspirer's AI ad management comparison and the Ryze guide to Google Ads tools cover management and attribution for Google, Meta and TikTok. Both are vendor-written and rank their own product first. They are useful for running campaigns on those networks, not for seeing who advertises inside ChatGPT.

Attribution tools like Cometly and Hyros answer a different question: which of your ads produced revenue. They help once you have campaigns running. They do not show competitors.
And SpyFu still shows what competitors buy in classic search, which remains worth checking next to the AI view.
A practical setup
Here is the workflow I would use.
Step 1: build a prompt set that mirrors buying intent
Start with 30 to 50 prompts across commercial, comparison and brand queries. Do not load it with organic informational prompts only, since ad-bearing responses will skew toward whatever you track most.
Step 2: run for at least a week
Ads do not appear on every run. Promptwatch's report warns that one run can hit an outlier day, so compare multiple days before you decide anything.
Step 3: record the full creative
For each competitor ad, log the headline, the claim being made, the landing page, and the triggering prompt. Claims are the most useful bit. If three competitors all say "free setup", you know what the market is fighting over.
Step 4: compare paid and organic
Check whether the advertiser is also cited organically on that prompt. Given the 3.63% overlap in SE Ranking's sample, many are renting visibility they have not earned. That is your opening: if you earn the citation, you get visibility without the bid.
Step 5: track your own campaigns separately
Use OpenAI's Ads Manager plus your own analytics for your own spend and conversions. As Rankshift points out, this is its own job, and the competitor trackers do not replace it. For attributing AI traffic, note that ChatGPT stopped sending a referrer on October 8, 2026, so tagged links (utm_source=chatgpt.com) matter more than they used to.
Which tool for which team
| Situation | Pick |
|---|---|
| You want ads, organic citations and content actions together | Promptwatch |
| You want the cheapest way to see competitor ad gaps | Rankshift |
| You already pay for SE Ranking | SE Ranking's Competitive Research |
| You run enterprise paid search at scale | Adthena |
| You want your Ads Manager data next to competitors | Gauge |
| You just want a market benchmark | Writesonic AI Ads Index (free) |
If you are comparing the wider category of AI visibility platforms, the GEO software directory at bestgeosoftware.com lists more options.
Limits you should know about
All of these tools sample. They do not see impression counts, bids, or targeting rules. A tracker that says a competitor ran an ad on a prompt is telling you it observed one, not how often real users saw it. Results can also vary by location and user type, which is why monitors pinned to a country matter. And since ad formats are changing fast, a tool's output today may not look the same in a quarter.
Do not make budget decisions off a single data point. Use the trackers to find patterns, then test them with your own small campaign.
Bottom line
OpenAI gives you your own numbers and nothing about anyone else, so a third-party tracker is the only way to see competitor advertisers and copy in ChatGPT. Pick one that captures the full creative on your prompts and puts paid next to organic. If a competitor is paying for a prompt where nobody is earning the citation, that is the cheapest gap to close.
Not sure where to start? Run the free AI visibility report to see how ChatGPT, Gemini and Perplexity currently talk about your brand, then decide which prompts deserve ad tracking.




