Product card clicks are the engagements ChatGPT Ads counts when a shopper taps a sponsored product card beneath an AI answer and lands on your merchant site, app, or an expanded product view. They are narrower than a campaign's total clicks, which also include text and recommendation card taps, so they isolate how your shopping unit performs.
If you sell through a product feed, this is the number that tells you whether your card art, price, and rating are pulling shoppers out of the conversation. Ignore it and you optimize blind, pouring CPC budget into cards that get seen in ChatGPT but never earn the tap that starts a purchase.
In ChatGPT Ads, a product card click is a billable engagement recorded the moment a user selects your sponsored shopping unit and is routed to its destination. It is the action OpenAI charges against in a cost-per-click campaign built on a product feed.
The product card is the shopping format that pulls image, title, price, and star ratings straight from your catalog feed and renders beneath the AI response, labeled as sponsored. A click fires when the shopper taps that card to reach your site or app, expand the card for more detail, or begin a direct checkout inside the thread. Each tap is attributed to the campaign, ad group, and ad that served it.
Product card clicks sit between the impression, where the card was shown but untouched, and the conversion, where a purchase or signup follows. They measure intent. They do not measure outcome. AirOps tracks the prompts that trigger your shopping cards so you can connect ChatGPT visibility to the clicks they drive.
Resources: see how to structure ChatGPT Ads campaigns that earn product card clicks
A product card click moves through the same pipeline every time, from the shopping query that surfaces your card to the tap that OpenAI logs and bills.
Query match: ChatGPT reads a shopping or commercial prompt and decides your product feed is relevant to the answer.
Card render: The sponsored product card renders beneath the response with your image, title, price, and rating, counting as an impression.
Shopper tap: The user taps the card, triggering the product card click and sending them to your destination URL or an expanded view.
Attribution: OpenAI attributes the click to the serving campaign, ad group, and ad, and charges your cost-per-click bid.
Reporting: The click rolls into Ads Manager alongside impressions, CTR, and spend, with no query-level or demographic breakdown.
That count tells you how many shoppers chose to act on your card. It does not tell you which prompts produced them, since OpenAI keeps query-level data closed.
Product card clicks are the first hard signal that ChatGPT shopping spend is working, and they shape whether you keep funding the channel or pull the budget back to Google and Meta. For a feed-based advertiser, this metric is where the case for the channel is won or lost.
They price your media: In a cost-per-click campaign, every product card click is a line item, so the volume and cost of these clicks decide your effective CPC and how far a test budget stretches.
They expose creative weakness: A card that earns impressions but few clicks is a failure mode you can see early, telling you the image, price, or rating is losing shoppers before diagnosis gets expensive.
They anchor the funnel: Clicks are the only step you can measure natively before conversions arrive, so they let you compare feeds and ad groups while attribution is still thin and unreliable.
SEO managers use product card clicks to see whether ChatGPT shopping placements pull the same high-intent shoppers their organic product pages target.
Growth marketers use product card clicks to compare cost-per-click efficiency across feeds before committing a larger ChatGPT Ads budget.
Content strategists use product card clicks to learn which product framings and prices earn taps, then feed those angles back into product page copy.
The share of card impressions that become product card clicks is your card's click-through rate, the clearest read on whether the creative and price earn attention once ChatGPT surfaces it, and it is the first number to move when you change the hero image.
Because feed campaigns bill per click, each product card click carries a bid-driven cost, so rising click volume raises spend directly and makes bid discipline central to the metric; a cheap click on an irrelevant query still costs you.
A product card click is attributed last-touch to the serving ad, and any downstream conversion is credited inside OpenAI's click-through window of seven, fourteen, or thirty days, which you set in reporting.
Measure whether your ChatGPT shopping cards earn action beyond exposure.
Isolate product card performance from text-ad clicks inside OpenAI Ads Manager.
Control cost-per-click spend by pausing feeds whose cards rarely get tapped.
Benchmark creative changes against click-through rate before conversions mature.
Surface which products resonate with ChatGPT's high-intent shopping audience.
Guide bid decisions in ChatGPT Ads using live click volume and cost.
Lead with a sharp hero image, because the card's visual does most of the work in earning the tap.
Keep price and rating accurate in your feed, since shoppers compare them against the AI's own recommendations before clicking.
Segment feeds by product category, so you can read product card clicks per segment instead of one blended number.
Set a disciplined cost-per-click bid, because ChatGPT's shopping audience clicks less often than search, so overbidding drains budget fast.
Pair the click data with a conversion pixel, so you learn which clicks actually turn into purchases.
Review click trends weekly, since the channel is young and performance shifts as OpenAI changes formats.
Avoid treating every product card click as equal value. A click from a loosely matched prompt can cost the same as a high-intent one, so judge clicks by what they convert to, and cut feeds that generate taps without downstream revenue.
AirOps: Tracks the ChatGPT prompts that surface your shopping cards and ties that visibility to the product card clicks they generate, so you can see which queries drive taps.
OpenAI Ads Manager: Reports product card clicks natively alongside impressions, CTR, and cost-per-click at the campaign, ad group, and ad level.
Google Analytics 4: Captures the on-site sessions that follow a product card click, helping you reconcile ChatGPT's click count against landing-page behavior.
Open Ads Manager: Log into OpenAI Ads Manager and locate the clicks column on an existing or draft campaign. This takes an afternoon and needs no new budget. Watching how clicks already accrue grounds everything that follows.
Connect your feed: Upload or sync your product catalog so ChatGPT can render shopping cards, which is what makes product card clicks possible.
Install tracking: Add OpenAI's conversion pixel and, where you can, the Conversions API, so clicks can be cleanly tied to on-site actions later.
Launch a CPC test: Run a small cost-per-click campaign with a modest bid and let it gather clicks across a handful of product groups. Keep the spend low enough to treat it as data collection.
Read and iterate: Compare product card clicks and click-through rate across feeds each week, then shift budget toward the cards earning the most qualified taps at the lowest cost.
Product card clicks count the taps shoppers make on your sponsored ChatGPT shopping card beneath an AI answer.
OpenAI logs them natively in Ads Manager and bills them under cost-per-click campaigns built on a product feed.
The data stays aggregate, with no query-level or demographic detail behind each click, limiting how deeply you can diagnose.
Dashboard click counts can outrun what your analytics records, so numbers need reconciling before you trust them.
The leverage is in the card's image, price, and rating, which decide whether impressions become clicks.
Total clicks count every tap your ChatGPT Ads campaign earns, while product card clicks count only the taps on your sponsored shopping cards. That distinction matters once you run more than one ad format. A campaign can serve a recommendation card and a product card to different shoppers, and lumping their clicks together hides which unit is actually pulling. Product card clicks let a feed-based advertiser see how the shopping unit performs on its own, separate from text or recommendation placements. Because OpenAI reports clicks at the campaign, ad group, and ad level, you can usually isolate the product card figure by structuring your account so shopping cards sit in their own ad groups. If everything shares one ad group, the native dashboard gives you a blended number and you lose that resolution. Set the account up for separation first, and the difference between the two metrics becomes something you can read and act on.
Review product card clicks at least weekly on a new ChatGPT Ads campaign, and daily for the first week while you confirm the feed is serving and tracking correctly. ChatGPT's shopping audience is smaller than search, so click volume accrues slowly, and checking hourly tells you nothing but noise. A weekly cadence gives enough data to compare cards fairly without overreacting to a single slow day. During the opening week, watch more closely for setup problems: a feed that stopped rendering, a destination URL that broke, or a sudden gap between dashboard clicks and site sessions. Once the campaign is stable, a weekly read of clicks, click-through rate, and cost is usually enough to guide bid and creative decisions. Push the cadence shorter only when you change something material, such as a new hero image or a bid adjustment, and you want to see how the card responds before the week closes.
Product card clicks vary week to week because the metric depends on how often ChatGPT decides your card is relevant to live prompts, and that demand shifts constantly. You do not buy a fixed number of impressions; your card surfaces when real users ask shopping questions your feed matches, so a quiet week of relevant queries means fewer cards shown and fewer clicks. Seasonality, news, and changing conversation patterns all move that query volume. OpenAI also tunes which prompts trigger ads and how formats render, and those platform changes can swing your click counts without any change on your side. Competitive pressure plays a part too, since other advertisers bidding on overlapping intent affect how often your card wins a placement. Add the known gap between dashboard clicks and analytics sessions, and some of the variance reflects measurement gaps instead of real demand. Treat week-to-week swings as a range to monitor, and look at multi-week trends before you judge performance.
Yes, you can influence product card clicks, though you cannot command them. The levers you control directly are the card itself and the bid: a stronger hero image, an accurate and competitive price, visible ratings, and a tight product title all raise the odds that a shopper taps once your card appears. Raising your cost-per-click bid helps your card win more placements, which gives it more chances to be clicked. What you cannot control is the prompt stream. You do not choose which questions users ask or exactly when ChatGPT judges your feed relevant, and there are no negative keywords to filter intent the way search offers. So your influence is real but indirect: you improve the rate at which impressions convert to clicks and you bid for more impressions, then ChatGPT's matching decides the raw opportunity. Focus your energy on feed quality and creative first, because those move click-through rate on every impression you do earn.
Treat a good number of product card clicks as one that hits an efficient cost and a healthy click-through rate for your category, because OpenAI publishes no official benchmark across advertisers. The company has said directly that it has no cross-advertiser performance baselines yet, so any single target you borrow from a blog is a guess with no basis. Shopping-intent cards tend to earn a higher click-through rate than text placements, but ChatGPT's overall click-through rate runs well below paid search, so judging your cards against Google numbers will mislead you. The practical move is to benchmark against yourself: set your own baseline in the first few weeks, then measure each creative and feed change against that. A good result is a card whose click-through rate and cost improve over your prior period while the clicks still convert on site. Volume alone is a weak signal, since cheap clicks on loosely matched prompts can inflate the count without adding value.