A standard event is a predefined, named conversion action that ChatGPT Ads recognizes by default, such as order_created, lead_created, or registration_completed. Unlike a custom event, which you name and define yourself, a standard event uses OpenAI's fixed taxonomy so the platform can optimize bidding and attribution against it without guesswork.
When you pick the standard event that matches your real conversion goal, you tell ChatGPT Ads exactly what success looks like and let it bid toward that outcome. Choose the wrong event or log everything as custom, and your optimization goals break, your reported conversions drift from reality, and wasted spend compounds before you notice.
Standard events are OpenAI's built-in vocabulary for the conversions advertisers track most: a purchase, a lead, a sign-up, a trial start, an added cart item. Each carries a fixed name and a typed data shape, so when you fire order_created with an amount and currency, ChatGPT Ads knows how to read and value it.
Every standard event has three parts: the name from OpenAI's taxonomy, a data payload like amount or currency, and an optional identifier that deduplicates the same conversion across the pixel and the Conversions API. You can send it from the browser, from your server, or both, and OpenAI matches it to your Pixel ID and your campaign's conversion goal.
A standard event sits alongside the custom event, which you define when no built-in name fits your action. Standard events are the safer default because the platform understands them natively for bidding and reporting. AirOps helps teams map each funnel stage to the right standard event and keep that tracking consistent as campaigns scale.
A standard event travels from a user action on your site to a counted conversion in OpenAI's Ads Manager. Here is the sequence.
Provision the pixel. Create a data source in the conversions tab of Ads Manager to generate your Pixel ID.
Install tracking. Load the oaiq pixel in your site's head, fire the server-side Conversions API, or run both for coverage.
Fire the event. When the action happens, send the matching standard event name with its typed payload, such as order_created with amount and currency.
Deduplicate. Pass the same Pixel ID, event name, and event identifier on browser and server so one conversion counts once.
Match and attribute. OpenAI ties the event to your campaign and its attribution window, then reports it.
The result tells you which ads drove which conversions inside the window you set. It does not reveal user-level identities, the conversation that led to the click, or demographic breakdowns.
Resources: see how AirOps connects your ChatGPT Ads performance and measurement in one connected system.
Your choice of standard event is the instruction ChatGPT Ads follows when it decides which impressions to buy. Pick the event that reflects revenue, and the platform spends toward revenue. Treat tracking as an afterthought, and you pay for clicks that never tie back to a business outcome. That single setting quietly shapes where most of your budget goes.
Cleaner optimization: Standard events feed OpenAI's bidding with a goal it understands natively, so automated bidding has a real target instead of a vague proxy.
Trustworthy reporting: Because every standard event has a fixed name and shape, conversions roll up consistently across campaigns, and your Ads Manager numbers reconcile with your own analytics.
Fewer costly mistakes: Logging a purchase as a generic custom event, or firing lead_created on every page load, inflates conversions and misdirects budget; the failure usually hides until spend has already scaled.
Growth marketers use standard events to point ChatGPT Ads optimization at order_created, so automated bidding chases revenue instead of raw traffic.
Demand gen leads use standard events to fire lead_created on qualified form submissions, feeding the campaign a goal that maps to pipeline.
SEO managers use standard events to track registration_completed on content sign-ups, connecting ChatGPT Ads conversions to the same actions they report elsewhere.
OpenAI publishes a fixed set of standard event names, each tied to a stage of a typical funnel, and choosing the name that matches your action is what lets the platform interpret the conversion correctly and compare it consistently across campaigns.
Every standard event carries structured fields such as amount, currency, or contents, and sending accurate values is what turns a raw event into something the platform can value and optimize against while a campaign is live.
When you fire the same conversion from both the pixel and the Conversions API, OpenAI relies on the shared Pixel ID, event name, and event identifier to count a single purchase once instead of double-counting it.
Align ChatGPT Ads bidding with your real conversion goal instead of a stand-in metric.
Report conversions consistently across campaigns in OpenAI's Ads Manager.
Deduplicate browser and server events with a shared Pixel ID so each conversion counts once.
Capture conversions the browser misses by pairing the pixel with the Conversions API.
Map every funnel stage to a named event the platform already understands.
Use a standard event whenever one fits: reach for the custom type only when no built-in name matches, because the platform optimizes standard events more reliably.
Match the event to a real outcome: fire order_created on confirmed purchases and lead_created on qualified submissions, so your optimization goal reflects value.
Run the pixel and Conversions API together: the server side catches what ad blockers and lost sessions drop from the browser.
Share one event identifier across both paths: deduplication keys on Pixel ID, event name, and that identifier, so a single conversion is never counted twice.
Send accurate payloads: pass real amount and currency values, because bidding toward revenue depends on the numbers you report.
Test before you scale: use the pixel's debug mode and the API's validate-only option to confirm events fire correctly first.
Avoid firing high-value events like order_created on page loads or test traffic to make dashboards look busy; inflated conversions train the platform on noise and burn budget on buyers who never existed.
AirOps: maps each stage of your funnel to the right standard event and keeps that conversion tracking consistent as you scale ChatGPT Ads campaigns.
Google Tag Manager: deploys the oaiq pixel and fires named standard events with typed payloads without hard-coding scripts on every page.
Google Analytics 4: gives you an independent conversion count to reconcile against the standard events reported in OpenAI's Ads Manager.
Map your funnel. List the conversions that matter, from content sign-ups to purchases, and match each to a standard event name. You can do this in an afternoon with no budget approval. Start here before touching any code.
Create a Pixel ID. In the conversions tab of Ads Manager, create a data source to get the Pixel ID every event will report to.
Install the pixel. Add the oaiq loader to your site's head and initialize it with your Pixel ID, through Google Tag Manager if you use it.
Fire your events. Send each standard event at the right moment with an accurate payload, and add the Conversions API for transactional conversions. Reuse one event identifier on both paths.
Validate and reconcile. Use debug and validate-only modes to confirm events fire, then compare Ads Manager totals against your own analytics before scaling spend.
A standard event is one of OpenAI's named conversion types, chosen from a fixed list instead of invented.
You fire a standard event from the oaiq pixel, the Conversions API, or both, with a typed payload like amount and currency.
The platform only understands the names in its taxonomy, so anything outside it must fall back to a custom event.
Firing the wrong event or an inflated one trains bidding on noise and wastes spend before anyone notices.
The biggest leverage is matching each standard event to a real revenue outcome, which points automated bidding at value.
A standard event uses a name from OpenAI's fixed taxonomy, while a custom event is one you name yourself when no built-in type fits. The practical difference is how the platform treats them. Because a standard event like order_created or lead_created has a known meaning and data shape, ChatGPT Ads can optimize bidding and report conversions against it with no extra interpretation. A custom event still gets tracked, but the platform has less built-in understanding of what it represents, and a custom event generally cannot serve as an optimization goal in the same way. The rule of thumb is simple: use a standard event whenever one reasonably matches the action you are tracking, and reserve custom events for the genuinely unusual conversions your business cares about that have no standard equivalent. Starting with standard events keeps your measurement legible to both the platform and your own team.
Fire a standard event the moment the action it represents actually completes: order_created on a confirmed purchase, lead_created on a submitted form, registration_completed on a finished sign-up. Timing matters more than volume. There is no fixed number you should track, but most advertisers do well with a small set that mirrors their real funnel, usually a handful spanning an early signal, a mid-funnel action, and the final revenue event. Tracking every possible event rarely helps, because events that do not map to a decision you will act on add noise instead of insight. Pick the one event that defines success for a campaign and make it the optimization goal, then add supporting events only if you will genuinely use them for analysis. The discipline is to measure what you will act on, and to fire each event once, accurately, at the point the conversion is real.
Standard event conversions often differ from your own analytics because the two systems count under different rules. ChatGPT Ads attributes conversions to its own click-through and view-through windows, so a sale your analytics credits to another channel may still appear as a standard event conversion here, or the reverse. Browser conditions add more variance: ad blockers, cookie restrictions, and lost sessions can stop the pixel from firing, which is exactly why pairing it with the server-side Conversions API closes part of the gap. Deduplication settings matter too, since a shared event identifier is what prevents the same purchase from being counted on both the pixel and the API. Some drift is normal and expected across any two measurement systems. You are aiming for a stable, explainable relationship instead of a perfect match, so when the numbers move you can tell whether tracking changed or performance did.
Yes, you directly control which standard event a ChatGPT Ads campaign optimizes toward, and this is one of the most important levers you have. In Ads Manager you choose the conversion event tied to a campaign, and the platform's automated bidding then works to produce more of that specific event. If you set order_created as the goal, bidding chases purchases; if you set lead_created, it chases leads. What you cannot do is make the platform optimize for a custom event the same way a standard event works, since custom events are limited as optimization goals. You also cannot change history: switching the goal shapes future delivery while leaving conversions already recorded untouched. The practical move is to decide what a campaign is really for, set the standard event that represents that outcome as the goal, and give the system enough conversions to learn from before you judge or change it.
A good standard event setup is one where every tracked event maps to a real business outcome, fires once and accurately, and reconciles closely with your own analytics. In practice, that means you have picked a clear primary event as your optimization goal, you run the pixel and the Conversions API together so coverage survives ad blockers and lost sessions, and you deduplicate with a shared identifier so nothing is double-counted. A healthy setup also shows stable, explainable numbers over time instead of wild swings you cannot account for. Beyond the mechanics, the real benchmark is whether the data changes what you do: if your standard events let you confidently shift budget toward the campaigns and audiences that drive revenue, the setup is working. If the numbers exist but nobody trusts or acts on them, the setup still has work to do, no matter how clean the implementation looks.