Ad event time is the reporting basis that credits each conversion to the date and time of the ad click or impression that earned it. It assigns conversions to when the ad interaction occurred, while conversion time, the alternative basis, assigns them to when the customer actually converted.
If you judge daily return on ad spend, the basis you choose decides which day each conversion lands on, and that changes the numbers you act on. Read the wrong basis and you will compare spend against conversions from different days, misprice your campaigns, and cut the ones that are actually working.
In OpenAI's Ads Manager, ad event time is the default way conversions are dated. Each conversion is counted on the day of the ad click or impression that received attribution credit. The conversion's actual date is recorded separately under the conversion-time basis.
Three things have to line up for a figure to appear under ad event time: a tracked ad interaction with a timestamp, a conversion event sent through the OpenAI Pixel or Conversions API, and an attribution window wide enough to connect them. OpenAI's Ads Manager offers click-through windows of 7, 14, or 30 days, with view-through set to off or one day. When the conversion falls inside that window, its credit is written back to the interaction date.
This puts ad event time next to conversion time and downstream of your attribution window, which sets how far back a conversion can reach. For teams measuring whether ChatGPT ads drive pipeline, the dating basis decides how spend and results line up on a calendar. AirOps helps marketers connect that ad-driven demand to the broader AI search visibility they are building across ChatGPT and other answer engines.
Resources: See how AirOps runs and measures AI-powered ads inside ChatGPT
Ad event time is produced by the attribution pipeline every ChatGPT ads conversion passes through. It runs in order, from the click to the row you read in a report.
Record the interaction: When someone clicks or views your ad, OpenAI logs the interaction with a timestamp and attaches a click reference to your landing page URL.
Capture the conversion: Your site or server fires the conversion event through the OpenAI Pixel or Conversions API, carrying its own event timestamp.
Match within the window: OpenAI ties the conversion back to the earlier interaction if it falls inside the attribution window you set.
Credit the interaction date: The matched conversion is written to the day of the ad click or impression, which becomes its ad event time.
Backfill the report: As late conversions arrive, OpenAI adds them to that same interaction day, so past dates keep filling in across the window.
Read this way, the report tells you which spending day produced results. It does not tell you when customers actually converted; the conversion-time view answers that.
How you date conversions decides how you spend. Because ad event time lines results up against the exact days you paid for clicks, it is the basis that tells you whether a ChatGPT ads campaign earns its budget day by day.
Spend and return share a calendar: Your cost sits on the click day, and ad event time puts the resulting conversions on that same day, so your daily cost per acquisition and return on ad spend reflect one coherent period.
Recent days read low until they settle: Attributed conversions can take 24 to 48 hours to appear, per OpenAI, so yesterday looks weak on an ad-event-time report; pause a campaign on that incomplete day and you kill a winner before the credit lands.
Attribution scope changes the verdict: The window you choose decides which conversions ever reach the interaction day, so a short window and a long one can make the same campaign look mediocre or strong.
SEO managers use ad event time to line ChatGPT ad conversions up against the days they spent, so paid and organic reporting share one timeline.
Growth marketers use ad event time to judge daily cost per acquisition before deciding which campaigns to scale or pause.
Demand gen leads use ad event time to tie pipeline from ChatGPT ads back to the spend dates their budgets are planned around.
The configurable span during which a conversion can still be credited back to an earlier ad click or impression, which directly controls which calendar day ends up holding the result and how far into the past a report can keep changing.
The alternative reporting basis that dates each conversion to the moment the customer actually completed the action, used for finance, cohort, and lifetime-value views that care about when outcomes happened instead of when ads were served.
The lag and gradual filling-in of recent days as late-arriving conversions are credited to their original interaction dates, which is why a fresh ad-event-time row almost always looks lower on the day it is first viewed.
Align daily spend with the conversions it produced for cleaner cost-per-acquisition math.
Compare ChatGPT ads performance day over day on the same basis as your ad costs.
See results natively in OpenAI's Ads Manager, where ad event time is the default column.
Switch to a conversion-time view when finance needs outcomes dated to when they happened.
Match the attribution window to your sales cycle so credited days reflect real buying behavior.
Pick one reporting basis for a given dashboard and keep it fixed, because switching between ad event time and conversion time mid-analysis makes trends impossible to read.
Wait a day or two before judging a recent date, since attributed conversions keep landing on their interaction day after the click.
Set your attribution window to match your real sales cycle, so the credited day reflects how long buyers actually take.
Label every export with its basis and window, because a number has no meaning until a reader knows how it was dated.
Reconcile ad-event-time totals against your own backend records weekly, so you catch tracking gaps before they distort optimization.
Keep deduplication working across pixel and Conversions API, because double-counted conversions inflate whichever day they land on.
Avoid optimizing against yesterday on an ad-event-time report. The recent days are still filling in, and acting on them treats an incomplete picture as final, which is the mistake even experienced buyers repeat.
AirOps: connects ChatGPT ad-driven demand to your wider AI search visibility, so conversions dated by ad event time sit alongside the citations and answer coverage AirOps tracks across answer engines.
OpenAI Ads Manager: reports conversions on ad event time by default and lets you toggle to a conversion-time column and set the attribution window.
Google Analytics 4: cross-checks on-site conversions and session timing so you can reconcile ChatGPT ad results against your own event timestamps.
Open your conversion columns: In OpenAI's Ads Manager, find the conversions column and confirm it is reporting on ad event time, which is the default. This takes minutes and needs no approval.
Confirm your attribution window: Check which click-through window and view-through setting your campaigns use, since this decides which conversions reach each interaction day.
Compare the two bases: Switch between ad event time and the conversion-time column on the same date range to see how far the daily numbers move.
Document the standard: Write down which basis and window each report uses, and share it so everyone reads the same numbers the same way.
Build a waiting rule: Set a team rule to let recent days settle before making spend decisions, so you optimize on stable ad-event-time data.
Ad event time dates each ChatGPT ads conversion to the ad click or impression that earned credit, using the interaction day instead of the conversion day.
It is measured inside OpenAI's Ads Manager, where it is the default conversion-reporting basis.
Its main constraint is the attribution window, which sets how far back a conversion can be credited.
Its main risk is judging recent days too early, before late conversions have backfilled their interaction dates.
The payoff is aligning spend and results on a single calendar, which makes daily cost-per-acquisition and return-on-ad-spend decisions trustworthy.
Ad event time and conversion time answer two different questions about the same conversion. Ad event time credits the conversion to the day of the ad click or impression that drove it, which is the view OpenAI's Ads Manager shows by default. Conversion time credits it to the day the customer actually completed the purchase or signup. The practical difference shows up on your calendar: on an ad-event-time report, a sale that happened today but came from a click five days ago is counted on that earlier click day, while a conversion-time report counts it today. Marketers tend to live in the ad-event-time view because advertising cost is also booked on the click day, so cost and results stay on matching dates. Finance and lifetime-value analysis often prefer conversion time because it reflects when revenue genuinely arrived. Neither is wrong; they serve different jobs, and the key is to know which one a given report uses.
Use ad event time whenever you are managing campaigns to spend efficiency, which is most day-to-day optimization work. Because your ad cost is recorded on the day of the click, dating conversions the same way lets you read cost per acquisition and return on ad spend for a single, coherent period. Reach for the conversion-time view in narrower situations: when finance needs revenue dated to the day it actually came in, when you are building cohorts around customer behavior, or when you are reconciling against a billing system that records the sale date. A good habit is to keep your optimization dashboards on ad event time and spin up conversion-time views on request, instead of flipping the default back and forth. Switching the basis in the middle of an analysis is where mistakes creep in, because the same campaign can look different under each one. Decide the basis before you pull the report, and label which one you used.
Ad event time keeps changing on recent days because conversions are credited backward to the interaction that earned them, and those conversions do not all arrive at once. Someone can click your ad today and convert four days from now; when they do, that future conversion is written back onto today's date. On top of that, newly attributed conversions take time to appear in Ads Manager at all, which delays the picture further. The combined effect is that any recent day starts out looking low and fills in over the following days, up to the length of your attribution window. Older days settle once their window has fully elapsed and no more conversions can be credited to them. This is normal behavior, and it does not signal a tracking fault. The fix is patience: let a date mature before you treat its ad-event-time numbers as final, and compare like-aged periods when you benchmark performance.
You can influence ad event time, but only indirectly, because it is a reporting output and not a dial you set. The strongest control is your attribution window: widening or narrowing the click-through window changes which conversions are eligible to be credited back to an interaction day, so it reshapes how results distribute across your calendar. Your tracking setup matters too. If your OpenAI Pixel or Conversions API events are missing, delayed, or duplicated, the days they land on will be wrong, so clean, deduplicated event delivery keeps ad event time accurate. You cannot, however, hand-pick which day a specific conversion lands on; that is determined by the interaction that won attribution. The honest answer is that you shape ad event time by configuring the window and by fixing your measurement plumbing, and then you read what the system reports. Treat it as a measurement you keep accurate, and spend your energy on the window and the tracking.
A reliable ad event time number is one from a date that has had time to settle and sits within a clearly defined attribution window. In practice, that means giving a recent day at least a couple of days before you trust it, since late conversions are still being credited back to it. It also means comparing periods of the same age, such as last week against the week before once both have fully matured, instead of a settled week against a still-filling one. There is no universal benchmark figure for the number itself, because a good cost per acquisition or return on ad spend depends entirely on your margins and sales cycle. What counts as good is internal consistency: the ad-event-time totals should roughly reconcile with your own backend records once the window has closed. If they diverge sharply, suspect a tracking or deduplication problem before you trust the trend.