Event value is the total monetary worth you assign to a conversion event in ChatGPT Ads, sent with a currency and an amount so OpenAI can report what each tracked action was actually worth. It differs from the per-unit product price, which prices a single item, because event value represents the whole event, such as a two-item order, after quantities are multiplied.
For a marketer running ChatGPT Ads, event value is what turns a raw conversion count into a revenue figure you can defend to finance. Send it wrong and your return on ad spend reporting breaks, making a profitable campaign look like a loss or the reverse.
In ChatGPT Ads conversion tracking, event value is the amount OpenAI records for a completed conversion event, expressed in minor currency units with a currency code. OpenAI recently redefined the underlying fields so a product's amount now means the price of a single unit, while the event value stands for the worth of the whole event once quantity is applied.
The value travels with the event through the OpenAI Measurement Pixel, a browser tag, or the Conversions API, which reports the same data server-side. For a two-unit order priced at 799 each, you send the unit amount and a quantity of two, and the event value resolves to 1,598. Amounts always pass in minor units, so 79900 reads as 799.00, and a mismatch here is a common reason reported revenue looks off by a factor of a hundred.
Event value sits beside the conversion count and feeds your return on ad spend: the count says how many actions happened, the value says what they were worth. It is separate from product price, a per-item feed attribute, and from your bid, which OpenAI optimizes toward the occurrence of an event. AirOps helps teams keep these conversion signals consistent across the pages and feeds that AI search surfaces read.
Event value moves from your website into OpenAI's reporting through a defined sequence, and each step has to be correct for the final revenue figure to mean anything.
Define events. Decide which standard events, such as a purchase or completed order, should carry a monetary value and which should not.
Install tracking. Add the Measurement Pixel to your site, wire up the Conversions API on your server, or run both for fuller coverage against ad blockers.
Set amounts. Send each unit's amount and a quantity so the event value resolves to the full order total instead of a single item.
Pass currency. Attach the currency code and keep the figure in minor units so OpenAI reads 79900 as 799.00 and not as a far larger sum.
Read reporting. Review the reported purchase value and return on ad spend in Ads Manager against what you spent.
The reported value tells you what your tracked conversions were worth and how efficiently spend produced revenue. It does not tell OpenAI how to bid, because conversion optimization still chases the occurrence of an event.
Event value is the line between a campaign you can fund and one you cannot, because it decides whether ChatGPT Ads shows up in your reporting as revenue or as an unexplained cost. Without it, you are approving spend on a channel you can only measure in clicks.
It makes ROAS real. Return on ad spend needs a numerator, and event value supplies it, so spend decisions rest on revenue instead of conversion counts alone.
It exposes profitable segments. With value attached to each event, you can see which products, markets, and audiences actually pay back the spend instead of treating every conversion as equal.
It prevents silent misreporting. The common failure mode is passing the value in the wrong units or sending a per-unit price as the whole-event amount, which inflates or deflates revenue by a large multiple and sends budget to the wrong campaigns for weeks before anyone notices.
Growth marketers use event value to calculate return on ad spend for ChatGPT Ads and compare it against their other paid channels.
Demand gen leads use event value to weight lead events by expected pipeline so high-value sign-ups stand out from low-intent ones.
SEO managers use event value to tie AI-surface conversions back to revenue when reporting the channel to leadership.
OpenAI expects monetary amounts in the smallest unit of the currency, so 799.00 is sent as 79900, and getting this wrong is the single most common cause of revenue reported off by a factor of a hundred.
The product amount is the price of one unit, while the event value is that unit amount multiplied by its quantity plus any order-level adjustments, so a two-item order carries an event value larger than either item on its own.
Running the Measurement Pixel and the Conversions API together lets the browser and your server both report the same valued event, which protects the value against ad blockers and dropped cookies.
Quantify return on ad spend for ChatGPT Ads instead of reporting bare conversion counts.
Compare ChatGPT Ads revenue efficiency directly against your other paid channels.
Identify which products and markets repay their spend using per-event revenue.
Weight lead and sign-up events by expected value so finance trusts the numbers.
Catch broken tracking early when reported revenue drifts far from your order system.
Send event value in minor units for every revenue event, because one unit error distorts your entire return on ad spend picture.
Pass the unit amount and quantity separately, so OpenAI resolves the whole-event total correctly instead of treating one item as the order.
Include an explicit currency code on every event, since mixed-currency accounts otherwise compare figures that do not match.
Run the Measurement Pixel and the Conversions API together, because server-side events recover value that browser tags lose to blockers.
Reconcile reported value against your own order system weekly, so drift surfaces before it misguides budget.
Keep event definitions stable across campaigns, because changing what carries value mid-flight breaks historical comparisons.
Avoid the mistake competent teams still make: sending a per-unit price as the whole-event amount. It quietly understates orders with multiple items, and because the figure still looks plausible, it can skew spend decisions for weeks before a reconciliation catches it.
AirOps: keeps conversion signals, product data, and the content behind them consistent across the pages and feeds AI search surfaces read, so your event value reflects what actually sells.
OpenAI Ads Manager: where you configure conversion events, review reported purchase value, and read return on ad spend for ChatGPT Ads campaigns.
Google Analytics 4: provides an independent revenue and conversion record you can reconcile against OpenAI's reported event value to catch unit or currency errors.
Audit events. This week, with no budget approval, list every conversion you want ChatGPT Ads to value and note which already carry a monetary amount in your checkout or lead flow.
Install the pixel. Add the OpenAI Measurement Pixel to your site and fire a valued event on your purchase or order confirmation page, passing amount, quantity, and currency.
Add the API. Wire up the Conversions API so the same valued events report server-side and survive ad blockers and cookie loss, using a shared event identifier to avoid double counting.
Verify units. Send a test order and confirm the reported value matches your order system in both amount and currency before you trust any of the numbers.
Report ROAS. Once values flow cleanly, calculate return on ad spend in Ads Manager and bring that figure into the channel reporting your team already reviews.
Event value is the monetary worth assigned to a ChatGPT Ads conversion event.
It is measured by sending an amount in minor units with a currency code through the Measurement Pixel or Conversions API.
The main constraint is that event value shapes reporting, while bidding still optimizes toward an event's occurrence.
The main risk is a unit or per-item error that misstates revenue by a large multiple.
The leverage is accurate return on ad spend, which lets you fund ChatGPT Ads on revenue instead of clicks.
Event value and product price answer two different questions in ChatGPT Ads. Product price is a per-unit feed attribute, the amount one item costs, and OpenAI treats the product amount as the price of a single unit. Event value is the worth of the whole conversion event, which you get by multiplying each unit amount by its quantity and adding any order-level adjustments. So a single order of three items priced at 50 each has a product price of 50 per item but an event value of 150. The distinction matters because reporting revenue and return on ad spend depend on the event total, while the per-unit price feeds how items are described and listed. If you send a per-unit price where OpenAI expects a whole-event value, multi-item orders are understated, and your revenue reporting drifts below reality until you reconcile it against your own order records.
Check event value closely at setup and then on a steady cadence afterward. Before you trust any reporting, run test conversions and confirm the reported value matches your order system in both amount and currency, because unit and currency errors are the most frequent problems and they are invisible until you look. After launch, reconcile OpenAI's reported purchase value against your own order or analytics records at least weekly, and immediately after any change to your checkout, feed, or tracking setup. These moments are when valued events most often break, since a developer edit or a new product type can silently alter how amounts are sent. If your account runs in several currencies or sells bundles and multi-item orders, check more often, because those cases carry the highest chance of a per-unit or minor-unit mistake. The goal is to catch drift within days, before a quarter of budget has followed bad numbers.
Reported event value varies from your order system for a few predictable reasons, and most are tracking artifacts instead of real differences. The most common is a unit error: amounts must be sent in minor units, so a value sent as 799 instead of 79900 reads as a hundredth of the true figure, or the reverse. Attribution windows are another source, because OpenAI counts conversions within a configured click-through window, so an order attributed to a different channel or falling outside the window will not appear. Deduplication matters too: if the Measurement Pixel and the Conversions API both report an event without a shared identifier, the value can double count. Refunds, cancellations, and taxes handled differently on each side also create gaps. The fix is to standardize what counts as a valued event, align currencies and units, use event identifiers for deduplication, and reconcile regularly so you know which gaps are expected and which signal a real break.
You control the event value you send, but you have limited control over how ChatGPT Ads acts on it. You decide which events carry a value, what amount and currency each one reports, and whether that data flows through the Measurement Pixel, the Conversions API, or both. What you cannot currently do is make OpenAI bid toward higher-value events: conversion optimization chases the occurrence of your selected event and ignores its monetary value, so a 500 order and a 50 order count the same to the optimizer. That means the practical influence of event value today is on reporting and your own decisions, while automated delivery stays untouched. Use it to measure return on ad spend, to compare segments, and to choose which campaigns deserve more budget, then act on those findings manually. If value-based optimization arrives later, clean and accurate event values will already be in place, so the honest move now is to get the measurement right and treat bidding as a separate lever.
A good event value setup is one where your reported revenue matches your order system within a small, explainable margin. Concretely, that means every revenue event carries an amount in the correct minor units and an explicit currency code, unit amounts and quantities are sent separately so multi-item orders resolve correctly, and the Measurement Pixel and Conversions API run together with event identifiers so nothing double counts. A healthy setup reconciles to your own records weekly with only expected gaps from attribution windows, refunds, or timing, and it holds steady after checkout or feed changes instead of drifting. You should be able to calculate return on ad spend from it and trust the number enough to shift budget on it. If your reported value swings wildly, lands a factor of a hundred off, or cannot be reconciled, the setup is not ready, and the right response is to fix units, currencies, and deduplication before you read anything into the figures.