Attributed sales are the monetary value of purchases that an ad platform credits to your ads when a shopper buys within a set attribution window after clicking or seeing one. They measure revenue earned, which separates them from conversion counts that tell you how many purchases happened without saying what each was worth.
If you judge a ChatGPT or retail-media campaign on attributed sales, you are deciding budget on a number the platform's own crediting rules shape as much as your creative does. Misread the attribution window and you will either kill a channel that quietly drives revenue or pour spend into one whose sales were already going to happen.
Reported inside an ad platform's dashboard, attributed sales represent the summed purchase value that the platform ties back to your campaigns. In ChatGPT Ads, OpenAI's reporting defines the figure precisely: the value of attributed order_created events, expressed in the ad account's currency. Each qualifying order adds the amount your store reported, so the figure is only as accurate as the values you send.
Three parts drive it. First, a conversion event carrying an amount and currency has to reach the platform through a pixel or server-side API. Second, that purchase has to match an earlier ad click or view through a stored identifier. Third, the match has to fall inside the attribution window you configured, or the sale goes uncredited.
Attributed sales sit above a plain conversion count and below return on ad spend, which divides this value by spend. Because the platform credits revenue it can see inside its own window, attributed sales rarely reconcile with the totals in your order system. AirOps helps you track the organic ChatGPT visibility that primes those paid purchases, so you read the paid number in context.
Attributed sales accumulate through a pipeline that runs every time a shopper who touched your ad completes a purchase. Each step has to succeed or the sale never reaches the column.
Tag the click. When someone clicks a ChatGPT ad, the platform appends a click identifier to the landing URL and stores it, so a later purchase can be traced back.
Fire the event. Your site or server sends an order_created event carrying the order amount and currency through the pixel or Conversions API.
Match the purchase. The platform matches that event to a prior click or view using the stored identifier and hashed customer data.
Apply the window. If the match falls inside your configured attribution window, the sale qualifies, and a click always takes precedence over a view.
Sum the value. The event amount is added to attributed sales at campaign, ad group, and ad level, usually 24 to 48 hours after the purchase.
The total tells you how much revenue the platform believes your ads produced inside its window. What it leaves out is incremental lift: whether those buyers would have purchased without the ad.
Attributed sales is the number that decides whether a channel keeps its budget, so a shaky figure sends real money to the wrong place. When you compare ChatGPT Ads against search or social, you are comparing attributed-sales figures that each platform calculated under its own rules, and the channel with the most generous rules can win the budget without earning it.
It ties spend to revenue instead of activity. A campaign can earn clicks and still produce little attributed sales, which is the signal to rework the offer before you scale.
It reveals where high-intent buyers come from. Adobe Analytics found that in July 2026 shoppers arriving via AI referrals converted at a rate 60% higher than non-AI traffic, so sales attributed to AI surfaces often carry unusual value.
It punishes silent tracking gaps. If your order_created event sends the line-item price instead of the full order total, attributed sales understates revenue and you may cut a channel that was actually winning.
Growth marketers use attributed sales to decide which ChatGPT Ads campaigns earn more budget at the weekly review and which get paused.
Demand gen leads use attributed sales to report paid-channel revenue to finance without overstating what the ads actually drove.
Content strategists use attributed sales to confirm that the product pages feeding accurate order values keep the whole reporting chain trustworthy.
The fixed period after a click or view during which a later purchase still counts toward your campaign, a setting that directly caps how much revenue your attributed sales can ever capture and makes cross-platform comparisons unreliable when the windows differ.
The amount and currency your store sends with each order_created event, the raw input that attributed sales adds up, which means a feed passing unit prices instead of order totals will quietly distort the revenue figure you report and trust.
The rule that assigns a sale to the final qualifying interaction and lets a click outrank a view, so two of your own campaigns touched before a purchase never each claim the same order inside one platform.
Connect ChatGPT Ads spend to actual purchase value in one dashboard you already check.
Compare campaigns by revenue earned instead of raw clicks won, which exposes offers that attract interest but little money.
Calculate return on ad spend directly, since attributed sales is the numerator that figure divides by spend.
Spot high-value AI-driven purchases that tend to convert above typical web traffic.
Export daily attributed sales to reconcile paid revenue against your own order system.
Fix one attribution window per platform and leave it, so your period-over-period numbers stay comparable across months.
Send the full order total in the event-level amount field, because attributed sales reads that value directly and inherits any mistake.
Pass events server-side through the Conversions API as well as the pixel, so blocked or abandoned browser sessions still report the sale.
Deduplicate browser and server events with a shared event ID, or a single purchase inflates attributed sales twice.
Reconcile attributed sales against your order system weekly, since platform windows and your books rarely agree on the same day.
Treat view-through attributed sales separately from click-through, because the two describe very different levels of buyer intent.
Avoid summing attributed sales across platforms into one revenue line. Each platform credits inside its own window, so adding ChatGPT Ads, search, and social totals double-counts shoppers who touched more than one channel and invents revenue you never earned.
AirOps: tracks how often ChatGPT cites and recommends your brand organically, so you can see the AI visibility that primes the paid purchases you later attribute.
OpenAI Ads Manager: reports attributed sales, return on ad spend, and conversions for ChatGPT Ads at campaign, ad group, and ad level, with configurable attribution windows.
Google Analytics 4: captures assisted conversions and AI-referred sessions, giving you a second read when native attribution files ChatGPT traffic under direct.
Audit your event values. This week, open a recent order_created payload and confirm the event-level amount carries the full order total in the correct currency. It needs no budget, just a developer console and ten minutes.
Install server-side tracking. Add the Conversions API alongside the measurement pixel so purchases report even when a browser blocks scripts or a session drops.
Set a deliberate window. Choose a click and view window that matches your real sales cycle, then document it so nobody quietly changes the setting later.
Reconcile against orders. Compare a full week of attributed sales to your order system and note the gap before you trust the number for spending decisions.
Tie it to bidding. Once attributed sales is accurate, feed order_created into conversion-optimized campaigns so the platform learns to bid toward real revenue.
Attributed sales is the purchase value an ad platform credits to your campaigns within an attribution window.
It is measured from order_created events that carry an amount and currency and match a prior ad click or view.
The attribution window caps what can be counted, so the same sale yields different totals on platforms with different windows.
Wrong event values or summed cross-platform totals quietly inflate or understate the figure you report.
The fastest gain comes from sending accurate order totals and reconciling against your own order system.
Attributed sales count only the purchase value an ad platform can tie to one of your campaigns, while total store sales count every order regardless of source. The two almost never match. Your order system records the full truth of what customers bought; the platform records a slice of that, limited to buyers it recognizes as having clicked or seen an ad inside the attribution window. Attributed sales can also exceed a short slice of total sales when two platforms both claim the same purchase, because each measures independently and neither knows about the other. Treat total sales as your financial ground truth and attributed sales as a channel's self-reported contribution. The gap between them is not an error to eliminate; it is organic demand, untracked sessions, and the limits of any one platform's view. Watch the direction of the gap over time instead of expecting the two numbers to reconcile on any given day.
Pick the window that matches how long your buyers actually take to decide, then keep it fixed. ChatGPT Ads lets you choose a click window of 7, 14, or 30 days and a view window of 0 or 1 day, applied to the conversion columns in reporting. A 7-day click window suits most everyday ecommerce, where purchases follow quickly. Longer consideration cycles, such as high-ticket or B2B purchases, justify 14 or 30 days so you do not lose credit for sales that close slowly. The view window captures people who saw an ad without clicking; set it to 1 day only if you want that softer signal reported separately. Changing the window affects reporting alone and leaves bidding or billing untouched, so a wider window will make attributed sales look larger without changing what you spent. The real rule is consistency: match the window across the platforms in your media plan, because otherwise the attributed-sales columns cannot be compared.
They are built from different inputs, so a gap is expected and never a bug. Attributed sales depends on three things lining up: the purchase event reaching the platform, that event matching a recognized ad click or view, and the match landing inside your window. Any purchase that misses one of those, such as a buyer who cleared cookies, used a different device, or converted after the window closed, never appears in attributed sales even though your order system logged the revenue. The reverse happens too: the figure can look inflated when the event sends a wrong amount, when a subscription auto-renews inside the window, or when a platform credits a sale that another channel also claims. Your books use one definition of a sale; the platform uses its own crediting convention. Instead of chasing an exact reconciliation, track the size and direction of the gap, and investigate when it swings sharply, because that usually points to a tracking break.
Partly, and the part you control is measurement itself, while the crediting stays fixed. You cannot change how the platform decides which sale belongs to which ad, and you should not try to, because that logic is fixed and last-touch. What you can influence is whether real purchases reach the platform accurately. Sending the correct order total in the event-level amount, firing order_created reliably from both the browser and the server, deduplicating with a shared event ID, and keeping your product feed values clean all determine whether attributed sales reflects true revenue. Widening the attribution window will also raise the reported figure, but that is a reporting choice and adds no real sales, so treat it carefully. The honest answer is that you influence attributed sales mainly by fixing your tracking and by running campaigns that genuinely drive purchases. Anything that nudges the number without more real revenue behind it is a vanity gain you will pay for at reconciliation.
There is no universal benchmark, and anyone quoting one is ignoring your margins. Attributed sales only means something next to spend, which is why return on ad spend, this value divided by spend, is the figure to judge. A good ROAS is one that clears your break-even after product cost, shipping, and fees, so a business with thin margins needs a far higher multiple than one selling software. Start by calculating the ROAS at which a campaign is profitable for you, then treat that as the floor. Because ChatGPT Ads is a newer surface, judge it against its own trend instead of against your mature search numbers: is attributed sales growing as you optimize, and is ROAS holding above your floor as you scale spend? Early campaigns often look weak until event tracking is clean, so confirm your measurement is accurate before you conclude the channel is underperforming. Direction and profitability matter more than any single target number.