A platform breakdown is a report in ChatGPT Ads Manager that segments your campaign performance by the ChatGPT surface where each impression, click, and conversion happened, such as the iOS app, Android app, or web. It differs from platform targeting, which decides where your ads are eligible to run; a platform breakdown tells you what actually happened across those surfaces after the campaign delivered.
If you read ChatGPT Ads results as one blended number, you will keep funding surfaces that lose money and starve the ones that convert. A platform breakdown shows where spend earns its keep, so you can shift budget toward the environment that produces your cheapest conversions.
Platform breakdown reporting is the Insights view in ChatGPT Ads Manager that splits every performance metric by the surface a user was on, so impressions, clicks, spend, and conversions are attributed to the iOS app, Android app, or web instead of one pooled total.
The breakdown is built from the platform value attached to each ad event. When your ad serves inside the native iOS or Android app, that event is tagged to the app group; when it serves in a browser on a phone or laptop, it is tagged to web. OpenAI has added granular web reporting, so the breakdown can separate desktop web, iOS web, and Android web while preserving the historical combined web total.
Platform breakdown sits downstream of platform targeting and alongside your other Insights segments, such as country and device. It describes delivery that already happened and does not change where future ads run. AirOps helps teams fold this platform-level data into a repeatable reporting workflow.
A platform breakdown runs the same way each time you open Insights: it pulls delivery data and groups it by surface.
Tag the event: Each impression, click, and conversion carries the platform it occurred on, set the moment your ad serves on a given ChatGPT surface.
Group by surface: Insights rolls those tagged events into platform groups, combining the iOS app, Android app, and web, or splitting web into desktop, iOS web, and Android web.
Attach metrics: The report attaches impressions, clicks, spend, and conversions to each group, so every surface carries its own cost and outcome.
Segment the view: You apply the Platform segment in the overview screen to see each surface side by side.
Compare and act: You read cost per conversion and return by surface, then move budget or adjust platform targeting for the next flight.
The breakdown tells you which surfaces earned your results and at what cost. It does not tell you why a surface underperforms; for that, look at landing pages, creative, and device behavior.
Resources: AirOps covers practical ways to plan and measure ChatGPT Ads campaigns.
Where your ChatGPT Ads budget goes is a buying decision, and a platform breakdown is the evidence behind it. Without it, you approve the next flight's spend on a blended average that hides which surfaces actually pay back. Treat it as the report you open before renewing or cutting a campaign.
Budget allocation: The breakdown shows cost per conversion by surface, so you can move spend toward the app or web environment that returns the most for each dollar.
Blind overspending (failure mode): If you ignore the breakdown and the iOS app quietly runs a conversion rate half that of web, you keep paying app-level costs for web-level results until the quarter closes.
Creative and landing-page fit: Surface-level data tells you whether an app audience and a browser audience respond differently, so you can brief creative and landing pages for the environment that drives them.
SEO managers use a platform breakdown to see whether ChatGPT web traffic converts better than app traffic before they invest in surface-specific landing content.
Content strategists use a platform breakdown to decide which creative and ad copy to prioritize for app audiences and browser audiences.
Growth marketers use a platform breakdown to reallocate daily budget toward the surface delivering the lowest cost per conversion each week.
Every metric in the breakdown depends on the platform tag assigned when your ad served, so clean attribution is what makes a surface-by-surface comparison trustworthy; if tagging is wrong, the whole report misleads you no matter how good the campaign looked.
The native app and the browser are genuinely different environments for checkout and form completion, so a gap between them in the breakdown usually reflects real behavior you can act on, especially for spend and conversion decisions.
ChatGPT Ads has expanded platform reporting over time, so the number of surfaces you can separate in the breakdown may differ from the ones you can target, and reading them as equivalent will distort your conclusions.
Pinpoint which ChatGPT surface delivers your cheapest conversions.
Shift budget toward the iOS app, Android app, or web based on evidence.
Catch an underperforming surface before it drains a full flight's budget.
Brief creative and landing pages for how app and browser audiences actually behave.
Align your reporting dimensions with the platform targeting controls in ChatGPT Ads Manager.
Report surface performance to stakeholders with numbers they can act on.
Check the breakdown on a fixed cadence, weekly or per flight, so budget shifts are based on current data instead of a one-time look.
Wait for enough conversions per surface before you judge it, because thin data makes one surface look better or worse than it is.
Pair the breakdown with your platform targeting so the surfaces you analyze match the ones you actually buy.
Read cost per conversion and return by surface instead of click volume alone, so you optimize toward revenue outcomes.
Separate web into desktop and mobile where the data supports it, so you can see whether browser performance hides two different audiences.
Avoid splitting campaigns by platform too early. Competent buyers often launch separate iOS, Android, and web campaigns before any surface has enough conversions, which fragments the data and slows the learning you built the breakdown to produce.
AirOps: builds repeatable reporting workflows that pull your ChatGPT Ads platform breakdown into a single view so surface performance feeds your next budget decision.
OpenAI Ads Manager: the native Insights screen where you apply the Platform segment to see impressions, clicks, spend, and conversions for each surface.
Google Analytics 4: connects downstream site and conversion behavior to the surface a user arrived from, adding context the in-platform breakdown alone cannot show.
Open Insights: This week, open ChatGPT Ads Manager, go to the overview screen, and apply the Platform segment to a live campaign. No budget approval needed. This gives you the raw view before any analysis.
Confirm your surfaces: Check which platforms appear in the breakdown, and note whether web is split into desktop and mobile or reported as one group.
Add the core metrics: Pull impressions, clicks, spend, and conversions into the view so each surface shows cost and outcome side by side.
Set a benchmark: Record cost per conversion and return for each surface over a comparable window, so later reads have a baseline to beat. A baseline turns future reads into decisions instead of guesses.
Act on the gap: Once a surface has enough conversions, move budget or adjust platform targeting toward the environment that performs, then recheck on your next cadence.
A platform breakdown segments ChatGPT Ads performance by the surface each impression, click, and conversion came from.
You produce it by applying the Platform segment in Insights, which groups metrics into the iOS app, Android app, and web.
The surfaces you can analyze may be fewer than the surfaces you can target, since reporting granularity has lagged buying controls.
The biggest risk is splitting campaigns by platform before any surface has enough conversions to judge.
The biggest gain is in budget: moving spend toward the surface with the lowest cost per conversion is where the breakdown pays off.
A platform breakdown and platform targeting work at opposite ends of the same campaign. Platform targeting is a setting you choose before a campaign runs, telling ChatGPT Ads which surfaces, the iOS app, Android app, or web, are eligible to show your ads. A platform breakdown is a report you read after delivery, showing how each of those surfaces actually performed on impressions, clicks, spend, and conversions. One controls where ads can appear; the other explains where your results came from. They work together: you set targeting to open or restrict surfaces, then use the breakdown to see which of the surfaces you allowed earned their spend. If the breakdown shows one surface consistently underperforming, that finding feeds back into your targeting decision for the next flight. Reading them as a loop, buy, measure, adjust, is how you turn surface data into steadily cheaper conversions over time.
Check the platform breakdown on a regular cadence tied to how fast your campaign spends. For most campaigns, a weekly read is enough to catch a surface drifting without reacting to daily noise that reflects normal variation. If you run short flights or spend heavily each day, move to every few days so you can act before a budget is gone. The key constraint is conversion volume: a surface needs enough conversions in the window you are reviewing for its cost per conversion to mean anything, so avoid judging a surface after only a handful of events. Pair your cadence with a fixed lookback window, such as the trailing seven or fourteen days, so each read compares like with like. When you make a change, give the campaign time to stabilize before the next read, because optimization and attribution both need a few days to settle into reliable numbers.
Surfaces vary in your platform breakdown because the people and the experiences behind each one are genuinely different. A user in the native iOS app is often mid-task and signed in, while someone reaching ChatGPT through a desktop browser may be researching across tabs with a different intent. Those differences show up as different click-through rates, conversion rates, and average order values across surfaces. The environment matters too: checkout flows, form length, and app-to-browser handoffs can each add friction on one surface and not another, so the same ad produces different outcomes. Audience composition plays a part, since app users and web users are not the same population. Attribution settings and conversion volume also shape what you see, because a low-volume surface swings more from a handful of conversions. None of this means a surface is broken; it means the breakdown is doing its job by exposing real variation you would otherwise average away.
You cannot change the numbers a platform breakdown reports, but you can change what it has to measure. The breakdown is a mirror of delivery; it reflects what already happened and does not take instructions. What you can influence is everything upstream of it. Adjust platform targeting to open or restrict the surfaces that feed the report, and the mix of surfaces in your next breakdown changes with it. Improve the creative, landing pages, and offer for a given surface, and that surface's conversion rate can improve in the following report. Shift budget toward a surface and it gathers more volume, which makes its numbers more stable and more trustworthy. So the honest answer is that you influence the breakdown indirectly, through the levers that shape delivery, never by editing the report itself. Treat a weak surface as a prompt to fix the experience or move the money, then read the next breakdown to confirm the change worked.
A good platform breakdown result is one where each surface you keep funding meets your campaign's cost-per-conversion or return target, and your spend is concentrated on the surfaces that clear it. There is no universal number, because a good cost per conversion depends on your margins, your offer, and your market. The useful benchmark is relative: compare each surface against your own blended target and against its own trend over previous flights. A healthy pattern shows one or two surfaces carrying most of the efficient conversions, with weak surfaces either improving after you act or getting less budget. Watch the spread between your best and worst surface; a very wide gap is an opportunity to reallocate, while a narrow gap means your targeting is already well matched to performance. Also check that each surface has enough conversions to trust, because a surface that looks excellent on five conversions has told you almost nothing yet. Progress over time beats any single snapshot.