Platform targeting is the ad-campaign setting that controls which devices, operating systems, and surfaces your ChatGPT ads are eligible to serve on, such as desktop web, mobile web, the iOS app, and the Android app. It governs where an ad can appear, separate from contextual targeting, which decides which conversation triggers it.
When you decide where to spend a ChatGPT Ads budget, platform targeting is the lever that keeps your spend on the surfaces your customers actually use and your landing pages actually support. Ignore it and you pay to show mobile-only offers to desktop users, or send app traffic to pages that break on a small screen.
In ChatGPT Ads, platform targeting defines the inventory your campaign can buy by filtering the ad auction down to the device types and app environments you select. You might run only on the ChatGPT mobile apps, only on desktop web, or on every surface at once. The setting decides eligibility before any conversation context or bid is considered.
Platform targeting works through a few selectable dimensions: operating system (iOS, Android), device class (mobile, tablet, desktop), and the surface where ChatGPT is accessed (native app versus browser). Each ad request carries signals about the user's device and environment. Your campaign only enters the auction when those signals match the platforms you turned on.
This sits alongside contextual targeting and audience targeting in a ChatGPT Ads campaign: contextual targeting reads the conversation, audience targeting matches known users, and platform targeting narrows the hardware and software your ad reaches. AirOps helps marketers track how brand mentions and referral traffic differ across these surfaces so platform choices are grounded in where AI answers actually convert.
Platform targeting runs as a filter layer inside the ChatGPT Ads auction. Before any ad is scored for relevance, the system checks the user's device and surface against the platforms your campaign allows.
Set scope: In the campaign setup, you select the operating systems, device types, and surfaces the ad may serve on.
Read the request: When a user opens a conversation, each ad opportunity carries device and environment signals.
Match and filter: The system compares those signals to your selections and drops the campaign from opportunities that do not match.
Enter the auction: Eligible opportunities move forward, where contextual and bid factors decide whether your ad wins the placement.
Log the surface: Each served impression records the platform it ran on, feeding your reporting.
The output tells you which surfaces delivered impressions, clicks, and conversions. It does not tell you whether a different platform mix would have performed better, so run a controlled test before you lock one in.
Where your ad runs decides whether your ChatGPT Ads budget reaches people who can act on it. Platform targeting turns that into a deliberate choice instead of leaving it to the auction defaults. For a channel this new, that control is how you keep early test spend honest and readable.
Budget efficiency: Serving only on the surfaces that convert stops you from paying for impressions on devices where your offer cannot close, which matters while CPMs on a young channel stay high.
Experience control: A mobile app user and a desktop user see different layouts, so matching the platform to a landing page that renders well protects your click-through and conversion rates.
Failure prevention: Without platform targeting, a campaign built around an iOS-only app install can still spend on Android users who hit a dead end, burning budget with no path to conversion.
Growth marketers use platform targeting to concentrate app-install campaigns on iOS and Android so budget only reaches users who can download the app.
Demand gen leads use platform targeting to keep desktop-heavy B2B offers on desktop web, where buyers research and forms convert.
Content strategists use platform targeting to align promoted articles with the surface whose layout reads best for long-form content.
A surface is where ChatGPT is used, such as the native app or a web browser, while a device is the hardware class like phone, tablet, or desktop; platform targeting lets you filter on both, and confusing the two leads to gaps where your ad never serves.
Platform targeting is a gate applied before the auction, so a tight selection shrinks the pool of impressions you can win; widen it to gather data early, then narrow once you see which surfaces actually drive the conversions you care about.
Because the same person can meet your brand on a phone app and later on desktop, judge platform performance by conversions attributed across surfaces instead of raw impression counts on any single one.
Direct your spend to the surfaces where your offer actually converts.
Protect conversion rates by matching ads to landing pages that render on the user's device.
Serve iOS and Android app-install ads only to users who can complete the download.
Read performance by surface so you can shift budget to what works.
Reduce wasted impressions on devices your campaign cannot serve.
Start broad across all surfaces, because early data shows you where conversions actually come from before you restrict spend.
Match each campaign's platform to its landing page, so users reach a page built for their device.
Separate iOS and Android into their own ad groups, since app behavior and conversion rates differ between them.
Check that tracking fires on every surface you enable, because a surface with broken measurement will look worse than it is.
Revisit platform settings after each reporting cycle, so your mix follows real performance as the channel matures.
Keep creative sized for the surface, because a desktop card and a full-width mobile image demand different assets.
Avoid the common mistake of copying a desktop campaign straight onto mobile surfaces and expecting the same result. The layouts, attention spans, and conversion paths differ enough that an unadjusted carry-over quietly drains budget while the dashboard still looks busy.
AirOps: Tracks how your brand's mentions, citations, and referral traffic from ChatGPT vary by surface, so you can see which platforms your paid and organic AI presence actually reach.
OpenAI Ads Manager: The native campaign tool where you set and adjust platform targeting for ChatGPT Ads, including device and surface selections.
Google Analytics 4: Measures how traffic from each platform behaves on your site after the click, from landing-page engagement to conversion.
Audit your surfaces: This week, check how your landing pages render on desktop, mobile web, and inside a mobile app view. Note any that break or load slowly. A broken mobile page is the fastest way to waste ad clicks.
Map offers to platforms: Decide which of your offers genuinely fit each surface, such as app installs for mobile and demo forms for desktop.
Set up tracking: Confirm your analytics and conversion tags fire correctly on every platform you plan to run, so the data you collect is trustworthy.
Launch broad, then read: Start a campaign across all relevant surfaces and let it gather a representative sample before you judge any single platform. Resist cutting a platform before it has had a fair share of impressions.
Tighten the mix: Once the data is clear, narrow platform targeting to the surfaces delivering your target cost per conversion and reallocate the freed budget.
Platform targeting decides the devices, operating systems, and surfaces where your ChatGPT ads can appear.
You set it in the campaign, and it filters each ad opportunity before the auction scores relevance for a placement.
The main constraint is reach: the tighter your platform selection, the fewer impressions your campaign can win.
The main risk is paying to reach devices where your offer or landing page cannot convert, no matter the bid.
The leverage comes from matching each offer to the surface that converts it and shifting budget there.
Platform targeting and contextual targeting answer two different questions: platform targeting decides the device and surface an ad can run on, while contextual targeting decides which conversation makes the ad relevant. In ChatGPT Ads, contextual targeting reads the topic and intent of the chat happening right before the ad, matching your campaign to the subject a user is actively exploring. Platform targeting never looks at the conversation at all. It only checks whether the user is on an eligible operating system, device class, and surface, and it applies that check before contextual relevance or bidding comes into play. You use them together: platform targeting narrows the pool of eligible impressions by hardware and software, and contextual targeting then selects the moments within that pool worth bidding on. Setting one well does not replace the other, and leaving either on its default can quietly undercut an otherwise sound campaign.
Review platform targeting at the end of every reporting cycle, which for most campaigns means weekly while you are still learning and every two to four weeks once performance stabilizes. Early on, you need frequent checks because a new channel gives you little prior data, and the first weeks are when you discover that one surface converts far better or worse than you assumed. Look at cost per conversion and conversion rate broken out by platform instead of blended totals, so a strong desktop result does not hide weak mobile spend. When you change a platform setting, give the campaign enough new impressions to produce a readable sample before you judge the change, since reacting to a day or two of thin data usually sends you chasing noise. As the channel matures, you can space reviews out, but never set platform targeting once and forget it.
Performance varies between iOS and Android because the two platforms attract different users, support different purchase flows, and carry different costs to reach. Spending power, app-store behavior, and the ease of completing a purchase or install all tend to differ across the two ecosystems, so the same creative and offer can convert at very different rates. Your own funnel adds more variance: if your app or checkout is more polished on one operating system, that platform will quietly outperform regardless of the ad itself. Measurement also differs, since tracking and attribution can behave differently across devices and browsers, which can make one platform look stronger or weaker than it really is. The practical response is to treat iOS and Android as separate lines in your campaign, give each its own budget and reporting, and judge them on their own cost per conversion before you decide where the next dollar should go.
Yes, platform targeting is one of the few parts of ChatGPT Ads you control directly, since you choose the operating systems, device classes, and surfaces your campaign is eligible to run on. This is a deliberate setting in the campaign, so unlike the conversation context that triggers an ad, it is fully in your hands. What you cannot control is which specific chat your ad lands in or whether an eligible user ever sees it, because contextual relevance and the auction still decide that. Your influence is at the eligibility level: you open or close whole surfaces, and the system respects those boundaries. Use that control carefully, because closing a platform removes every impression it could have delivered, including the ones that might have converted well. The better approach is to open surfaces broadly first, gather real performance data, and then use your direct control to tighten the mix toward what works.
A good platform targeting setup is one where every surface you run on has a clear reason to be there and a tracked path to conversion. There is no universal benchmark percentage, because the right mix depends on your offer: an app install belongs on mobile operating systems, while a complex B2B demo often belongs on desktop web. Judge the setup by outcomes instead of breadth. Strong signs are a cost per conversion that holds steady or improves as you narrow platforms, reporting that is clean and separated by surface, and landing pages that render and convert well everywhere you serve. Warning signs are impressions piling up on a platform with no conversions, or a blended cost per conversion that looks fine only because one strong surface is masking a wasteful one. In short, a good setup is deliberate, measurable, and regularly pruned, so spend keeps flowing to the surfaces that actually earn it.