Android app inventory is the set of sponsored ad placements that ChatGPT Ads can deliver inside the ChatGPT app on Android phones and tablets, shown below the assistant's answer. It is a distinct surface from Android web, where someone opens ChatGPT in a mobile browser and the placement counts as web inventory instead.
When you set platform targeting, choosing this surface decides whether your budget reaches Android app users at all. Skip that choice and your campaign runs across every surface by default, which hides how the Android app converts compared with iOS or the web and wastes spend on placements you cannot see clearly.
In ChatGPT Ads, Android app inventory is one of the platform surfaces an advertiser can target, represented by the value android_app in the campaign targeting object. You select it at the campaign level through platform targeting, OpenAI's control for choosing which ChatGPT surfaces a campaign can deliver on.
The android_app value maps only to the native ChatGPT app on Android. A person who opens ChatGPT in an Android browser falls under web inventory, which covers desktop and mobile web together. Platform targeting is optional, so a campaign with no platform set stays eligible on every surface, including the Android app.
Android app inventory sits alongside iOS app inventory and web inventory as the three broad platform groups. In the Ads Manager Insights view, OpenAI later split reporting into five labels, separating Android app from Android web. AirOps helps teams treat these surfaces as a measurable ad channel and tie their performance back to pipeline.
Android app inventory delivers through the same campaign setup as every other ChatGPT surface, with one added targeting choice. Here is how a placement reaches an Android app user.
Set targeting. In your campaign, add android_app to the included platforms, or leave platform targeting empty to stay eligible everywhere.
Match context. When an Android app user sends a prompt, the ads system reads the conversation's context and intent against your context hints, title, and landing page.
Enter the auction. Eligible ads compete in a relevance-weighted auction, and bid guidance signals whether your bid can win delivery.
Render the card. The winning ad shows as a sponsored card below the assistant's answer inside the Android app, labeled and separate from the response.
Report by platform. Insights segments results by surface, so you can read Android app performance on its own.
Segmenting by platform tells you how the Android app surface performs against iOS and web. It does not tell you why a user behaved differently, so pair it with landing-page and conversion data.
Platform targeting turns the Android app from a default assumption into a deliberate budget decision. Whether you treat Android app inventory as its own line item changes what you can measure and where your spend lands. For a media buyer weighing ChatGPT Ads against other channels, that choice decides whether you can prove the Android app pulls its weight or whether its results stay buried in a blended, all-surface number you cannot act on with confidence.
Device behavior varies. An Android app user checks out and fills forms differently from desktop, so blended reporting can hide a weak surface.
Budget control gets specific. Targeting android_app keeps spend on the users you want and off surfaces that convert poorly for you.
Measurement stays honest. Without platform segmentation, a weak Android app conversion rate hides inside a blended number, and you keep funding dead placements.
SEO managers use Android app inventory to isolate ChatGPT Android app traffic and compare its landing-page behavior against organic AI referrals.
Growth marketers use Android app inventory to shift budget toward the mobile app surface when it converts better than web or iOS.
Demand gen leads use Android app inventory to report ChatGPT Android app performance separately in channel reviews, instead of hiding it in a blended figure.
Platform targeting is the optional campaign-level control where you choose which ChatGPT surfaces can serve your ads, with Android app, iOS app, and web as the broad groups, each set through the targeting.platforms.included field in the campaign object or the Ads Manager dropdown.
Web inventory covers ChatGPT opened in any browser, whether desktop or mobile, which means a session in an Android phone's browser counts as web, while a session inside the installed ChatGPT app counts as Android app inventory, a split that trips up buyers who assume Android means the app.
Insights segmentation lets you break reporting out by surface, so Android app results sit on their own line, though OpenAI's documentation notes device reporting has at times grouped into broader buckets, which is why a dedicated analytics check still matters.
Isolate ad delivery to the ChatGPT Android app surface instead of every surface at once.
Compare Android app conversion against iOS app and web in the Insights report.
Protect budget from surfaces that convert poorly for your offer.
Target the android_app value directly through Ads Manager or the Advertiser API.
Read platform-level results to justify where ChatGPT Ads spend should go next.
Set platform targeting explicitly on every campaign, so delivery never defaults to surfaces you did not choose.
Keep android_app separate from iOS app early, because device behavior and conversion often differ.
Match landing pages to mobile app users, since app-to-browser handoffs can break form completion.
Segment Insights by platform before judging results, so the Android app is not hidden in a blended number.
Pair platform reporting with your own analytics, because OpenAI groups some device data into broad buckets.
Test the Android app surface on its own budget, so you can read its true cost per click.
Avoid treating Android app and Android web as the same thing. A browser session on an Android phone belongs to web inventory, so a campaign set to android_app alone will skip every Android user who reaches ChatGPT through a browser, and your results will understate Android reach.
AirOps: Plans and measures AI ad surfaces like the ChatGPT Android app as a growth channel, tying platform-level delivery back to pipeline and revenue.
ChatGPT Ads Manager: OpenAI's console where you set the android_app platform target and segment Insights to read Android app performance.
Google Analytics 4: Measures how traffic from the ChatGPT Android app behaves after the click, so you can see conversion differences by device.
Audit your current targeting. Open your existing ChatGPT Ads campaigns and check whether platform targeting is set on each one. An empty field means you are already serving on the Android app by default, bundled with every other surface.
Separate the surfaces. Create or edit a campaign to include android_app on its own, so Android app delivery stops sharing a budget with iOS and web.
Align the landing page. Confirm your destination page loads fast and completes forms cleanly for someone arriving from a mobile app session, since a slow page wastes Android app clicks you paid for.
Turn on platform reporting. In Insights, segment by platform so Android app results report separately from the start, giving you a clean baseline.
Read and reallocate. After enough delivery, compare Android app cost and conversion against other surfaces, then move budget toward the surface that earns it.
Android app inventory is the sponsored placement surface inside the ChatGPT app on Android devices.
You reach it by adding the android_app value to a campaign's platform targeting in Ads Manager or the API.
The main constraint is that an Android browser session counts as web inventory, while only the native app counts here.
The main risk is leaving platform targeting empty, which buries Android app results in a blended number.
The payoff comes from reading each surface on its own, then moving budget to the one that converts.
The difference comes down to which surface the user is on. Android app inventory means someone is inside the installed ChatGPT application on an Android phone or tablet. Android web means that same person opened ChatGPT in a mobile browser, and ChatGPT Ads files that session under web inventory, which also includes desktop browsers. The platform targeting value android_app reaches only the installed app. Select android_app alone and your campaign will skip every Android owner who uses the browser version. That catches many buyers off guard, because they assume Android points to the phone itself when it means the installed app. When OpenAI expanded Insights reporting, it eventually broke these out as separate labels for Android app and Android web, so you can read each one. The practical move is to decide whether you want the app audience, the browser audience, or both, then set platform targeting to match that intent.
Set it as soon as you have a reason to treat the mobile app audience differently from the web. A new advertiser testing ChatGPT Ads can leave platform targeting open at first, collect a week or two of delivery, then segment Insights to see how the Android app compares. Turn on android_app targeting once the data shows a real gap in cost or conversion between surfaces, or when your landing page clearly performs better for app users than browser users. There is no fixed schedule from OpenAI for this, so let your own numbers set the timing. Revisit the setting whenever you change creative, budget, or landing pages, since those often shift how each surface performs. Also check it after any ChatGPT Ads product update, because OpenAI has changed how platform targeting and reporting work more than once since launch, and a change can reset assumptions you made earlier.
Performance varies because the person, the device, and the path to conversion all change with the surface. An Android app user often sits in a different mindset from someone on a desktop browser, where research and multitasking are easier. The handoff matters too. Tapping an ad inside the app can push a user into an in-app browser or out to a separate one. Every extra step risks a drop before the form or checkout loads. Screen size, autofill behavior, and payment options differ across Android devices, which shapes whether a mobile visitor finishes what they started. Your own funnel adds more variance, since a page built for desktop can frustrate an app user. None of this is unique to ChatGPT Ads, but the single-surface nature of a conversational placement makes the differences easy to miss. Segmenting by platform is how you turn that variance into something you can read and act on.
You cannot set a precise budget split by surface. ChatGPT Ads lets you include or exclude the android_app surface through platform targeting, but it does not give you a slider to send a fixed share of spend to the Android app and the rest elsewhere. What you can do is control eligibility. Set a campaign to android_app only and all of its budget becomes eligible for the Android app surface. To weight surfaces differently, the common workaround is to run separate campaigns, one per surface, each with its own budget, so you decide how much each can spend. Automated bidding then optimizes within whatever surfaces a campaign is eligible for. Custom audience bid multipliers adjust bids for matched users, but they work at the audience level and do not key off the surface itself. So you influence the Android app's share mainly through how you structure and fund your campaigns, then let reporting guide where the money should move next.
There is no public benchmark for the Android app surface specifically, so treat good as better than your own alternatives. ChatGPT Ads is young, and OpenAI has not published surface-level norms for cost per click or conversion rate. OpenAI has suggested starting cost-per-click bids in a modest range for new campaigns, but that is general guidance that applies to all surfaces, with no Android app target attached. The useful benchmark is internal. Compare the Android app's cost per result and conversion rate against iOS and web in the same account, over the same period, with the same creative. An Android app that delivers results at a lower or similar cost to your other surfaces is earning its budget. When it lags after a fair test, trim its share and move spend to what converts. Hold the comparison steady over time, because a young platform shifts, and a single week rarely tells the full story.