Platform Targeting on ChatGPT Ads: the Setting Nobody Configures
By Keshav Parsai 6 min read
In brief
ChatGPT Ads targets five platforms and now reports by platform in Insights. Launch broad, read the split, then prune the surfaces that underperform.
Last verified: 2 October 2026 | Version: 2.0 | Next scheduled review: 2 January 2027
Five checkboxes sit in the campaign setup, all ticked by default, and almost nobody touches them. OpenAI's campaign documentation lists them plainly: Android app, Android web, desktop web, iOS app, iOS web.
The first version of this page argued that you had to decide these in advance, because reporting had no platform dimension and whatever you left on was fused into one number. That is no longer true. OpenAI's campaign documentation now says you can segment Insights by Platform to see results for each of the five. The setting still deserves a deliberate answer, but the answer has changed.
Two audiences, five labels
Start from what a surface indicates about the person on the other side of it.
Someone on desktop web is at a computer. For a B2B product that is overwhelmingly a working-hours context: a work machine, a work task, a work problem being thought through in a work session.
Someone in the iOS or Android app is on a phone, usually outside a desk context. That is where personal use concentrates: evenings, commutes, household admin and general curiosity.
Mobile web sits awkwardly between the two and is the least interpretable of the five. It is a phone, but in a browser, which on iOS often means a link opened from elsewhere rather than a session someone started deliberately.
That is a claim about context, not performance. A B2B advertiser with all five surfaces on is buying two materially different populations under one bid and one budget. The difference now is that you can see each population's results afterwards.
What platform reporting changes
Before, the only way to read performance by platform was to split surfaces into separate campaigns before you had any evidence. With a Platform segment in Insights, the sequence used on Google and Meta becomes available: leave the surfaces on, let the campaign accumulate spend, read the breakdown, then prune.
Two things have not changed.
Targeting is still campaign level. You can see the split per platform, but bid and budget are shared across every surface the campaign includes. If desktop converts and app traffic does not, the fix is to edit the campaign's platform selection or move one surface into its own campaign, not to bid it down.
A segment only reads if each slice has enough data. On a channel where a campaign often struggles to spend 25 USD a day, five slices of a small budget may each be too thin to mean anything. Read platform splits at the same sample thresholds you would use for any other decision.
What to actually do
For a first campaign, start with all five surfaces on unless you have a specific reason not to. Underdelivery is the most common failure on this channel, and restricting surfaces cuts your match surface before you know whether the campaign can spend at all.
Once the campaign has delivered enough for the platform segment to read, look at cost per qualified lead or conversion by platform, not just CTR. If one group of surfaces is clearly worse, remove it or move it into a separate campaign with its own budget.
The exception is a B2B advertiser whose conversion event happens almost entirely at a desk, such as a demo request that needs a work email. Starting on desktop web alone is still defensible there, as a deliberate choice rather than a default you never revisit.
This is reasoning, not measurement
OpenAI publishes no performance data by platform, and no advertiser has published a controlled comparison. The claim that desktop web carries a cleaner B2B context comes from how people generally use computers and phones, not from observed ChatGPT Ads results.
It is equally possible that surface matters less here than on display channels, because matching happens on conversation meaning. Someone researching compliance software in the iOS app at nine in the evening may be exactly the buyer you want. The Platform segment is how you find out for your own account.
What we cannot tell you
- Whether any platform performs better than another. OpenAI publishes no performance data by platform and no advertiser has published a comparison.
- How ChatGPT usage divides between app and web. OpenAI has not published a surface breakdown of its user base.
- Whether platform targeting will move to ad group level. Not on any published roadmap.
- What mobile web traffic represents. Still the least interpretable of the five surfaces.
Quick answers
What platform targeting options does ChatGPT Ads have? Five: Android app, Android web, desktop web, iOS app and iOS web. You select one or more at campaign level.
Can I see performance by platform? Yes. OpenAI's documentation says you can segment Insights by Platform to see results for each of the five. Earlier guidance, including the first version of this page, said this was not possible.
Which platform should a B2B advertiser choose? Usually all five to start, then prune from the Platform segment once each slice has enough data. Desktop web only is defensible when the conversion almost always happens at a desk.
Does restricting platforms hurt delivery? It reduces your match surface, and underdelivery is already the common failure on this channel.
Is platform targeting set per ad group? No. It sits at campaign level, alongside objective, budget, dates, locations and custom audiences.
Sources
| Claim | Source | Tier |
|---|---|---|
| Five platforms, selected at campaign level | OpenAI Help Center, Create Campaigns for ChatGPT Ads, 2 Oct 2026 | Confirmed, primary |
| Insights can be segmented by Platform | OpenAI Help Center, Create Campaigns for ChatGPT Ads, 2 Oct 2026 | Confirmed, primary |
| Minimum daily budget of 25 USD in the US, varying by billing currency | OpenAI Help Center, Create Campaigns for ChatGPT Ads, 2 Oct 2026 | Confirmed, primary |
| No published performance data by platform | OpenAI documentation, by absence | Absent |
| Desktop web as the cleanest B2B context | Newtation reasoning from usage context | Inference, ours |
Related reading
- How to Structure a ChatGPT Ads Account When There Are No Keywords
- Location Targeting Inside ChatGPT Ads: Country, Region, City, DMA and Postal Code
- How to Measure ChatGPT Ads, and What You Cannot Measure Yet
- How Many Ad Groups Should One ChatGPT Campaign Have?
Changelog
2 October 2026, v2.0. OpenAI now documents a Platform segment in Insights. Withdrew the central claim that platform results could never be separated, and changed the recommended default from desktop-only to launch broad and prune from the segment.
12 September 2026, v1.0. First publication.
Field kit
Tools
Site
OpenAI (primary)
Share this
Keshav studies how AI systems retrieve, verify, and cite brand information. At InPromptAds, he leads source research and turns platform documentation into practical guidance for advertisers.
