Platform information: ChatGPT Ads are developing and platform requirements can change. This article was last reviewed on 11 September 2026. Shays Media rechecks material platform information before relying on it for client decisions.
Status tells you something is not serving. Diagnosis tells you why.
A campaign showing no delivery is a status, not an explanation. It's tempting, when nothing's happening, to start changing things — the budget, the ad copy, the context hints, all at once — in the hope that something in that pile of changes fixes it. That's a bad way to run any advertising account, and it's a particularly bad way to run one on a platform still in beta, where the actual cause is often something specific and checkable rather than a vague "the algorithm doesn't like it."
The better approach is to work through causes in a sensible order, identify the actual issue, and fix that one thing. If you change everything at once and delivery starts, you won't know which change fixed it — which means you won't know what to protect the next time you touch the account.
We can't diagnose your specific account from a general article. What follows is the order worth checking, not a diagnosis of what's wrong with yours.
1. Account verification and setup
Before anything else: is the advertiser account itself fully verified and set up? Campaigns won't deliver until required account setup and review is complete. If verification is still pending, or account information is incomplete, that's a plausible blocker worth ruling out before changing campaign settings.
2. Billing and payment
ChatGPT Ads uses postpay billing, which requires a billing profile and a valid payment method in place before delivery can happen. A missing or failed payment method, or a billing profile that wasn't fully completed, will stop delivery regardless of how well the campaign itself is built.
3. Account information completeness
Beyond verification and billing specifically, incomplete account information — a missing logo, an unconfirmed business detail — can sit as an open item that blocks progress further down the setup chain. Worth a full check, not just a glance.
4. Campaign, ad group and ad status
Each level of the hierarchy — campaign, ad group, ad — has its own status. It's possible for a campaign to be active while an ad group inside it is paused, or for an ad group to be live while the individual ad hasn't cleared review. Check status at all three levels, not just the top one, because a healthy-looking campaign can still contain a blocked ad group or ad underneath it.
5. Dates
A simple one, easy to overlook: check start and end dates on the campaign. A campaign that technically exists but hasn't started yet, or has already ended, will show no delivery for a reason that has nothing to do with anything else on this list.
6. Review and policy status
Check whether the relevant campaign and ad assets have completed any required review, and whether Ads Manager is showing a policy-related status. OpenAI's Advertising Policies govern what's eligible to run and where. A review or policy issue is a different problem from a technical setup issue and should be diagnosed from the status and reason shown rather than guessed at.
7. Budget, bid and delivery settings
Check that the daily budget is actually set at a level capable of delivering. Two things to rule out here: first, the minimum daily budget varies by currency (currently £15 for GBP accounts, $25 for USD), so double-check the account is above that floor for its currency; second, for campaigns running a fixed bid alongside a daily budget, the budget may also need to be at least the highest effective applicable bid, or delivery can be constrained even with an otherwise reasonable budget. Also remember that a daily budget is treated as an average over a seven-day period. Budget and bid settings can constrain delivery even when the rest of the setup is technically correct. This is separate from the question of whether the budget is commercially meaningful, which we cover in How Much Do ChatGPT Ads Cost in the UK?.
8. Landing page reachability and destination integrity
The destination page needs to be valid and reachable, and it must not be blocked from OpenAI's crawlers — verify the current exact crawler name against live documentation if you're checking robots rules directly. Beyond pure accessibility, OpenAI's current Ad Policies also require destination integrity: the landing page needs to be consistent with the advertiser, the creative and the offer end-to-end, not just technically reachable. A landing page that's down, behind a login wall, blocked by a robots directive, or that doesn't genuinely match what the ad promises can prevent delivery or trigger a policy issue even when every setting inside Ads Manager looks correct. This is easy to miss because part of the check lives outside Ads Manager — on your own website or hosting setup — so the destination itself needs inspecting rather than only the campaign settings.
9. Upload or format issues
Where relevant, confirm that uploaded creative assets — including the image — were accepted and meet the current format and specification requirements. Use the live Ads Manager status and current creative specifications rather than assuming an uploaded asset is valid simply because it is present in the account.
10. Reporting delay vs actual non-delivery
Finally, before concluding delivery has genuinely stopped, rule out the possibility that what you're looking at is a reporting delay rather than an actual delivery problem. Dashboards and reporting don't always update in real time — OpenAI's current launch guidance notes that reporting may lag ad delivery by up to around seven hours, a figure worth reconfirming at the point you rely on it, since it's exactly the kind of detail that can change. A campaign that's actually delivering can sometimes look flat in reporting for a period before catching up. This is worth checking last, not first, because it's tempting to assume it's "just reporting" and skip the checks above that might reveal a real, fixable issue.
As a general timing reference, current OpenAI launch guidance says active campaigns should begin receiving impressions and clicks within 24 hours — worth treating as current platform guidance rather than a delivery guarantee, and rechecking at the time, since it's the kind of expectation a beta platform can revise.
The principle behind the order
This list is ordered roughly from "most likely to block everything" to "most likely to be a narrower issue," but the actual order matters less than the discipline: identify the specific cause before making a change, and change one thing at a time where practical. If you jump straight to rewriting ad copy because delivery looks weak, and the actual cause was an incomplete billing profile, you've made an unrelated change, possibly a worse one, without fixing anything — and now you don't know whether your next attempt worked because of the copy change or in spite of it.
This is the same principle behind our own management approach more broadly: diagnosis before changing, one major variable at a time where practical, and documenting what was actually checked and changed. It's slower than firing off a batch of edits and hoping. It's also the only way to actually learn what was wrong.
When to bring in help
If you've worked through this list and delivery still isn't happening, or the review/policy status is the blocker and it's not clear what needs to change, that's a reasonable point to get a second set of eyes on the actual account rather than continuing to guess from outside it. Our Setup + Management service includes exactly this kind of diagnosis as part of ongoing monitoring, and it's also something we can look at directly if you're managing the account yourself and have hit a wall.
Check Your Fit if you're earlier in the process, or get in touch about Setup + Management if you have a live account with a delivery problem you'd like a second opinion on.
A few questions people ask at this stage
Why isn't my ad showing at all? There's no single answer — it could be account verification, billing, campaign or ad-group status, review/policy status, budget level, landing-page accessibility, or a reporting delay. The useful approach is working through causes in order rather than guessing at one and changing it.
Should I just change my budget, copy and targeting all at once to see if something works? No — that's the approach this article argues against. Changing several things at once means that even if delivery starts, you won't know which change actually fixed it, which leaves you no better equipped for next time.
Can you guarantee my ads will start delivering if I fix these things? No. Fixing an identified cause removes a known blocker; it doesn't guarantee any particular level of delivery, which remains subject to OpenAI's own review, policy and platform decisions.