Back to Insights

ChatGPT Ads Just Landed in Europe: What It Actually Means for Your Brand

September 1, 2026|12 min read|Kilian Dreher

On August 24, 2026, OpenAI expanded ChatGPT Ads into 31 European countries, Denmark, Germany, France, Spain, Italy, Sweden, Norway, the Netherlands, and Austria among them. A week later, on August 31, self-serve access to ChatGPT Ads Manager opened for advertisers based in Europe, meaning you no longer need an agency partner or a direct line to OpenAI's sales team to run one. OpenAI puts weekly ChatGPT usage north of 1 billion people, and for the first time, European advertisers can put a paid message in front of a meaningful slice of them without leaving ads.openai.com.

That's the news. The more useful question is what it actually means if you run paid media for an e-commerce or DTC brand: is this a new channel worth budget, or a shiny distraction from the two platforms that already work? We run a data-driven Meta Ads scaling playbook for e-commerce clients every day, and our honest read is: it's neither a replacement for Meta nor Google, and it's not nothing either. Here's what's actually true about the format, the targeting, the pricing, and whether it's worth your time this quarter.

Table of Contents


What Is ChatGPT Ads Manager, Exactly?

ChatGPT Ads Manager is OpenAI's self-serve advertising platform, live at ads.openai.com. It lets you build, launch, and track campaigns that show a single sponsored card underneath a relevant ChatGPT answer, not inside it. OpenAI has been explicit on this point: ads don't influence what the model says, they're served by a separate layer after the response is generated, and every unit carries a clear "Sponsored" label.

The format is deliberately narrow compared to what you're used to on Meta or Google:

ElementSpec
Advertiser nameText, brand name
Favicon or logoSmall icon, brand mark
Title16 to 24 characters
Copy line32 to 48 characters
ImageOne 1:1 square image
Landing URLDirect link to your page

That's it. No carousels, no video, no multiple headline variations rotating for you to test. One card, tight character limits, one square image. It rewards clear product photography and a title that reads like an answer, not a slogan, because the user just got a helpful response and your ad needs to feel like a continuation of that, not an interruption.


The Europe Launch: What Actually Changed

ChatGPT Ads didn't appear out of nowhere. The rollout has moved in clear stages:

  • February 2026: Closed US pilot. Minimum spend around $200,000, CPMs around $60. Enterprise-only, invite-based.
  • May 5, 2026: Self-serve Ads Manager opens in the US at ads.openai.com, no minimum spend.
  • May 12, 2026: Product feed ads ship, letting retailers connect a catalog instead of building creative per product.
  • August 24, 2026: Expansion to 31 European markets, including Denmark and most of the DACH and Nordic region.
  • August 31, 2026: Self-serve Ads Manager access opens for European advertisers, the point where this stopped being a US-only story.

The six-month gap between the US pilot and the European rollout tracks OpenAI's usual pattern: prove the ad layer works and doesn't tank trust in the US market first, then extend it once the mechanics (auction, privacy controls, labeling) are stable enough to ship into GDPR-governed markets. If you're based in Europe, this is the first point at which you can actually open an account and spend without going through a partner.


Who Actually Sees These Ads

This is the detail most coverage of the launch skips past, and it changes how you should think about reach: ChatGPT Ads only show to users on the Free and Go tiers. Plus, Pro, Business, Enterprise, and Edu subscribers see zero ads, at any spend level.

Two implications worth sitting with. First, your addressable audience is structurally different from a paid subscriber base, you're reaching people who haven't (yet) decided ChatGPT is worth paying for, which skews toward lighter, more casual usage rather than the power users running it as a daily work tool. Second, OpenAI has an obvious incentive to keep growing that free-tier pool rather than converting all of it to paid, since it's now a monetizable audience in its own right. Neither of these makes the channel bad. They just mean "1 billion weekly users" isn't the number that matters for your media plan, the free-and-Go segment of that is.


How ChatGPT Ads Actually Work: Targeting, Bidding, and Pricing

This is where ChatGPT Ads genuinely diverge from what you know. There's no cookie-based retargeting, no pixel, no lookalike audience built from your existing customer list. OpenAI matches ads to the live, real-time context of the conversation itself, what you're actually asking about right now, not a profile built from your last three months of browsing. A user asking about running shoes for flat feet can surface a relevant ad; the same user asking about running shoes for flat feet last Tuesday doesn't get retargeted with the same product today unless they're back in that context.

Auctions run second-price, the same mechanic Google popularized: you don't pay your maximum bid, you pay just above the next-highest bidder. Campaigns support both CPC (cost-per-click) and CPM (cost-per-thousand-impressions) bidding, plus geo-targeting. Pricing has moved fast since February:

  • Closed pilot: ~$60 CPM, $200,000 minimum spend
  • Current self-serve: roughly $25 CPM, or $2 to $8 per click
  • OpenAI's suggested starting point: a maximum CPC around $3 to $5
  • Minimum spend on self-serve: none

That CPM is still meaningfully above Meta's typical average, which is one reason letting AI run your Meta budget autonomously is a different conversation from letting an algorithm bid into a brand-new auction with thin historical data. Early auctions on any new ad platform tend to be under-competed, and under-competed auctions are usually where the best early CPMs live, but that window closes as more advertisers show up.


ChatGPT Ads vs. Meta Ads vs. Google Ads: What's Different, What's the Same

Google AdsMeta AdsChatGPT Ads
What it targetsSearch keywords, explicit intentInterest, behavior, lookalikesLive conversational context
Pricing modelCPC, keyword-auctionCPM/CPC, generally the cheapest of the threeCPM or CPC, currently ~3x Meta's average CPM
Creative formatText + extensionsImage, video, carousel, multiple formatsOne sponsored card: title, copy, 1:1 image
Cookies/pixels usedYesYes, extensivelyNo, contextual matching only
Attribution/measurementMature, multi-touchMature, pixel + CAPIEarly, no multi-touch attribution yet
Best-fit funnel stageHigh-intent, bottom-of-funnelDiscovery, top-of-funnelDecision-stage, mid-funnel

The useful mental model: Google owns the moment someone searches for something they already know they want. Meta owns the moment someone discovers they want something they didn't know existed. ChatGPT Ads sit in a third moment, someone is actively working through a decision inside a conversation, comparing options, asking follow-up questions, and your product shows up as part of that reasoning process rather than interrupting a scroll or capturing a keyword.

What's the same: it's still an auction, it's still pay-per-click-or-impression, and creative quality still matters more than the platform wants to admit. What's genuinely different: no behavioral targeting to lean on, a single rigid ad format instead of a dozen placements to test, and attribution that's nowhere near what Google Ads or Meta's Conversions API gives you today. If your reporting stack depends on clean multi-touch attribution, budget for that gap explicitly, don't assume you'll get a clean answer on what a ChatGPT ad actually drove.


The E-commerce Angle: Product Feed and Shopping Ads

The part of this launch that matters most for DTC and e-commerce specifically shipped back on May 12: product feed ads. Instead of building a sponsored card for every SKU by hand, you connect your product catalog and OpenAI generates ads automatically from your existing product names, images, prices, and attributes, at a scale OpenAI says supports up to 1 million SKUs per advertiser.

That's the detail that makes this worth a second look if you run a catalog with real depth. Manual creative-per-product was the actual blocker to running this at any scale before May, not interest from advertisers. If your Shopify feed is already clean (accurate titles, decent product photography, current pricing), turning that into eligible ChatGPT ad inventory is close to a data-plumbing problem, not a creative production one. That's a meaningfully lower lift than what a new Meta or TikTok creative test cycle demands.


Should You Test It Now? A Founder's Framework

Three questions, in order:

  1. Do you have headroom in your top-of-funnel or branded-search budget, not your core Meta prospecting budget? ChatGPT Ads is a decision-stage channel with immature attribution. Pulling budget from a Meta campaign that's already converting at a known CAC to fund an unproven channel is the wrong trade. Pulling from a line you were already treating as exploratory is the right one.
  2. Can you tolerate not knowing exactly what it drove for four to eight weeks? Without multi-touch attribution, you're reading directional signals (branded search lift, direct traffic, coupon code usage) rather than a clean last-click number. If your reporting culture demands precision on day one, wait until OpenAI ships better measurement.
  3. Is your product catalog or landing page actually ready for a decision-stage visitor? Someone who reached your link from inside a ChatGPT conversation has already done more reasoning than someone who clicked a cold Meta ad. A weak PDP wastes that intent faster than it wastes cold traffic. If your conversion fundamentals aren't solid, fixing those will move the needle more than any new channel, and it's the same logic behind reducing e-commerce CPA generally: efficiency work compounds regardless of which channel you're testing.

If those three check out, a small capped pilot makes sense, not a reallocation. We're running exactly that on zentric.digital this month: a small, deliberately capped ChatGPT Ads test, so we can publish real setup screenshots, actual CPC and CPM numbers, and a step-by-step walkthrough instead of speculating from OpenAI's documentation. That follow-up is coming once we have first-hand numbers to show, not projections.


What This Means for Your 2026 Ad Budget

Don't restructure your media plan around this yet. Treat it the way we treat any new AI-native tool before it earns a permanent line in the budget: a small, bounded test with a clear kill criterion, run in parallel with the channels that are already working, not instead of them. That's the same discipline behind how we evaluate and adopt new AI tooling at Zentric more broadly, new doesn't mean ready, and ready doesn't mean it replaces what's already proven.

What is worth doing now, regardless of whether you spend a dollar on the platform: get your product feed clean. Accurate titles, current pricing, real product images. That work pays off whether it's feeding ChatGPT's product ads, Google's Merchant Center, or Meta's catalog, and it's the actual prerequisite for testing this channel well whenever you decide to.


Frequently Asked Questions

Q: What is ChatGPT Ads Manager?

A: ChatGPT Ads Manager is OpenAI's self-serve advertising platform at ads.openai.com, where advertisers build and manage sponsored cards that appear beneath relevant ChatGPT answers, clearly labeled as sponsored content.

Q: When did ChatGPT Ads launch in Europe?

A: OpenAI expanded ChatGPT Ads to 31 European countries starting August 24, 2026. Self-serve Ads Manager access for European advertisers opened on August 31, 2026.

Q: Which European countries have ChatGPT Ads?

A: The expansion covers 31 markets, including Denmark, Germany, France, Spain, Italy, Sweden, Norway, the Netherlands, Austria, and Switzerland.

Q: How much do ChatGPT Ads cost?

A: Self-serve pricing is currently around $25 CPM or $2 to $8 per click, with OpenAI suggesting a starting max CPC of $3 to $5. There's no minimum spend on self-serve, a sharp drop from the closed pilot's $60 CPM and $200,000 minimum.

Q: Are ChatGPT Ads better than Meta or Google Ads for e-commerce?

A: Not a replacement for either. Google captures explicit search intent, Meta captures discovery through behavioral targeting, and ChatGPT Ads sit in a decision-stage moment with no cookie-based targeting and immature attribution. It's a channel worth a small, capped test alongside Meta and Google, not a substitute for them yet.


Key Takeaways

  • ChatGPT Ads opened self-serve access for European advertisers on August 31, 2026, following an August 24 expansion into 31 countries, Denmark included.
  • Ads only show to Free and Go tier users, Plus, Pro, Business, Enterprise, and Edu accounts are ad-free, so the addressable audience is narrower than the "1 billion weekly users" headline suggests.
  • Targeting is entirely contextual, no cookies or pixels, matched to the live conversation, with a second-price auction and current self-serve pricing around $25 CPM or $2 to $8 per click.
  • The format is one sponsored card (title, short copy, one square image), a real constraint compared to Meta's multiple formats, and attribution is nowhere near Google or Meta's maturity yet.
  • Product feed ads (up to 1 million SKUs) are the detail that matters most for e-commerce brands, turning a clean product catalog into ad inventory without per-SKU creative work.
  • Test with a small, capped budget pulled from exploratory spend, not from a Meta line that's already converting, and don't restructure your 2026 plan around this yet.

Evaluating whether a new channel like this actually belongs in your media plan, or where your budget is better spent fixing what's already converting, is exactly the kind of call we make for e-commerce and DTC clients every week. Book a discovery call and we'll walk through what we'd test first for your account.

Ready to Scale Profitably?

Book your free discovery call and let us map out the next growth moves for your e-commerce brand.