Is ChatGPT Ads Worth It? What Early Testing Reveals

Q: Is ChatGPT Ads worth it?

Worth it if: you need to defend your brand footprint. Real people are already asking ChatGPT to compare named brands head-to-head. If you are invisible on your own comparison prompts, you are handing high-intent customers directly to competitors who do show up. Not worth it yet if: you are budget-constrained or held to strict, near-term ROAS targets. Attribution is entirely manual, there is zero backend revenue data, and no one can prove whether these CPCs actually beat Google for the same search intent. Which type of advertiser you are matters far more than which channel you choose.

Mery Hayles
2 days ago
7 min read

One Adthena account's real performance data, and a market-wide monitoring view across 14,000+ advertisers, point to the same answer, and to the same warning.

TLDR: Is ChatGPT Ads worth it?

Worth it if: you need to defend your brand footprint. Real people are already asking ChatGPT to compare named brands head-to-head. If you are invisible on your own comparison prompts, you are handing high-intent customers directly to competitors who do show up.

Not worth it yet if: you are budget-constrained or held to strict, near-term ROAS targets. Attribution is entirely manual, there is zero backend revenue data, and no one can prove whether these CPCs actually beat Google for the same search intent. Which type of advertiser you are matters far more than which channel you choose.

Key takeaways

40.2% of monitored, commercially relevant ChatGPT prompts returned a paid ad between March and July 2026.
Conversational, context-rich ad copy beat keyword-style copy by 39% relative CTR in our own account (1.07% vs 0.77%)
In a nine-day monitoring window (8 to 16 July 2026), the US had roughly 4.1x as many distinct advertisers running ChatGPT ads as the UK (5,260 vs 1,275), closely matching our account's own CPC-gap prediction
People are already asking ChatGPT to compare named brands head-to-head: real conversational demand, though whether it's incremental to Google is still an open question
Brands active generally can still be entirely absent from their own head-to-head comparison prompts, handing that moment to a named competitor instead. We saw this directly with a UK holiday park query in July
What's still unproven: revenue and ROAS impact. Reason enough to move cautiously if you're budget constrained or under strict ROAS pressure

The question isn't really "is it worth it" any more

Ask most PPC teams whether ChatGPT Ads is worth their time and you'll get a shrug: it's beta, the budgets are small, the reporting is thin. All fair. But that framing assumes the decision is still ahead of the market. It isn't. Across the commercially relevant prompts we monitor, 40.2% now return a paid ad inside ChatGPT. Whatever "worth it" means, a large chunk of the market has already answered the question with its wallet.

The better question is whether you're in that 40%, or watching it happen to your category from the outside. To answer that properly, we pulled together two very different vantage points: what it's actually like to run a live ChatGPT Ads account, and what our own monitoring sees happening market-wide, across ChatGPT ad activity we track. One is a single, detailed account. The other is a wide lens across thousands of advertisers. They turned out to agree with each other more than we expected.

What it's actually like once you're in the account

Joss Froggatt has been running Adthena's own ChatGPT Ads Manager account since the end of May. His conclusion after two months: "Twenty-five years of Google Ads logic does not map onto a conversational ad platform like ChatGPT Ads without friction." A few specifics worth sitting with:

Write for a conversation, not a query. In an A/B test on the account, a rich, conversational context hint returned a 1.07% CTR against 0.77% for a thinner, keyword-style list, a 39% relative lift. You can't exact-match a conversation, and the platform doesn't report back at the query level the way Google does: there's no search query report, and there likely never will be one in the form PPC teams are used to.

Daily budgets behave like a promise, not a suggestion. A $25/day minimum-budget test spent exactly $25.00 every single day, without the usual Google-style drift above or below the cap. It reads like a bug for the first couple of days; it isn't. The read is that ChatGPT Ads throttles delivery across the day so the total lands exactly on the cap, worth knowing, and worth noting that the same discipline doesn't appear to hold for lifetime budgets, where spend can move much faster with far less pacing control.

Reporting is still a primary-school data set. Impressions, clicks, cost, CPC, conversions: that's roughly it. Anything more exotic, like CPA or conversion rate, means building it yourself.

What we see watching everyone else do the same thing

That's the view from inside one account. Adthena's own monitoring gives us the view from outside all of them, tracking real ad creative, real prompts, and real advertisers across ChatGPT at a market level. A few things stand out.

The market is already crowded. Our tracked advertiser universe across ChatGPT now runs to more than 14,000 distinct advertisers, and in a nine-day snapshot alone (8 to 16 July 2026) we saw over 7,900 distinct advertisers actively serving ads. Big retail and travel names dominate the leaderboard by volume: Booking.com, Target, Home Depot, Zalando, and L.L.Bean topped our latest pull, with Wayfair and Best Buy also among the most active. Worth flagging: some brands that featured prominently in earlier monitoring, didn't show up at all in this window. That could mean a paused campaign, a shifted strategy, or simply the natural churn of a young, fast-moving platform.

These aren't abstract placements, they're landing in genuinely specific moments. In our latest pull, we saw Wayfair's ad for a Frigidaire stainless steel French door refrigerator served against the prompt "best french door refrigerator." We saw Home Depot's ad for a GE Profile ADA dishwasher served against "dishwasher capacity comparison guide." We saw Expedia's ad for Hawaii island packages served against "best inter island flights for travel." This isn't a future state, it's what ChatGPT users are already being shown today, mid-conversation, at the exact moment they're deciding what to buy.

The US/UK gap: one account guessed it, market data confirmed it

Here's where the two views line up in a way we didn't expect going in. Joss's account data showed a real CPC gap between markets ($4.52 average CPC in the US against $3.05 in the UK), and, from what he could see inside one account, estimated the US advertiser base at "around 300% bigger" than the UK's.

Pulling a nine-day window from our own market-wide monitoring (8 to 16 July 2026) rather than one account's view: the US had 5,260 distinct advertisers running ChatGPT ads against 1,275 in the UK, a 4.1x gap. That's about as close to an exact match to a single account's own estimate as you'll ever see, and it's a useful sanity check both ways: the account-level read wasn't a guess, and the market-wide number confirms the UK genuinely is an earlier-stage market, not just for one advertiser's experience, but across the board.

That gap won't last. The US platform has only had a few months' head start. Lighter competition right now is an opportunity with a shrinking window, not a permanent feature of the market.

A different kind of demand, and the risk of missing it

There's a reasonable case that ChatGPT Ads isn't simply Google Ads wearing a new interface. It may be capturing a different moment in the decision journey, not just the same searchers in a new box. The prompts we track aren't keyword fragments, they're often full comparison questions asked in the exact language a shopper would use out loud like "best family holiday parks hoseasons vs parkdean." That's someone actively weighing named brands against each other in one conversation, in a format Google's results page doesn't really replicate.

We want to be careful here: we can't yet prove this is an incremental audience. Untangling whether these are the same searchers simply showing up in a new channel, or a genuinely different pool of people reaching for ChatGPT before they'd ever open Google, is a harder question than either of our data sources can answer alone. That needs proper cross-channel audience research, not a single account or a monitoring snapshot. What we can say is that the behavior is real and already happening at volume, and it changes what "missing out" means.

That's where the sharper, more measurable pattern in our market-wide data comes in. Brands that are clearly active in ChatGPT ads generally can still be completely absent from the exact prompts where someone is comparing them by name against a named competitor, arguably the highest-intent moment a prompt can capture. We saw this directly in our July data: for the prompt "airbnb vs booking.com", neither Airbnb nor Booking.com appeared, a guidebook add-on brand did instead. In both cases, the named brands weren't losing that moment on price or ad creative. They lost it because they never showed up to compete for it at all.

When it's not worth it yet

None of the above should read as an unqualified "get in now." If direct conversions are the job and budget is tight, or you're held to a strict near-term ROAS number, the honest read is: not yet, at least not with serious spend.

We have engagement data, not revenue data. The platform's own reporting stops at impressions, clicks, cost, CPC, and conversions: no query-level report, no CPA or true conversion rate without downloading everything and building it yourself. So attribution is thin by construction, and everything in this piece, the CTR lift, the CPCs, the ad-frequency stat, sits above the point where a click becomes a sale. Neither Joss's account testing nor our market monitoring can currently show what any of this does to actual revenue. That's the gap that decides "worth it" for most finance teams, and right now it has to be answered with your own conversion tracking, not with anything in this article.

Cost is unproven, not necessarily high. A $4.52 US CPC and $3.05 UK CPC aren't obviously expensive or cheap in isolation, that depends on category, and on how those numbers stack up against what you already pay on Google for the same intent. Neither of our sources has that side-by-side benchmark yet, so treat any "ChatGPT Ads is cheaper or more expensive than Google" claim you hear right now, including implicitly from this piece, as unproven either way.

Add to that a platform still in beta (lifetime budgets reportedly pace far more loosely than daily ones, and today's CPCs and advertiser counts are a snapshot, not a stable baseline to plan a year of spend around), and holding off, if defending this quarter's CAC is the actual job, is a defensible call.

In summary

So, is ChatGPT Ads worth it? The honest answer splits in two, and both halves are backed by real data now, not guesses.

If you're worried about competitors quietly picking up customers, revenue, or lifetime value from you on a new discovery platform: worth it. Real people are already asking ChatGPT to compare named brands head-to-head, over 14,000 advertisers are already active in our tracked universe, and brands invisible on their own comparison prompts are handing that moment to whoever does show up instead.

If you're optimizing for direct conversions right now and budget is tight, or you're held to a strict near-term ROAS target: not worth it yet, at least not with serious spend. Nobody, including us, can currently show you the revenue behind these numbers - tracking and attribution is still very limited.

Most PPC teams are some mix of both. The honest move isn't picking a side in the abstract, it's being clear-eyed about which one your business actually is before deciding how much budget this gets.