ChatGPT ads UK explained by a Suffolk Webflow agency: who sees them, how targeting compares with Google and Meta, and who should wait. 4 verdicts inside.

ChatGPT now sells ads to UK advertisers, but for most Suffolk service firms the honest answer today is to wait, because targeting control matters more than price.
Picture a plumber in Bury St Edmunds with a small monthly ad budget. He's made up, but you probably know someone like him. A LinkedIn post has just told him ChatGPT ads are open to UK advertisers, and he's wondering whether to move the money by Friday.
Our answer is "not yet" for most local service firms. Price isn't the sticking point. As far as we can tell, you can't yet steer these ads round a 20-mile patch the way you can on Google.
We have a bias, and you should read the verdict with it in front of you. FutureProofs sells Webflow website plans with SEO included, so we do well when owners put effort into being found rather than renting attention. Ad management isn't in our published plans, so we're not pitching it here.
Yes, as far as we can tell. OpenAI has opened ChatGPT advertising to UK advertisers, but we haven't linked its announcements here, so we're not quoting launch dates or spend rules.
Check OpenAI's own ads pages for the current terms. The product is young. It changes month by month. Anything dated in a blog post, ours included, can go stale quickly.
Here's what we'd look for before spending a pound.
It's a young product. Google Ads has years of reporting, exclusions and bidding tools behind it. For the wider picture on AI answers and small-business search, we covered how Google's own AI answers changed small-business search in a separate post.
Expect a small, clearly labelled sponsored panel inside a conversation, shown to some ChatGPT users and not to others.
Think of a headline, a short description and a link to your landing page. Check OpenAI's current ad specs for the exact limits. That's all. You get a sentence or two to persuade someone to leave a chat they were enjoying.
Not everyone using ChatGPT will be shown your ad, and OpenAI's help pages set out who is included. Take a Sudbury accountancy firm selling to company directors and professionals. Some of those buyers may sit outside the group that sees ads, so your reachable audience could be smaller than the headline user numbers suggest.
OpenAI publishes its own policy on how ads relate to answers and on what advertisers can see. We can't test any of that from outside. Read the policy yourself before you rely on it.
A firm like that still does its real selling after the click, on its professional services website. The ad only opens the door.

Not with the control a local business is used to. We haven't found documentation showing location targeting to the standard Google Ads offers. If OpenAI has since published it, that changes our verdict.
On Google Ads you can draw a radius round Stowmarket or Colchester, limit it to office hours, and see the exact search phrases behind each click. On ChatGPT, as we read it, you describe the kind of conversation you'd like to appear beside, such as "people asking about boiler repairs". That's a topic. It says nothing about place.
Go back to our made-up plumber, who covers Bury St Edmunds and the villages around it. A homeowner in Leeds asking about boiler repairs is the right conversation and the wrong customer, and he'd still pay for the click. An untargeted click is money gone, whatever the invoice calls it.
If your business lives or dies by a postcode, ask which tool lets you say "only here". Our own take on web design for trades starts from the same idea. A site should speak to the 20 miles you actually drive.
ChatGPT ads give you inferred intent and, so far as we can see, thinner control. Google gives you stated intent and the most control. Meta interrupts people but has deep audience tools.
The bottom row matters most for a small firm. Google's search terms report is the quiet hero of nearly every profitable small account. You read what people typed, cut the phrases that waste money, and the account gets better each week. There's more on that side on our Google Ads page.
Meta sits in the middle. It works for reaching people who didn't know they needed you, and less well when someone needs a plumber tonight.

We can't give you a UK rate card, because we can't source one. Bid and budget guidance lives inside OpenAI's own ads tools, so read it there before planning a test. Treat any bid ceiling as a cap on what you offer, not a forecast of what you'll pay.
Confident CPM figures do the rounds on social media. We've left them out because none came with a named, dated source.
A small budget buys a handful of clicks, and a handful of clicks tells you almost nothing. You can't tell a bad ad from bad luck. By the time the result is readable, the month's money has gone. That's true of any new channel. It bites harder when you can't see search terms or narrow by place.
On Google you'd set the radius and the hours, add negative keywords, and check the search terms report on Tuesday morning. Cutting a losing term that day is how small accounts turn profitable. A channel with thinner reporting doesn't let you do that yet.
If an agency opens the account for you, ask for it to be in your business's name, so you keep the history if you ever part ways.
Add the time for tracking, a landing page and someone to read the results. It helps to know what a monthly plan already includes before you add another channel to the pile.
Local service firms should wait. National online shops can run a small, tracked test. Anyone unprofitable on Google or Meta should fix that first.
If you do test, keep it boring. Prove your tracking works first, with a test enquiry or order. Pick one product or service and run it for a month. Judge it on cost per enquiry, not clicks, and leave Google running so you've got something to compare against.
Shops should read our page on web design for online shops, which covers what a site needs before any paid traffic arrives. Shops are built on Shopify on a monthly plan from £495 a month, with the Shopify SEO included. For a local shop, start with our guide to taking a shop online.
A national shop with working tracking and a profitable Google account has more reason to test than the local firms we're telling to wait. If that's you, test it.
And our own alternative, being cited in AI answers, isn't instant. We can't promise it either. It runs slower than buying a click, and there's no dashboard to watch.

Yes, as far as we can tell. OpenAI has opened advertising to UK advertisers, so a small business can look at opening an account. Being allowed in doesn't make it a good fit. Before you put money in, work out how you'd target your own area and how you'd count enquiries.
We can't source a UK rate card, so we won't quote one. Any per-click or CPM figure you've seen on social media is unconfirmed unless it names a source and a date. Check the bid and budget guidance inside OpenAI's own ads tools before you plan a test.
Not everyone using it. OpenAI decides which users are shown ads, and its help pages set out the current rules. Your reachable audience is likely smaller than the headline user numbers, which matters if your buyers are professionals who may be outside that group.
Only if Google Ads isn't working. A channel that hands you search terms, a radius and a schedule is hard to beat for a local firm. With spare money, test ChatGPT on its own with a fixed cap, and leave the profitable account running for a clean comparison at the end.
That's for OpenAI to answer, and we can't audit it from outside. Its ads policy pages set out what it says about ads and answers, and about what advertisers can see. Read those before relying on anyone's summary, including ours.
Cheaper in cash, slower in time. Earned mentions can't be guaranteed, and they depend on clear, specific pages that answer real questions. An ad gives you a faster test and a dashboard. On a thin budget, building the pages first usually makes more sense than renting a click.
Check your tracking, your Google account and your own pages before you pay for a new channel. It's all free and takes an afternoon.
Open Google Ads and read the search terms report for the last 30 days. Note which phrases brought real enquiries and which brought nothing. Then ask ChatGPT the questions your customers ask, like "emergency plumber near Bury St Edmunds", and see who gets named and why. Our post on how to earn a ChatGPT citation explains what those answers tend to reward. If you're weighing a second channel, splitting a small ad budget is a useful companion read.
Then look at your service pages. Do they state prices, areas and timescales in plain words? Pages like that earn their keep whether the visitor arrives from an ad or an answer. If you're local, SEO in Bury St Edmunds shows how we think about it.
Send us your current site and the towns you cover. We'll tell you which pages are missing, even if you never hire us. Use our contact page, or see the 3 plans on our pricing page, from £189 a month.
Written by the FutureProofs team in Suffolk. Updated 10 Oct 2026.